1<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01//EN" "http://www.w3.org/TR/html4/strict.dtd"> 2<html> 3 <head> 4 <meta charset="windows-1252" /> 5 <meta http-equiv="Content-Type" content="text/html; charset=windows-1252" /> 6 <meta name="viewport" content="width=device-width, initial-scale=1" /> 7 <title>WSC 2023 Proceedings</title> 8 <link href="includes/css/jquery-ui.css" rel="stylesheet" type="text/css" /> 9 <link 10 href="includes/css/shared_styles.css" 11 rel="stylesheet" 12 type="text/css" 13 /> 14 <link 15 href="includes/css/block_styles.css?v=1" 16 rel="stylesheet" 17 type="text/css" 18 /> 19 <link 20 href="includes/css/jquery.qtip.min.css" 21 rel="stylesheet" 22 type="text/css" 23 /> 24 <link 25 href="includes/css/font-awesome-4.1.0/css/font-awesome.min.css" 26 rel="stylesheet" 27 type="text/css" 28 /> 29 <link 30 href="includes/css/user_generated.css" 31 rel="stylesheet" 32 type="text/css" 33 /> 34 <link href="archive_styles.css" rel="stylesheet" type="text/css" /> 35 <style> 36 div.banner_top, 37 div.banner_top .site_title, 38 div.banner_top .no_logo_banner_right, 39 div.logo_banner, 40 div.logo_banner .user_name, 41 #header { 42 background-color: #0066cc; 43 color: #ffffff; 44 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 45 sans-serif; 46 font-size: 15px; 47 text-transform: none; 48 } 49 50 div.logo_banner .site_title a, 51 div.banner_top .site_title a, 52 #header #site_title a { 53 color: #ffffff; 54 text-decoration: none; 55 } 56 57 .documentation_box { 58 background: #f0e0bc; 59 color: #000000; 60 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 61 sans-serif; 62 font-size: 15px; 63 } 64 65 .filter_bar, 66 .filter_bar_w_legend { 67 background-color: #c9ddf9; 68 } 69 .filter_bar_w_legend .instr, 70 .filter_bar .instr { 71 background-color: #a2c2fc; 72 } 73 74 #footer { 75 background-color: #eaeaea; 76 color: #999999; 77 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 78 sans-serif; 79 font-size: 15px; 80 text-transform: none; 81 } 82 #footer a { 83 color: #777777; 84 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 85 sans-serif; 86 font-size: 15px; 87 text-transform: none; 88 } 89 90 .contents .input .title, 91 .contents .input_box .title { 92 background-color: #247bf4; 93 } 94 .contents .input, 95 .contents .input_box, 96 .contents .input table tr th, 97 .contents .input_box table tr th, 98 .form .block-content { 99 background-color: #f6f9fe; 100 } 101 .contents .input .instr, 102 .contents .input_box .instr, 103 .multi_block_button, 104 .form .block .instr { 105 background-color: #dbe8fa; 106 } 107 .contents .input .odd, 108 .contents .input_box .odd, 109 .form .block-content .odd { 110 background-color: #edf3fc; 111 } 112 .contents .input .even, 113 .contents .input_box .even, 114 .form .block-content .even { 115 background-color: #dbe8fa; 116 } 117 118 #actions_col .block-title { 119 background-color: #244a84; 120 color: #ffffff; 121 } 122 #actions_col .block-title a { 123 color: #ffffff; 124 } 125 #actions_col .block-content { 126 background-color: #bdd2f8; 127 } 128 #actions_col .block-content .instr { 129 background-color: #b0cbfc; 130 } 131 #actions_col .block-content .odd { 132 background-color: #c2d6fb; 133 } 134 #actions_col .block-content .even { 135 background: #b0cbfc; 136 } 137 138 body.in_iframe { 139 background-color: #fffff7; 140 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 141 sans-serif; 142 font-size: 15px; 143 } 144 .pagedoc { 145 background-color: #fffff7; 146 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 147 sans-serif; 148 font-size: 15px; 149 } 150 .page_box { 151 background-color: #fffff7; 152 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 153 sans-serif; 154 font-size: 15px; 155 } 156 .page_box_in_iframe { 157 background-color: #fffff7; 158 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 159 sans-serif; 160 font-size: 15px; 161 } 162 .contents { 163 background-color: #fffff7; 164 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 165 sans-serif; 166 font-size: 15px; 167 } 168 .contents_options { 169 background-color: #fffff7; 170 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 171 sans-serif; 172 font-size: 15px; 173 } 174 #top-links { 175 background-color: #fffff7; 176 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 177 sans-serif; 178 font-size: 15px; 179 } 180 .fullscreen { 181 background-color: #fffff7; 182 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 183 sans-serif; 184 font-size: 15px; 185 } 186 .subtabs .fg_tab { 187 border-bottom-color: #fffff7; 188 } 189 .subtabs .fg_tab div { 190 background-color: #fffff7; 191 border-bottom-color: #fffff7; 192 } 193 .fullscreen_schedule { 194 background-color: #fffff7; 195 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 196 sans-serif; 197 font-size: 15px; 198 } 199 .contents input, 200 .contents input_box, 201 .contents textarea { 202 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 203 sans-serif; 204 font-size: 15px; 205 } 206 207 /*Possible fix for buttons using the wrong font -Nathan*/ 208 /*:not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) > button, 209 input[type=button], 210 input[type=reset], 211 input[type=submit] { 212 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, sans-serif !important; 213 }*/ 214 215 .qtip.rm-qtip { 216 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 217 sans-serif; 218 font-size: 15px; 219 } 220 #cboxContent { 221 background-color: #fffff7; 222 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 223 sans-serif; 224 font-size: 15px; 225 } 226 227 /* For now, use the main site background color for tool tips. */ 228 .qtip.qtip-rm, 229 .qtip.qtip-rm .qtip-titlebar { 230 background-color: #fffff7; 231 } 232 233 /* Not sure where this should live. */ 234 #actions_col .block-title-text { 235 font-size: 15px; 236 } 237 #related_col .block-title-text { 238 font-size: 15px; 239 } 240 241 /* For jquery-ui. */ 242 .ui-widget { 243 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 244 sans-serif; 245 font-size: 15px; 246 } 247 248 .arrow-slidedown { 249 background-color: #fffff7; 250 color: #000000; 251 } 252 253 .contents .output_box .title { 254 background-color: #d89655; 255 } 256 .block-content, 257 .block2-content, 258 .contents .output_box table tr th, 259 .contents .output_box { 260 background-color: #f9f4e6; 261 } 262 .output_box_instr, 263 .contents .output_box .instr, 264 .block .instr, 265 .block-content .instr { 266 background-color: #f2e0c5; 267 } 268 .odd, 269 .contents .output_box .odd, 270 .block-content .odd { 271 background-color: #f6ead5; 272 } 273 .even, 274 .contents .output_box .even, 275 .block-content .even { 276 background-color: #f2e0c5; 277 } 278 279 a:link, 280 a:visited, 281 a:active, 282 .clickable, 283 a.clickable, 284 a.clickable:link, 285 a.clickable:visited, 286 a.clickable:active, 287 .ttip_object_info_blue, 288 .ttip_object_info_blue_no_clone, 289 .ttip_object_info_blue_wide, 290 .ttip_object_info_blue_wide_no_clone, 291 .ttip_object_info_blue_very_wide, 292 .ttip_object_info_blue_very_wide_no_clone, 293 .ttip_object_info_blue_extra_wide, 294 .ttip_object_info_blue_extra_wide_no_clone, 295 .ttip_object_info_blue_modal, 296 .ttip_object_info_blue_modal_no_clone, 297 .colorbox_object_info, 298 span.menu_item_label, 299 .page_box_print .contents A, 300 .page_box_print #footer a { 301 color: #0000ee; 302 } 303 304 /* Light Links */ 305 .light_link a, 306 .light_arrow, 307 .light_link a:link, 308 .light_link a:active, 309 .light_link a:visited, 310 .light_clickable, 311 a.light_clickable, 312 a.light_clickable:link, 313 a.light_clickable:active, 314 a.light_clickable:visited { 315 color: #5088f0; 316 } 317 318 /* user hovers */ 319 a:hover, 320 .light_link a:hover, 321 .light_arrow:hover, 322 .light_clickable:hover, 323 a.light_clickable:hover, 324 .hover_link:hover, 325 .ttip_object_info_blue:hover, 326 .ttip_object_info_blue_no_clone:hover, 327 .ttip_object_info_blue_wide:hover, 328 .ttip_object_info_blue_wide_no_clone:hover, 329 .ttip_object_info_blue_very_wide:hover, 330 .ttip_object_info_blue_very_wide_no_clone:hover, 331 .ttip_object_info_blue_extra_wide:hover, 332 .ttip_object_info_blue_extra_wide_no_clone:hover, 333 .ttip_object_info_blue_modal:hover, 334 .ttip_object_info_blue_modal_no_clone:hover, 335 .ttip_object_info:hover, 336 .ttip_object_info_no_clone:hover, 337 .ttip_object_info_wide:hover, 338 .ttip_object_info_wide_no_clone:hover, 339 .ttip_object_info_very_wide:hover, 340 .ttip_object_info_very_wide_no_clone:hover, 341 .ttip_object_info_extra_wide:hover, 342 .ttip_object_info_extra_wide_no_clone:hover, 343 .ttip_object_info_modal:hover, 344 .ttip_object_info_modal_no_clone:hover, 345 .colorbox_object_info:hover, 346 .subtabs .fg_tab:hover div, 347 .subtabs .fg_tab:hover A, 348 .subtabs .bg_tab:hover, 349 .subtabs .bg_tab:hover A { 350 color: #0000ee; 351 } 352 353 ul.rm_mega_menu li.mega > div, 354 ul.rm_mega_menu > li.mega-link > a:hover, 355 .disp_details_header, 356 .disp_details_sub_header, 357 .disp_details I, /* This is deprecated, since it clashes with font awesome using I tags. */ 358 .disp_red, 359 .disp_label { 360 color: #b32626; 361 } 362 363 div.active_toggle_button { 364 background-color: #5088f0; 365 } 366 367 /* Default button */ 368 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 369 > button, 370 input[type="button"], 371 input[type="reset"], 372 input[type="submit"] { 373 border-color: #0000ee; 374 color: #0000ee; 375 } 376 377 /* Small, big, and save buttons */ 378 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 379 > button.small-button, 380 input[type="button"].small-button, 381 input[type="reset"].small-button, 382 input[type="submit"].small-button, 383 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 384 > button.big-button, 385 input[type="button"].big-button, 386 input[type="reset"].big-button, 387 input[type="submit"].big-button, 388 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 389 > button.save-button, 390 input[type="button"].save-button, 391 input[type="reset"].save-button, 392 input[type="submit"].save-button { 393 background-color: #0000ee; 394 } 395 396 /* Light button */ 397 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 398 > button.light-button, 399 input[type="button"].light-button, 400 input[type="reset"].light-button, 401 input[type="submit"].light-button { 402 border-color: #5088f0; 403 color: #5088f0; 404 } 405 406 /* Light-save button */ 407 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 408 > button.light-save-button, 409 input[type="button"].light-save-button, 410 input[type="reset"].light-save-button, 411 input[type="submit"].light-save-button { 412 border-color: #5088f0; 413 background-color: #5088f0; 414 } 415 416 /* Default button hover */ 417 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 418 > button:hover, 419 input[type="button"]:hover, 420 input[type="reset"]:hover, 421 input[type="submit"]:hover { 422 background-color: #0000ee; 423 } 424 425 /* Light button hover */ 426 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 427 > button:hover.light-button, 428 input[type="button"]:hover.light-button, 429 input[type="reset"]:hover.light-button, 430 input[type="submit"]:hover.light-button { 431 background-color: #5088f0; 432 } 433 434 /* Default button disabled */ 435 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 436 > button[disabled], 437 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 438 > button[disabled]:hover, 439 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 440 > button[disabled]:active, 441 input[type="button"][disabled], 442 input[type="button"][disabled]:hover, 443 input[type="button"][disabled]:active, 444 input[type="reset"][disabled], 445 input[type="reset"][disabled]:hover, 446 input[type="reset"][disabled]:active, 447 input[type="submit"][disabled], 448 input[type="submit"][disabled]:hover, 449 input[type="submit"][disabled]:active { 450 color: #0000ee; 451 } 452 453 /* Big, save buttons disabled */ 454 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 455 > button[disabled].big-button, 456 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 457 > button[disabled]:hover.big-button, 458 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 459 > button[disabled]:active.big-button, 460 input[type="button"][disabled].big-button, 461 input[type="button"][disabled]:hover.big-button, 462 input[type="button"][disabled]:active.big-button, 463 input[type="reset"][disabled].big-button, 464 input[type="reset"][disabled]:hover.big-button, 465 input[type="reset"][disabled]:active.big-button, 466 input[type="submit"][disabled].big-button, 467 input[type="submit"][disabled]:hover.big-button, 468 input[type="submit"][disabled]:active.big-button, 469 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 470 > button[disabled].save-button, 471 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 472 > button[disabled]:hover.save-button, 473 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 474 > button[disabled]:active.save-button, 475 input[type="button"][disabled].save-button, 476 input[type="button"][disabled]:hover.save-button, 477 input[type="button"][disabled]:active.save-button, 478 input[type="reset"][disabled].save-button, 479 input[type="reset"][disabled]:hover.save-button, 480 input[type="reset"][disabled]:active.save-button, 481 input[type="submit"][disabled].save-button, 482 input[type="submit"][disabled]:hover.save-button, 483 input[type="submit"][disabled]:active.save-button { 484 background-color: #0000ee; 485 } 486 487 /* Light button disabled */ 488 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 489 > button[disabled].light-button, 490 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 491 > button[disabled]:hover.light-button, 492 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 493 > button[disabled]:active.light-button, 494 input[type="button"][disabled].light-button, 495 input[type="button"][disabled]:hover.light-button, 496 input[type="button"][disabled]:active.light-button, 497 input[type="reset"][disabled].light-button, 498 input[type="reset"][disabled]:hover.light-button, 499 input[type="reset"][disabled]:active.light-button, 500 input[type="submit"][disabled].light-button, 501 input[type="submit"][disabled]:hover.light-button, 502 input[type="submit"][disabled]:active.light-button { 503 color: #5088f0; 504 } 505 506 /* Light-save button disabled */ 507 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 508 > button[disabled].light-save-button, 509 :not(.mce-btn):not(.mce-window-head):not(.ui
509-datepicker-buttonpane) 510 > button[disabled]:hover.light-save-button, 511 :not(.mce-btn):not(.mce-window-head):not(.ui-datepicker-buttonpane) 512 > button[disabled]:active.light-save-button, 513 input[type="button"][disabled].light-save-button, 514 input[type="button"][disabled]:hover.light-save-button, 515 input[type="button"][disabled]:active.light-save-button, 516 input[type="reset"][disabled].light-save-button, 517 input[type="reset"][disabled]:hover.light-save-button, 518 input[type="reset"][disabled]:active.light-save-button, 519 input[type="submit"][disabled].light-save-button, 520 input[type="submit"][disabled]:hover.light-save-button, 521 input[type="submit"][disabled]:active.light-save-button { 522 background-color: #5088f0; 523 } 524 525 #related_col .block-title { 526 background-color: #bbbbbb; 527 color: #000000; 528 } 529 #related_col .block-title a { 530 color: #0000ff; 531 } 532 #related_col .block-content { 533 background-color: #e5e5e5; 534 } 535 #related_col .block-content .instr { 536 background-color: #d0d0d0; 537 } 538 #related_col .block-content .odd { 539 background-color: #dedede; 540 } 541 #related_col .block-content .even { 542 background: #d0d0d0; 543 } 544 545 .role_stype_bar { 546 background-color: #e0ab76; 547 color: #000000; 548 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 549 sans-serif; 550 font-size: 15px; 551 text-transform: none; 552 } 553 554 .tabs .fg_tab { 555 background-color: #f0dbbc; 556 color: #cc1a1a; 557 border-bottom-color: #f0dbbc; 558 } 559 .tab_menu_label:hover, 560 .active .tab_menu_label, 561 .tab_no_menu_label:hover, 562 .tab_no_menu_label:hover a { 563 color: #cc1a1a; 564 } 565 .subtabs { 566 background-color: #f0dbbc; 567 color: #000000; 568 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 569 sans-serif; 570 font-size: 15px; 571 text-transform: none; 572 } 573 .subtabs .divider, 574 .subtabs .bg_tab, 575 .subtabs .bg_tab a { 576 background-color: #f0dbbc; 577 border-top-color: #f0dbbc; 578 color: #000000; 579 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 580 sans-serif; 581 font-size: 15px; 582 text-transform: none; 583 } 584 .subtabs .fg_tab { 585 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 586 sans-serif; 587 font-size: 15px; 588 color: #cc1a1a; 589 text-transform: none; 590 } 591 592 /* 593 uncomment this to make the subtabs follow the selected tab color instead 594 of the link color 595 596 .subtabs .bg_tab:hover a { 597 color: #CC1A1A; 598 } 599 .subtabs .fg_tab:hover a, { 600 color: #CC1A1A; 601 } 602 */ 603 604 .subtabs .fg_tab a { 605 color: #cc1a1a; 606 } 607 608 .tabs { 609 background: #e8ca9b; 610 } 611 .tabs .divider, 612 .tabs .bg_tab { 613 background-color: #e8ca9b; 614 border-top-color: #e8ca9b; 615 color: #000000; 616 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 617 sans-serif; 618 font-size: 15px; 619 text-transform: none; 620 } 621 .tabs .fg_tab { 622 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 623 sans-serif; 624 font-size: 15px; 625 color: #cc1a1a; 626 text-transform: none; 627 } 628 .tab_menu_label, 629 .tab_no_menu_label { 630 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 631 sans-serif; 632 font-size: 15px; 633 } 634 .qtip.qtip-rm-tab-menu { 635 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 636 sans-serif; 637 font-size: 15px; 638 } 639 640 div.rm_mega_menus_container, 641 ul.rm_mega_menu.darker, 642 ul.rm_mega_menu > li.mega > a, 643 ul.rm_mega_menu.darker > li.mega > a, 644 ul.rm_mega_menu > li.mega-link > a, 645 ul.rm_mega_menu.darker > li.mega-link > a, 646 ul.rm_mega_menu > li.mega-label > span { 647 background-color: #777777; 648 border-color: #777777; 649 color: #ffffff; 650 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 651 sans-serif; 652 font-size: 15px; 653 text-transform: none; 654 } 655 ul.rm_mega_menu > li.mega.selected > a { 656 background-color: #ffffff; 657 border-color: #ffffff; 658 color: #777777; 659 } 660 ul.rm_mega_menu > li.mega:hover > a { 661 color: #cc1a1a; 662 } 663 div.rm_mega_menus_container 664 ul.rm_mega_menu 665 > li.mega 666 > div.menu_dropdown { 667 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 668 sans-serif; 669 font-size: 15px; 670 } 671 .contents input, 672 .contents input_box, 673 .contents textarea { 674 font-family: Optima, Helvetica, Verdana, "Lucida Grande", Arial, 675 sans-serif; 676 font-size: 15px; 677 } 678 </style> 679
679<script 680 src="includes/jquery/jquery-1.8.3.min.js" 681 type="text/javascript" 682 ></script>
682 683
683<script 684 src="includes/jquery/jquery-ui.min.js" 685 type="text/javascript" 686 ></script>
686 687
687<script 688 src="includes/jquery/jquery.hoverIntent.minified.js" 689 type="text/javascript" 690 ></script>
690 691
691<script src="includes/jquery/basic.js" type="text/javascript"></script>
691 692
692<script 693 src="includes/jquery/jquery.qtip.min.js" 694 type="text/javascript" 695 ></script>
695 696
696<script src="includes/jquery/jquery.highlight-4.js"></script>
696 697
697<script src="includes/jquery/jquery.instaFilter.js"></script>
697 698
698<script type="text/javascript"> 699 $(function () { 700 setup_info_links(); 701 }); 702 </script>
702 703
703<script src="includes/jquery/user_generated.js"></script>
703 704
704<script type="text/javascript"> 705 $(document).ready(function () { 706 $program_sections = $("#sections-container"); 707 $("#program_filter").InstaFilter($program_sections, { 708 typing_pause: 500, 709 search_unit_selector: ".slot-entry", 710 search_unit_ancestor_fields_selector: 711 ".session-title, .session-chair, .room-name", 712 }); 713 714 // this ensures there is enough vertical space below the search field to keep the 715 // page from scrolling during a search 716 $program_sections.css("min-height", $(window).height()); 717 }); 718 </script>
718 719 </head> 720 <body> 721 <a name="top" tabindex="-1"></a> 722 <div class="centered"> 723 <div class="header"> 724 <div class="header-logo" style="float: left"></div> 725 <div 726 class="header-title" 727 style=" 728 float: left; 729 margin-top: 15px; 730 height: 80px; 731 font-size: 30px; 732 font-weight: bold; 733 " 734 > 735 <span class="page-title">WSC 2023 Proceedings</span> 736 </div> 737 <div style="clear: both"></div> 738 </div> 739 </div> 740 <br /> 741 <div class="centered nav-links"> 742 <br /><span class="page-links" 743 ><a href="at_a_glance.html">Overview</a></span 744 > 745 | <span class="page-links">By Program Track</span> | 746 <span class="page-links"><a href="by_auth.html">Author Index</a></span 747 ><br /> 748 </div> 749 <br /> 750 <div id="main-content-box"> 751 <div class="centered"> 752 <table class="cellspacing10px" role="presentation"> 753 <tr> 754 <td> 755 <div class="anchor-link"> 756 <a href="#ptrack101">Advanced Tutorials</a> 757 </div> 758 </td> 759 <td> 760 <div class="anchor-link"> 761 <a href="#ptrack128">Logistics Supply Chains Transportation</a> 762 </div> 763 </td> 764 <td> 765 <a href="#ptrack119">Simulation and Artificial Intelligence</a> 766 </td> 767 </tr> 768 <tr> 769 <td> 770 <div class="anchor-link"> 771 <a href="#ptrack111">Agent-based Simulation</a> 772 </div> 773 </td> 774 <td> 775 <div class="anchor-link"> 776 <a href="#ptrack123">Manufacturing and Industry 4.0</a> 777 </div> 778 </td> 779 <td> 780 <div class="anchor-link"> 781 <a href="#ptrack120">Simulation as Digital Twin</a> 782 </div> 783 </td> 784 </tr> 785 <tr> 786 <td> 787 <div class="anchor-link"> 788 <a href="#ptrack103">Analysis Methodology</a> 789 </div> 790 </td> 791 <td> 792 <div class="anchor-link"> 793 <a href="#ptrack124">MASM: Semiconductor Manufacturing</a> 794 </div> 795 </td> 796 <td> 797 <div class="anchor-link"> 798 <a href="#ptrack106">Simulation in Education</a> 799 </div> 800 </td> 801 </tr> 802 <tr> 803 <td> 804 <div class="anchor-link"> 805 <a href="#ptrack110">Aviation Modeling and Analysis</a> 806 </div> 807 </td> 808 <td> 809 <div class="anchor-link"> 810 <a href="#ptrack108" 811 >Military and National Security Applications</a 812 > 813 </div> 814 </td> 815 <td> 816 <div class="anchor-link"> 817 <a href="#ptrack121">Simulation Optimization</a> 818 </div> 819 </td> 820 </tr> 821 <tr> 822 <td> 823 <div class="anchor-link"> 824 <a href="#ptrack112">Complex and Resilient Systems</a> 825 </div> 826 </td> 827 <td> 828 <div class="anchor-link"> 829 <a href="#ptrack117">Modeling Methodology</a> 830 </div> 831 </td> 832 <td> 833 <div class="anchor-link"> 834 <a href="#ptrack122" 835 >Uncertainty Quantification and Robust Simulation</a 836 > 837 </div> 838 </td> 839 </tr> 840 <tr> 841 <td> 842 <div class="anchor-link"> 843 <a href="#ptrack104">Data Science for Simulation</a> 844 </div> 845 </td> 846 <td><a href="#ptrack131">Professional Development</a></td> 847 <td> 848 <div class="anchor-link"><a href="#ptrack133">Vendor</a></div> 849 </td> 850 </tr> 851 <tr> 852 <td> 853 <div class="anchor-link"> 854 <a href="#ptrack129" 855 >Environment Sustainability and Resilience</a 856 > 857 </div> 858 </td> 859 <td> 860 <div class="anchor-link"> 861 <a href="#ptrack118">Project Management and Construction</a> 862 </div> 863 </td> 864 <td> 865 <div class="anchor-link"><a href="#ptrack132">Plenary</a></div> 866 </td> 867 </tr> 868 <tr> 869 <td> 870 <div class="anchor-link"> 871 <a href="#ptrack102">Introductory Tutorials</a> 872 </div> 873 </td> 874 <td> 875 <div class="anchor-link"> 876 <a href="#ptrack109">Reliability Modeling and Simulation</a> 877 </div> 878 </td> 879 <td> 880 <div class="anchor-link"><a href="#ptrack138">Poster</a></div> 881 </td> 882 </tr> 883 <tr> 884 <td> 885 <div class="anchor-link"> 886 <a href="#ptrack116">Healthcare and Life Sciences</a> 887 </div> 888 </td> 889 <td> 890 <div class="anchor-link"> 891 <a href="#ptrack107">Scientific Applications</a> 892 </div> 893 </td> 894 <td> 895 <div class="anchor-link"> 896 <a href="#ptrack139">PhD Colloquium</a> 897 </div> 898 </td> 899 </tr> 900 <tr> 901 <td> 902 <div class="anchor-link"> 903 <a href="#ptrack105">
903Hybrid Simulation</a> 904 </div> 905 </td> 906 <td> 907 <div class="anchor-link"> 908 <a href="#ptrack130">Simulation Around the World</a> 909 </div> 910 </td> 911 <td> 912 <div class="anchor-link"><a href="#other">Other</a></div> 913 </td> 914 </tr> 915 </table> 916 <br /> 917 918 <div class="centered"> 919 <br /><span class="page-links" 920 ><strong 921 ><a href="/wsc23papers/wsc2023-papers.zip" 922 >Download All Papers</a 923 ></strong 924 ></span 925 > 926 927 <span class="page-links" 928 ><strong 929 ><a href="/wsc23papers/wsc2023-case.zip" 930 >Download All Case Studies</a 931 ></strong 932 ></span 933 ><br /> 934 </div> 935 936 <hr /> 937 </div> 938 <div class="righted"> 939 <input 940 id="program_filter" 941 name="program_filter" 942 placeholder="search" 943 size="40" 944 type="text" 945 /> 946 </div> 947 <div class="centered" id="sections-container"> 948 <table role="presentation"> 949 <tr> 950 <td align="left"> 951 <div class="area-section"> 952 <div class="centered"> 953 <a name="ptrack132" tabindex="-1"></a> 954 <div class="section-title">Plenary</div> 955 </div> 956 <div class="section-entry"> 957 <div class="session-entry"> 958 <span class="session-event-type">Plenary</span 959 ><span class="type-track-spacer"> · </span 960 ><span class="program-track">Plenary</span><br /> 961 <div class="session-title"> 962 Opening Plenary: Modeling for Energy Resilience: How DOE 963 Uses Simulation to Model and Manage Everything from the 964 Power Grid to the Strategic Petroleum Reserve 965 </div> 966 <div class="session-chair"> 967 Chair: Bahar Biller (SAS Institute, Inc)<br /> 968 </div> 969 <div class="slot-entry"> 970 <a name="prog101" tabindex="-1"></a> 971 <div class="slot-title-line"> 972 <span class="slot-title" 973 >Modeling for Energy Resilience: How DOE Uses 974 Simulation to Model and Manage Everything from the 975 Power Grid to the Strategic Petroleum Reserve</span 976 > 977 </div> 978 <div class="slot-authors"> 979 Ann Dunkin (Department of Energy) 980 </div> 981 <div class="slot-abstract"> 982 <div> 983 <a 984 class="clickable no-decoration" 985 id="vhsjs_view_758_1707793552_880812" 986 onclick="$('#vhsjs_view_758_1707793552_880812').hide(); 987 $('#vhsjs_hide_758_1707793552_880812').show(); 988 $('#757_1707793552_880804').slideDown(function() { 989 if (typeof Masonry === 'function') { 990 $('.use_masonry').masonry(); 991 }; 992 993 });" 994 ><i class="fa fa-caret-right"></i> 995 <span class="hover_link">Abstract</span></a 996 ><a 997 class="clickable no-decoration" 998 id="vhsjs_hide_758_1707793552_880812" 999 onclick="$('#757_1707793552_880804').hide(function() { 1000 if (typeof Masonry === 'function') { 1001 $('.use_masonry').masonry(); 1002 }; 1003 }); 1004 $('#vhsjs_hide_758_1707793552_880812').hide(); 1005 $('#vhsjs_view_758_1707793552_880812').show();" 1006 style="display: none" 1007 ><i class="fa fa-caret-down"></i>
1008 <span class="hover_link">Abstract</span></a 1009 > 1010 <div 1011 data-display-control="758_1707793552_880812" 1012 id="757_1707793552_880804" 1013 style="display: none" 1014 > 1015 <div class="arrow-slidedown"> 1016 <blockquote> 1017 The U.S. Department of Energy’s 1018 responsibilities run the gamut from managing the 1019 nuclear stockpile and the strategic petroleum 1020 reserve to running the power grid in 36 states 1021 to performing basic and applied research to 1022 protect national security, ensure stable power 1023 sector operations and accelerate the clean 1024 energy transition. Leveraging the power of 1025 DOE’s computing infrastructure, including 1026 the world’s fastest supercomputer, 1027 simulation models are used to accelerate 1028 advancements in nearly every field of research 1029 across DOE. Through a series of examples 1030 highlighting grid management, cybersecurity, 1031 cavern modeling and fundamental physical 1032 phenomena, this keynote will illuminate how DOE 1033 applies modeling and simulation to both research 1034 and operations. 1035 </blockquote> 1036 </div> 1037 </div> 1038 </div> 1039 </div> 1040 <div class="slot-urls"></div> 1041 <a href="/wsc23papers/prog101.pdf" target="_blank">pdf</a 1042 ><br /> 1043 1044 <div> 1045 <br /><iframe 1046 src="https://player.vimeo.com/video/894605270?h=aaf53c335d" 1047 width="640" 1048 height="360" 1049 frameborder="0" 1050 allow="autoplay; fullscreen; picture-in-picture" 1051 allowfullscreen 1052 ></iframe> 1053 <p> 1054 <a href="https://vimeo.com/894605270" 1055 >Welcome and Keynote</a 1056 > 1057 from 1058 <a href="https://vimeo.com/user139157832">INFORMS</a> 1059 on <a href="https://vimeo.com">Vimeo</a>. 1060 </p> 1061 </div> 1062 </div> 1063 </div> 1064 <div class="session-entry"> 1065 <span class="session-event-type">Plenary</span 1066 ><span class="type-track-spacer"> · </span 1067 ><span class="program-track">Plenary</span><br /> 1068 <div class="session-title"> 1069 Titans of Simulation: Resilience of Supply Chains and the 1070 Role of Simulation 1071 </div> 1072 <div class="session-chair"> 1073 Chair: John Shortle (George Mason University)<br /> 1074 </div> 1075 <div class="slot-entry"> 1076 <a name="prog104" tabindex="-1"></a> 1077 <div class="slot-title-line"> 1078 <span class="slot-title" 1079 >Resilience of Supply Chains and the Role of 1080 Simulation</span 1081 > 1082 </div> 1083 <div class="slot-authors"> 1084 John Fowler (Arizona State University) 1085 </div> 1086 <div class="slot-abstract"> 1087 <div> 1088 <a 1089 class="clickable no-decoration" 1090 id="vhsjs_view_760_1707793552_8862867" 1091 onclick="$('#vhsjs_view_760_1707793552_8862867').hide(); 1092 $('#vhsjs_hide_760_1707793552_8862867').show(); 1093 $('#759_1707793552_8862786').slideDown(function() { 1094 if (typeof Masonry === 'function') { 1095 $('.use_masonry').masonry(); 1096 }; 1097 1098 });" 1099 ><i class="fa fa-caret-right"></i>
1100 <span class="hover_link">Abstract</span></a 1101 ><a 1102 class="clickable no-decoration" 1103 id="vhsjs_hide_760_1707793552_8862867" 1104 onclick="$('#759_1707793552_8862786').hide(function() { 1105 if (typeof Masonry === 'function') { 1106 $('.use_masonry').masonry(); 1107 }; 1108 }); 1109 $('#vhsjs_hide_760_1707793552_8862867').hide(); 1110 $('#vhsjs_view_760_1707793552_8862867').show();" 1111 style="display: none" 1112 ><i class="fa fa-caret-down"></i> 1113 <span class="hover_link">Abstract</span></a 1114 > 1115 <div 1116 data-display-control="760_1707793552_8862867" 1117 id="759_1707793552_8862786" 1118 style="display: none" 1119 > 1120 <div class="arrow-slidedown"> 1121 <blockquote> 1122 Supply chain resilience refers to the capacity 1123 of a supply chain to proactively prepare for 1124 unforeseen events, effectively address 1125 disruptions, and bounce back from them while 1126 ensuring the sustained smooth operation of the 1127 supply chain at the preferred level of 1128 connectivity and management of its structure and 1129 functions. Recent disruptive events including 1130 the Covid-19 pandemic and the Russian invasion 1131 of Ukraine have caused an increased emphasis on 1132 supply chain resilience. In this presentation, 1133 we discuss strategies to prepare for, address, 1134 and bounce back from (potential) disruptions and 1135 the role that simulation can play in enhancing 1136 supply chain resilience. 1137 </blockquote> 1138 </div> 1139 </div> 1140 </div> 1141 </div> 1142 <div class="slot-urls"></div> 1143 <a href="/wsc23papers/prog104.pdf" target="_blank">pdf</a 1144 ><br /> 1145 1146 <div> 1147 <br /><iframe 1148 src="https://player.vimeo.com/video/912587418?h=4462675fc2" 1149 width="640" 1150 height="360" 1151 frameborder="0" 1152 allow="autoplay; fullscreen; picture-in-picture" 1153 allowfullscreen 1154 ></iframe> 1155 <p> 1156 <a href="https://vimeo.com/912587418" 1157 >WSC 2023 Titan of Simulation - John Fowler</a 1158 > 1159 from 1160 <a href="https://vimeo.com/user139157832">INFORMS</a> 1161 on <a href="https://vimeo.com">Vimeo</a>. 1162 </p> 1163 </div> 1164 </div> 1165 </div> 1166 <div class="session-entry"> 1167 <span class="session-event-type">Plenary</span 1168 ><span class="type-track-spacer"> · </span 1169 ><span class="program-track">Plenary</span><br /> 1170 <div class="session-title"> 1171 Titans of Simulation: Ensuring Food Security under Climate 1172 Change: How Simulation Can Help in Making Agricultural 1173 Supply Chains More Resilient 1174 </div> 1175 <div class="session-chair"> 1176 Chair: John Shortle (George Mason University)<br /> 1177 </div> 1178 <div class="slot-entry"> 1179 <a name="prog103" tabindex="-1"></a> 1180 <div class="slot-title-line">
1181 <span class="slot-title" 1182 >Ensuring Food Security under Climate Change: How 1183 Simulation Can Help in Making Agricultural Supply 1184 Chains More Resilient</span 1185 > 1186 </div> 1187 <div class="slot-authors"> 1188 Enver Yücesan (INSEAD) 1189 </div> 1190 <div class="slot-abstract"> 1191 <div> 1192 <a 1193 class="clickable no-decoration" 1194 id="vhsjs_view_762_1707793552_8895886" 1195 onclick="$('#vhsjs_view_762_1707793552_8895886').hide(); 1196 $('#vhsjs_hide_762_1707793552_8895886').show(); 1197 $('#761_1707793552_8895802').slideDown(function() { 1198 if (typeof Masonry === 'function') { 1199 $('.use_masonry').masonry(); 1200 }; 1201 1202 });" 1203 ><i class="fa fa-caret-right"></i> 1204 <span class="hover_link">Abstract</span></a 1205 ><a 1206 class="clickable no-decoration" 1207 id="vhsjs_hide_762_1707793552_8895886" 1208 onclick="$('#761_1707793552_8895802').hide(function() { 1209 if (typeof Masonry === 'function') { 1210 $('.use_masonry').masonry(); 1211 }; 1212 }); 1213 $('#vhsjs_hide_762_1707793552_8895886').hide(); 1214 $('#vhsjs_view_762_1707793552_8895886').show();" 1215 style="display: none" 1216 ><i class="fa fa-caret-down"></i> 1217 <span class="hover_link">Abstract</span></a 1218 > 1219 <div 1220 data-display-control="762_1707793552_8895886" 1221 id="761_1707793552_8895802" 1222 style="display: none" 1223 > 1224 <div class="arrow-slidedown"> 1225 <blockquote> 1226 Climate change and the resulting increased 1227 frequency of unpredictable extreme weather 1228 events create new operational challenges for the 1229 commercial seed industry, which is a key pillar 1230 of a sustainable and secure global food supply. 1231 More specifically, extreme weather events 1232 translate into two main effects on agricultural 1233 production: Higher yield variability and lower 1234 expected yields. In recent years, extreme 1235 weather events already caused reductions in the 1236 yields of cereals, maize, and other staple 1237 crops. It is also projected that a warming of 1238 +2C (+4C) would increase the coefficient of 1239 variation of corn yield by 62% (192%) in six 1240 countries that collectively account for 73% of 1241 global production. In this presentation, we 1242 first examine how the increased likelihood of 1243 extreme weather events affects agricultural 1244 supply chains in terms of R&D, production 1245 planning, contracting, allocation, and storage 1246 decisions. We then discuss the key challenges 1247 associated with each stage and highlight how 1248 simulation can help address them under increased 1249 volatility. 1250 </blockquote> 1251 </div> 1252 </div> 1253 </div> 1254 </div> 1255 <div class="slot-urls"></div> 1256 <a href="/wsc23papers/prog103.pdf" target="_blank">pdf</a 1257 ><br /> 1258 <div> 1259 <br /><iframe 1260 src="https://player.vimeo.com/video/912590810?h=1dceac718e" 1261 width="640" 1262 height="360" 1263 frameborder="0" 1264 allow="autoplay; fullscreen; picture-in-picture" 1265 allowfullscreen 1266 ></iframe> 1267 <p> 1268 <a href="https://vimeo.com/912590810" 1269 >WSC 2023 Titan of Simulation - Enver Yucesan</a 1270 > 1271 from 1272 <a href="https://vimeo.com/user139157832">INFORMS</a> 1273 on <a href="https://vimeo.com">Vimeo</a>. 1274 </p> 1275 </div> 1276 </div> 1277 </div> 1278 </div> 1279 <div class="centered"> 1280 <div class="top-link"><a href="#top">Return to Top</a></div> 1281 </div> 1282 <hr /> 1283 </div> 1284 1285 <div class="area-section"> 1286 <div class="centered"> 1287 <a name="ptrack101" tabindex="-1"></a> 1288 <div class="section-title">Advanced Tutorials</div> 1289 </div> 1290 <div class="centered track-chair">
1291 <span class="track-chair-role" 1292 >Track Coordinator - Advanced Tutorials: </span 1293 ><span class="track-chair-names" 1294 >Henry Lam (Columbia University), Giulia Pedrielli (Arizona 1295 State University)</span 1296 > 1297 </div> 1298 <div class="section-entry"> 1299 <div class="session-entry"> 1300 <span class="session-event-type">Tutorial</span 1301 ><span class="type-track-spacer"> · </span 1302 ><span class="program-track">Advanced Tutorials</span><br /> 1303 <div class="session-title"> 1304 Screening Simulated Systems for Optimization 1305 </div> 1306 <div class="session-chair"> 1307 Chair: Eunhye Song (Georgia Institute of Technology)<br /> 1308 </div> 1309 <div class="slot-entry"> 1310 <a name="inv192" tabindex="-1"></a> 1311 <div class="slot-authors"> 1312 Jinbo Zhao (Texas A&M University), Javier Gatica 1313 (Pontificia Universidad Catolica de Chile), and David 1314 Eckman (Texas A&M University) 1315 </div> 1316 <div class="slot-abstract"> 1317 <div> 1318 <a 1319 class="clickable no-decoration" 1320 id="vhsjs_view_2_1707793550_8328831" 1321 onclick="$('#vhsjs_view_2_1707793550_8328831').hide(); 1322 $('#vhsjs_hide_2_1707793550_8328831').show(); 1323 $('#1_1707793550_8328671').slideDown(function() { 1324 if (typeof Masonry === 'function') { 1325 $('.use_masonry').masonry(); 1326 }; 1327 1328 });" 1329 ><i class="fa fa-caret-right"></i> 1330 <span class="hover_link">Abstract</span></a 1331 ><a 1332 class="clickable no-decoration" 1333 id="vhsjs_hide_2_1707793550_8328831" 1334 onclick="$('#1_1707793550_8328671').hide(function() { 1335 if (typeof Masonry === 'function') { 1336 $('.use_masonry').masonry(); 1337 }; 1338 }); 1339 $('#vhsjs_hide_2_1707793550_8328831').hide(); 1340 $('#vhsjs_view_2_1707793550_8328831').show();" 1341 style="display: none" 1342 ><i class="fa fa-caret-down"></i> 1343 <span class="hover_link">Abstract</span></a 1344 > 1345 <div 1346 data-display-control="2_1707793550_8328831" 1347 id="1_1707793550_8328671" 1348 style="display: none" 1349 > 1350 <div class="arrow-slidedown"> 1351 <blockquote> 1352 Screening procedures for ranking and selection 1353 have received less attention than selection 1354 procedures, yet they serve as a cheap and 1355 powerful tool for decision making under 1356 uncertainty. Research on screening procedures 1357 has been less active in recent years, just as 1358 the advent of parallel computing has 1359 dramatically reshaped how selection procedures 1360 are designed and implemented. As a result, 1361 screening procedures used in modern practice 1362 continue to largely operate offline on fixed 1363 data. In this tutorial, we provide an overview 1364 of screening procedures with the goal of 1365 clarifying the current state of research and 1366 laying out opportunities for future development. 1367 We discuss several guarantees delivered by 1368 screening procedures and their role in different 1369 decision-making settings and investigate their 1370 impact on screening power and sampling 1371 efficiency in numerical experiments. We also 1372 study the implementation of screening procedures 1373 in parallel computing environments and how they 1374 can be combined with selection procedures. 1375 </blockquote> 1376 </div> 1377 </div> 1378 </div> 1379 </div> 1380 <div class="slot-urls"></div> 1381 <a href="/wsc23papers/001.pdf" target="_blank">pdf</a 1382 ><br /> 1383 </div> 1384 </div> 1385 <div class="session-entry">
1386 <span class="session-event-type">Tutorial</span 1387 ><span class="type-track-spacer"> · </span 1388 ><span class="program-track">Advanced Tutorials</span><br /> 1389 <div class="session-title"> 1390 Practical Impact and Academia Are Not Antonyms 1391 </div> 1392 <div class="session-chair"> 1393 Chair: Russell R. Barton (Pennsylvania State 1394 University)<br /> 1395 </div> 1396 <div class="slot-entry"> 1397 <a name="inv130" tabindex="-1"></a> 1398 <div class="slot-authors"> 1399 Shane Henderson (Cornell University) 1400 </div> 1401 <div class="slot-abstract"> 1402 <div> 1403 <a 1404 class="clickable no-decoration" 1405 id="vhsjs_view_4_1707793551_0719159" 1406 onclick="$('#vhsjs_view_4_1707793551_0719159').hide(); 1407 $('#vhsjs_hide_4_1707793551_0719159').show(); 1408 $('#3_1707793551_0719018').slideDown(function() { 1409 if (typeof Masonry === 'function') { 1410 $('.use_masonry').masonry(); 1411 }; 1412 1413 });" 1414 ><i class="fa fa-caret-right"></i> 1415 <span class="hover_link">Abstract</span></a 1416 ><a 1417 class="clickable no-decoration" 1418 id="vhsjs_hide_4_1707793551_0719159" 1419 onclick="$('#3_1707793551_0719018').hide(function() { 1420 if (typeof Masonry === 'function') { 1421 $('.use_masonry').masonry(); 1422 }; 1423 }); 1424 $('#vhsjs_hide_4_1707793551_0719159').hide(); 1425 $('#vhsjs_view_4_1707793551_0719159').show();" 1426 style="display: none" 1427 ><i class="fa fa-caret-down"></i> 1428 <span class="hover_link">Abstract</span></a 1429 > 1430 <div 1431 data-display-control="4_1707793551_0719159" 1432 id="3_1707793551_0719018" 1433 style="display: none" 1434 > 1435 <div class="arrow-slidedown"> 1436 <blockquote> 1437 This tutorial discusses principles and 1438 strategies for the interplay between applied 1439 work with organizations and an academic research 1440 agenda. I emphasize lessons I have learned 1441 through my own work and my own mistakes, with 1442 special focus on some high-stakes settings, 1443 including advising Cornell University’s 1444 response to the COVID-19 pandemic and work with 1445 the emergency services, among other 1446 applications. 1447 </blockquote> 1448 </div> 1449 </div> 1450 </div> 1451 </div> 1452 <div class="slot-urls"></div> 1453 <a href="/wsc23papers/002.pdf" target="_blank">pdf</a 1454 ><br /> 1455 </div> 1456 </div> 1457 <div class="session-entry"> 1458 <span class="session-event-type">Tutorial</span 1459 ><span class="type-track-spacer"> · </span 1460 ><span class="program-track">Advanced Tutorials</span><br /> 1461 <div class="session-title"> 1462 Statistical Limit Theorems in Distributionally Robust 1463 Optimization 1464 </div> 1465 <div class="session-chair"> 1466 Chair: Henry Lam (Columbia University)<br /> 1467 </div> 1468 <div class="slot-entry"> 1469 <a name="inv214" tabindex="-1"></a> 1470 <div class="slot-authors"> 1471 Jose Blanchet (Stanford University) and Alexander 1472 Shapiro (Georgia Institute of Technology) 1473 </div> 1474 <div class="slot-abstract"> 1475 <div> 1476 <a 1477 class="clickable no-decoration" 1478 id="vhsjs_view_6_1707793551_0812967" 1479 onclick="$('#vhsjs_view_6_1707793551_0812967').hide(); 1480 $('#vhsjs_hide_6_1707793551_0812967').show(); 1481 $('#5_1707793551_0812864').slideDown(function() { 1482 if (typeof Masonry === 'function') { 1483 $('.use_masonry').masonry(); 1484 }; 1485 1486 });" 1487 ><i class="fa fa-caret-right"></i>
1488 <span class="hover_link">Abstract</span></a 1489 ><a 1490 class="clickable no-decoration" 1491 id="vhsjs_hide_6_1707793551_0812967" 1492 onclick="$('#5_1707793551_0812864').hide(function() { 1493 if (typeof Masonry === 'function') { 1494 $('.use_masonry').masonry(); 1495 }; 1496 }); 1497 $('#vhsjs_hide_6_1707793551_0812967').hide(); 1498 $('#vhsjs_view_6_1707793551_0812967').show();" 1499 style="display: none" 1500 ><i class="fa fa-caret-down"></i> 1501 <span class="hover_link">Abstract</span></a 1502 > 1503 <div 1504 data-display-control="6_1707793551_0812967" 1505 id="5_1707793551_0812864" 1506 style="display: none" 1507 > 1508 <div class="arrow-slidedown"> 1509 <blockquote> 1510 The goal of this paper is to develop a 1511 methodology for the systematic analysis of 1512 asymptotic statistical properties of data-driven 1513 DRO formulations based on their corresponding 1514 non-DRO counterparts. We illustrate our approach 1515 in various settings, including both 1516 phi-divergence and Wasserstein uncertainty sets. 1517 Different types of asymptotic behaviors are 1518 obtained depending on the rate at which the 1519 uncertainty radius decreases to zero as a 1520 function of the sample size and the geometry of 1521 the uncertainty sets. 1522 </blockquote> 1523 </div> 1524 </div> 1525 </div> 1526 </div> 1527 <div class="slot-urls"></div> 1528 <a href="/wsc23papers/003.pdf" target="_blank">pdf</a 1529 ><br /> 1530 </div> 1531 </div> 1532 <div class="session-entry"> 1533 <span class="session-event-type">Tutorial</span 1534 ><span class="type-track-spacer"> · </span 1535 ><span class="program-track">Advanced Tutorials</span><br /> 1536 <div class="session-title"> 1537 Digital Twins: Features, Models, and Services 1538 </div> 1539 <div class="session-chair"> 1540 Chair: Feng Ju (Arizona State University)<br /> 1541 </div> 1542 <div class="slot-entry"> 1543 <a name="inv200" tabindex="-1"></a> 1544 <div class="slot-authors"> 1545 Andrea Matta (Politecnico di Milano, Via La Masa 1) and 1546 Giovanni Lugaresi (KU Leuven) 1547 </div> 1548 <div class="slot-abstract"> 1549 <div> 1550 <a 1551 class="clickable no-decoration" 1552 id="vhsjs_view_8_1707793551_0887475" 1553 onclick="$('#vhsjs_view_8_1707793551_0887475').hide(); 1554 $('#vhsjs_hide_8_1707793551_0887475').show(); 1555 $('#7_1707793551_088738').slideDown(function() { 1556 if (typeof Masonry === 'function') { 1557 $('.use_masonry').masonry(); 1558 }; 1559 1560 });" 1561 ><i class="fa fa-caret-right"></i> 1562 <span class="hover_link">Abstract</span></a 1563 ><a 1564 class="clickable no-decoration" 1565 id="vhsjs_hide_8_1707793551_0887475" 1566 onclick="$('#7_1707793551_088738').hide(function() { 1567 if (typeof Masonry === 'function') { 1568 $('.use_masonry').masonry(); 1569 }; 1570 }); 1571 $('#vhsjs_hide_8_1707793551_0887475').hide(); 1572 $('#vhsjs_view_8_1707793551_0887475').show();" 1573 style="display: none" 1574 ><i class="fa fa-caret-down"></i>
1575 <span class="hover_link">Abstract</span></a 1576 > 1577 <div 1578 data-display-control="8_1707793551_0887475" 1579 id="7_1707793551_088738" 1580 style="display: none" 1581 > 1582 <div class="arrow-slidedown"> 1583 <blockquote> 1584 This work provides an overview of digital twins, 1585 digital replicas of real entities conceived to 1586 support analysis, improvements, and optimal 1587 decisions. Specifically, it aims to better 1588 clarify what digital twins are by pointing out 1589 their main features, what they can do to support 1590 their related physical twins, and which models 1591 they use. An illustrative example together with 1592 a few selected application examples is used to 1593 better describe digital twins. A discussion on 1594 the actual challenges and research opportunities 1595 is also reported. 1596 </blockquote> 1597 </div> 1598 </div> 1599 </div> 1600 </div> 1601 <div class="slot-urls"></div> 1602 <a href="/wsc23papers/004.pdf" target="_blank">pdf</a 1603 ><br /> 1604 </div> 1605 </div> 1606 <div class="session-entry"> 1607 <span class="session-event-type">Tutorial</span 1608 ><span class="type-track-spacer"> · </span 1609 ><span class="program-track">Advanced Tutorials</span><br /> 1610 <div class="session-title"> 1611 Bootstrapping and Batching for Output Analysis 1612 </div> 1613 <div class="session-chair"> 1614 Chair: Sara Shashaani (North Carolina State University)<br /> 1615 </div> 1616 <div class="slot-entry"> 1617 <a name="inv190" tabindex="-1"></a> 1618 <div class="slot-authors"> 1619 Raghu Pasupathy (Purdue University) 1620 </div> 1621 <div class="slot-abstract"> 1622 <div> 1623 <a 1624 class="clickable no-decoration" 1625 id="vhsjs_view_10_1707793551_0952969" 1626 onclick="$('#vhsjs_view_10_1707793551_0952969').hide(); 1627 $('#vhsjs_hide_10_1707793551_0952969').show(); 1628 $('#9_1707793551_0952864').slideDown(function() { 1629 if (typeof Masonry === 'function') { 1630 $('.use_masonry').masonry(); 1631 }; 1632 1633 });" 1634 ><i class="fa fa-caret-right"></i> 1635 <span class="hover_link">Abstract</span></a 1636 ><a 1637 class="clickable no-decoration" 1638 id="vhsjs_hide_10_1707793551_0952969" 1639 onclick="$('#9_1707793551_0952864').hide(function() { 1640 if (typeof Masonry === 'function') { 1641 $('.use_masonry').masonry(); 1642 }; 1643 }); 1644 $('#vhsjs_hide_10_1707793551_0952969').hide(); 1645 $('#vhsjs_view_10_1707793551_0952969').show();" 1646 style="display: none" 1647 ><i class="fa fa-caret-down"></i> 1648 <span class="hover_link">Abstract</span></a 1649 > 1650 <div
1651 data-display-control="10_1707793551_0952969" 1652 id="9_1707793551_0952864" 1653 style="display: none" 1654 > 1655 <div class="arrow-slidedown"> 1656 <blockquote> 1657 We review bootstrapping and batching as devices 1658 for statistical inference in simulation output 1659 analysis. Bootstrapping, discovered in the late 1660 1970s and developed over the ensuing three 1661 decades, is widely held as being among the 1662 important scientific discoveries of the previous 1663 century due primarily to its facility for 1664 general statistical inference. By contrast, 1665 batching was introduced in the 1960s but was 1666 developed within the simulation community (in 1667 the 1980s) for the narrower contexts of variance 1668 parameter estimation and confidence interval 1669 construction. In recent years, however, there 1670 has been increasing realization that batching, 1671 much like bootstrapping, can be used also for 1672 general statistical inference, and that batching 1673 often compares favorably with bootstrapping in 1674 dependent data contexts. Bootstrapping and 1675 batching have tremendous applicability for 1676 uncertainty quantification in simulation, and 1677 are prime candidates for adoption in simulation 1678 software. We describe the general principles 1679 underlying bootstrapping and batching, outline 1680 guarantees, and discuss implementation. 1681 </blockquote> 1682 </div> 1683 </div> 1684 </div> 1685 </div> 1686 <div class="slot-urls"></div> 1687 <a href="/wsc23papers/005.pdf" target="_blank">pdf</a 1688 ><br /> 1689 </div> 1690 </div> 1691 <div class="session-entry"> 1692 <span class="session-event-type">Tutorial</span 1693 ><span class="type-track-spacer"> · </span 1694 ><span class="program-track">Advanced Tutorials</span><br /> 1695 <div class="session-title"> 1696 Coarse-Grained Simulations of DNA and RNA Systems with 1697 oxDNA and oxRNA Models: Tutorial 1698 </div> 1699 <div class="session-chair"> 1700 Chair: Wei Xie (Northeastern University)<br /> 1701 </div> 1702 <div class="slot-entry"> 1703 <a name="inv198" tabindex="-1"></a> 1704 <div class="slot-authors"> 1705 Matthew Sample, Michael Matthies, and Petr Sulc (Arizona 1706 State University) 1707 </div> 1708 <div class="slot-abstract"> 1709 <div> 1710 <a 1711 class="clickable no-decoration" 1712 id="vhsjs_view_12_1707793551_102048" 1713 onclick="$('#vhsjs_view_12_1707793551_102048').hide(); 1714 $('#vhsjs_hide_12_1707793551_102048').show(); 1715 $('#11_1707793551_1020386').slideDown(function() { 1716 if (typeof Masonry === 'function') { 1717 $('.use_masonry').masonry(); 1718 }; 1719 1720 });" 1721 ><i class="fa fa-caret-right"></i>
1722 <span class="hover_link">Abstract</span></a 1723 ><a 1724 class="clickable no-decoration" 1725 id="vhsjs_hide_12_1707793551_102048" 1726 onclick="$('#11_1707793551_1020386').hide(function() { 1727 if (typeof Masonry === 'function') { 1728 $('.use_masonry').masonry(); 1729 }; 1730 }); 1731 $('#vhsjs_hide_12_1707793551_102048').hide(); 1732 $('#vhsjs_view_12_1707793551_102048').show();" 1733 style="display: none" 1734 ><i class="fa fa-caret-down"></i> 1735 <span class="hover_link">Abstract</span></a 1736 > 1737 <div 1738 data-display-control="12_1707793551_102048" 1739 id="11_1707793551_1020386" 1740 style="display: none" 1741 > 1742 <div class="arrow-slidedown"> 1743 <blockquote> 1744 We present a tutorial on setting-up the oxDNA 1745 coarse-grained model for simulations of DNA and 1746 RNA nanotechnology. The model is a popular tool 1747 used both by theorists and experimentalists to 1748 simulate nucleic acid systems both in biology 1749 and nanotechnology settings. The tutorial is 1750 aimed at new users asking "Where should I start 1751 if I want to use oxDNA". We assume no prior 1752 background in using the model. This tutorial 1753 shows basic examples that can get a novice user 1754 started with the model, and points the 1755 prospective user towards additional reading and
1756 online resources depending on which aspect of 1757 the model they are interested in pursuing. 1758 </blockquote> 1759 </div> 1760 </div> 1761 </div> 1762 </div> 1763 <div class="slot-urls"></div> 1764 <a href="/wsc23papers/006.pdf" target="_blank">pdf</a 1765 ><br /> 1766 </div> 1767 </div> 1768 <div class="session-entry"> 1769 <span class="session-event-type">Tutorial</span 1770 ><span class="type-track-spacer"> · </span 1771 ><span class="program-track">Advanced Tutorials</span><br /> 1772 <div class="session-title"> 1773 Importance Sampling Strategy for Heavy-tailed Systems with 1774 Catastrophe Principle 1775 </div> 1776 <div class="session-chair"> 1777 Chair: Henry Lam (Columbia University)<br /> 1778 </div> 1779 <div class="slot-entry"> 1780 <a name="inv213" tabindex="-1"></a> 1781 <div class="slot-title-line"> 1782 <span class="slot-title" 1783 >Importance Sampling Strategy for Heavy-Tailed Systems 1784 with Catastrophe Principle</span 1785 > 1786 </div> 1787 <div class="slot-authors"> 1788 Xingyu Wang and Chang-Han Rhee (Northwestern University) 1789 </div> 1790 <div class="slot-abstract"> 1791 <div> 1792 <a 1793 class="clickable no-decoration" 1794 id="vhsjs_view_14_1707793551_1087675" 1795 onclick="$('#vhsjs_view_14_1707793551_1087675').hide(); 1796 $('#vhsjs_hide_14_1707793551_1087675').show(); 1797 $('#13_1707793551_1087577').slideDown(function() { 1798 if (typeof Masonry === 'function') { 1799 $('.use_masonry').masonry(); 1800 }; 1801 1802 });" 1803 ><i class="fa fa-caret-right"></i> 1804 <span class="hover_link">Abstract</span></a 1805 ><a 1806 class="clickable no-decoration" 1807 id="vhsjs_hide_14_1707793551_1087675" 1808 onclick="$('#13_1707793551_1087577').hide(function() { 1809 if (typeof Masonry === 'function') { 1810 $('.use_masonry').masonry(); 1811 }; 1812 }); 1813 $('#vhsjs_hide_14_1707793551_1087675').hide(); 1814 $('#vhsjs_view_14_1707793551_1087675').show();" 1815 style="display: none" 1816 ><i class="fa fa-caret-down"></i> 1817 <span class="hover_link">Abstract</span></a 1818 > 1819 <div 1820 data-display-control="14_1707793551_1087675" 1821 id="13_1707793551_1087577" 1822 style="display: none" 1823 > 1824 <div class="arrow-slidedown"> 1825 <blockquote> 1826 Large deviations theory has a long history of 1827 providing powerful machinery for designing 1828 efficient rare-event simulation techniques. 1829 However, traditional large deviations theory 1830 fails to provide useful bounds in heavy-tailed 1831 contexts, and designing efficient rare-event 1832 simulation algorithms for heavy-tailed systems 1833 has been considered challenging. Recent 1834 developments in the theory of heavy-tailed large 1835 deviations enable designing a strongly efficie
1835nt 1836 importance sampling scheme that is universally 1837 applicable to a wide range of rare events. This 1838 tutorial aims to provide an accessible overview 1839 of the recent developments in the large 1840 deviations theory for heavy-tailed stochastic 1841 processes, which is followed by a detailed 1842 account of the design principle behind the 1843 strongly efficient importance sampling scheme 1844 for such processes. The implementations of the 1845 general principle are demonstrated through a few 1846 specific heavy-tailed rare events that arise in 1847 stochastic approximation, finance, and queueing 1848 theory contexts. 1849 </blockquote> 1850 </div> 1851 </div> 1852 </div> 1853 </div> 1854 <div class="slot-urls"></div> 1855 <a href="/wsc23papers/007.pdf" target="_blank">pdf</a 1856 ><br /> 1857 </div> 1858 </div> 1859 </div> 1860 <div class="centered"> 1861 <div class="top-link"><a href="#top">Return to Top</a></div> 1862 </div> 1863 <hr /> 1864 </div> 1865 <div class="area-section"> 1866 <div class="centered"> 1867 <a name="ptrack111" tabindex="-1"></a> 1868 <div class="section-title">Agent-based Simulation</div> 1869 </div> 1870 <div class="centered track-chair"> 1871 <span class="track-chair-role" 1872 >Track Coordinator - Agent-Based Simulation: </span 1873 ><span class="track-chair-names" 1874 >Andrew J. Collins (Old Dominion University), Chris Kuhlman 1875 (University of Virginia)</span 1876 > 1877 </div> 1878 <div class="section-entry"> 1879 <div class="session-entry"> 1880 <span class="session-event-type">Technical Session</span 1881 ><span class="type-track-spacer"> · </span 1882 ><span class="program-track">Agent-based Simulation</span 1883 ><br /> 1884 <div class="session-title"> 1885 Military and Homeland Security Agent-based Modeling 1886 </div> 1887 <div class="session-chair"> 1888 Chair: Berry Gerrits (University of Twente)<br /> 1889 </div> 1890 <div class="slot-entry"> 1891 <a name="con136" tabindex="-1"></a> 1892 <div class="slot-title-line"> 1893 <span class="slot-title" 1894 >Squashing Bugs and Improving Design: Using Data 1895 Farming to Support Verification and Validation of 1896 Military Agent-Based Simulations</span 1897 > 1898 </div> 1899 <div class="slot-authors"> 1900 Susan K. Aros and Mary L. McDonald (Naval Postgraduate 1901 School) 1902 </div> 1903 <div class="slot-abstract"> 1904 <div> 1905 <a 1906 class="clickable no-decoration" 1907 id="vhsjs_view_16_1707793551_2844257" 1908 onclick="$('#vhsjs_view_16_1707793551_2844257').hide(); 1909 $('#vhsjs_hide_16_1707793551_2844257').show(); 1910 $('#15_1707793551_2844098').slideDown(function() { 1911 if (typeof Masonry === 'function') { 1912 $('.use_masonry').masonry(); 1913 }; 1914 1915 });" 1916 ><i class="fa fa-caret-right"></i>
1917 <span class="hover_link">Abstract</span></a 1918 ><a 1919 class="clickable no-decoration" 1920 id="vhsjs_hide_16_1707793551_2844257" 1921 onclick="$('#15_1707793551_2844098').hide(function() { 1922 if (typeof Masonry === 'function') { 1923 $('.use_masonry').masonry(); 1924 }; 1925 }); 1926 $('#vhsjs_hide_16_1707793551_2844257').hide(); 1927 $('#vhsjs_view_16_1707793551_2844257').show();" 1928 style="display: none" 1929 ><i class="fa fa-caret-down"></i> 1930 <span class="hover_link">Abstract</span></a 1931 > 1932 <div 1933 data-display-control="16_1707793551_2844257" 1934 id="15_1707793551_2844098" 1935 style="display: none" 1936 > 1937 <div class="arrow-slidedown"> 1938 <blockquote> 1939 Verification and validation of complex 1940 agent-based human behavior simulation models is 1941 a challenging endeavor, particularly since a 1942 dearth of real-world data makes it impossible to 1943 use most traditional validation methods. Data 1944 farming techniques have stepped up to the 1945 challenge, proving to be a valuable tool for 1946 verification and validation of complex models. 1947 In this paper we demonstrate how data farming 1948 and analysis aids in the verification and 1949 validation of complex models by presenting 1950 specific examples pertaining to WRENCH, an 1951 agent-based simulation model that represents 1952 complex interactions between security forces and 1953 civilians during civil security stability 1954 operations. We first provide an overview of data 1955 farming and its relevance for verification and 1956 validation of military agent-based simulation 1957 models, then give an overview of WRENCH, and 1958 finally demonstrate with examples how we have 1959 used data farming to aid in the verification and 1960 validation of WRENCH. 1961 </blockquote> 1962 </div> 1963 </div> 1964 </div> 1965 </div> 1966 <div class="slot-urls"></div> 1967 <a href="/wsc23papers/008.pdf" target="_blank">pdf</a 1968 ><br /> 1969 </div> 1970 <div class="slot-entry"> 1971 <a name="con277" tabindex="-1"></a> 1972 <div class="slot-title-line"> 1973 <span class="slot-title" 1974 >Beyond Accuracy: Cybersecurity Resilience Evaluation 1975 of Intrusion Detection System against DoS Attacks 1976 using Agent-based Simulation</span 1977 > 1978 </div> 1979 <div class="slot-authors"> 1980 Jeongkeun Shin, Geoffrey B. Dobson, L. Richard Carley, 1981 and Kathleen M. Carley (Carnegie Mellon University) 1982 </div> 1983 <div class="slot-abstract"> 1984 <div> 1985 <a 1986 class="clickable no-decoration" 1987 id="vhsjs_view_18_1707793551_312476" 1988 onclick="$('#vhsjs_view_18_1707793551_312476').hide(); 1989 $('#vhsjs_hide_18_1707793551_312476').show(); 1990 $('#17_1707793551_3124676').slideDown(function() { 1991 if (typeof Masonry === 'function') { 1992 $('.use_masonry').masonry(); 1993 }; 1994 1995 });" 1996 ><i class="fa fa-caret-right"></i>
1997 <span class="hover_link">Abstract</span></a 1998 ><a 1999 class="clickable no-decoration" 2000 id="vhsjs_hide_18_1707793551_312476" 2001 onclick="$('#17_1707793551_3124676').hide(function() { 2002 if (typeof Masonry === 'function') { 2003 $('.use_masonry').masonry(); 2004 }; 2005 }); 2006 $('#vhsjs_hide_18_1707793551_312476').hide(); 2007 $('#vhsjs_view_18_1707793551_312476').show();" 2008 style="display: none" 2009 ><i class="fa fa-caret-down"></i> 2010 <span class="hover_link">Abstract</span></a 2011 > 2012 <div 2013 data-display-control="18_1707793551_312476" 2014 id="17_1707793551_3124676" 2015 style="display: none" 2016 > 2017 <div class="arrow-slidedown"> 2018 <blockquote> 2019 Machine Learning has become increasingly popular 2020 in developing Intrusion Detection Systems (IDS) 2021 for cybersecurity. However, the focus has mainly 2022 been on achieving high detection accuracy rather 2023 than evaluating the impact on cybersecurity 2024 resiliency. In this paper, we use agent-based 2025 simulation to investigate the impact of 2026 different IDS algorithms on the cybersecurity 2027 resiliency of organizations under DoS attacks. 2028 Our simulation includes a server agent equipped 2029 with either Naive Bayes or SMO-based IDS, and a 2030 cybercriminal agent capable of launching 2031 different types of Denial of Service attacks. 2032 Our results suggest that the choice of IDS 2033 algorithm can significantly affect an 2034 organization’s cybersecurity resiliency 2035 against DoS attacks. Specifically, while SMO 2036 shows better overall accuracy on the KDD Cup 2037 1999 dataset, Naive Bayes-based IDS proves more 2038 effective in practice due to its better-balanced 2039 detection rates across different types of DoS 2040 attacks. Our findings have important 2041 implications for improving organizations’ 2042 cybersecurity posture. 2043 </blockquote> 2044 </div> 2045 </div> 2046 </div> 2047 </div> 2048 <div class="slot-urls"></div> 2049 <a href="/wsc23papers/009.pdf" target="_blank">pdf</a 2050 ><br /> 2051 </div> 2052 <div class="slot-entry"> 2053 <a name="inv167" tabindex="-1"></a> 2054 <div class="slot-title-line"> 2055 <span class="slot-title" 2056 >Using Evolutionary Model Discovery to Develop Robust 2057 Policies</span 2058 > 2059 </div> 2060 <div class="slot-authors"> 2061 Alex Isherwood, Matthew Koehler, and David Slater (MITRE 2062 Corporation) 2063 </div> 2064 <div class="slot-abstract"> 2065 <div> 2066 <a 2067 class="clickable no-decoration" 2068 id="vhsjs_view_20_1707793551_314819" 2069 onclick="$('#vhsjs_view_20_1707793551_314819').hide(); 2070 $('#vhsjs_hide_20_1707793551_314819').show(); 2071 $('#19_1707793551_314811').slideDown(function() { 2072 if (typeof Masonry === 'function') { 2073 $('.use_masonry').masonry(); 2074 }; 2075 2076 });" 2077 ><i class="fa fa-caret-right"></i>
2078 <span class="hover_link">Abstract</span></a 2079 ><a 2080 class="clickable no-decoration" 2081 id="vhsjs_hide_20_1707793551_314819" 2082 onclick="$('#19_1707793551_314811').hide(function() { 2083 if (typeof Masonry === 'function') { 2084 $('.use_masonry').masonry(); 2085 }; 2086 }); 2087 $('#vhsjs_hide_20_1707793551_314819').hide(); 2088 $('#vhsjs_view_20_1707793551_314819').show();" 2089 style="display: none" 2090 ><i class="fa fa-caret-down"></i> 2091 <span class="hover_link">Abstract</span></a 2092 > 2093 <div 2094 data-display-control="20_1707793551_314819" 2095 id="19_1707793551_314811" 2096 style="display: none" 2097 > 2098 <div class="arrow-slidedown"> 2099 <blockquote> 2100 Agent-based models can be a powerful tool for 2101 evaluating the impact of policy decisions on a 2102 population. However, analyses are traditionally 2103 beholden to one set of rules hypothesized at the 2104 conception of the model. Modelers make 2105 assumptions of agent behavior that are not 2106 necessarily governed by data and the actual 2107 behavior of the true population can vary. 2108 Evolutionary Model Discovery provides a solution 2109 to this problem by leveraging genetic algorithms 2110 and genetic programming to explore the plausible 2111 set of rules that can explain agent behavior. 2112 Here we describe an initial use of the EMD 2113 system to develop robust policies in a resource 2114 constrained environment. In this instance, we 2115 extend the NetLogo implementation of the Epstein 2116 Rebellion model model of civil violence as a 2117 sample problem. We use the EMD framework to 2118 generate plausible populations and then develop 2119 policy responses for the government that are 2120 robust across the plausible populations. 2121 </blockquote> 2122 </div> 2123 </div> 2124 </div> 2125 </div> 2126 <div class="slot-urls"></div> 2127 <a href="/wsc23papers/010.pdf" target="_blank">pdf</a 2128 ><br /> 2129 </div> 2130 </div> 2131 <div class="session-entry"> 2132 <span class="session-event-type">Technical Session</span 2133 ><span class="type-track-spacer"> · </span 2134 ><span class="program-track">Agent-based Simulation</span 2135 ><br /> 2136 <div class="session-title"> 2137 Healthcare Agent-based Modeling 2138 </div> 2139 <div class="session-chair"> 2140 Chair: Xueying Liu (Virginia Polytechnic Institute and 2141 State University)<br /> 2142 </div> 2143 <div class="slot-entry"> 2144 <a name="con128" tabindex="-1"></a> 2145 <div class="slot-title-line"> 2146 <span class="slot-title" 2147 >An Iterative Analysis Method Using Causal Discovery 2148 Algorithms to Enhance ABM as a Policy Tool</span 2149 > 2150 </div> 2151 <div class="slot-authors"> 2152 Shuang Chang, Takashi Kato, Yusuke Koyanagi, Kento 2153 Uemura, and Koji Maruhashi (Fujitsu Laboratories Ltd.) 2154 </div> 2155 <div class="slot-abstract"> 2156 <div> 2157 <a 2158 class="clickable no-decoration" 2159 id="vhsjs_view_22_1707793551_3252032" 2160 onclick="$('#vhsjs_view_22_1707793551_3252032').hide(); 2161 $('#vhsjs_hide_22_1707793551_3252032').show(); 2162 $('#21_1707793551_325195').slideDown(function() { 2163 if (typeof Masonry === 'function') { 2164 $('.use_masonry').masonry(); 2165 }; 2166 2167 });" 2168 ><i class="fa fa-caret-right"></i>
2169 <span class="hover_link">Abstract</span></a 2170 ><a 2171 class="clickable no-decoration" 2172 id="vhsjs_hide_22_1707793551_3252032" 2173 onclick="$('#21_1707793551_325195').hide(function() { 2174 if (typeof Masonry === 'function') { 2175 $('.use_masonry').masonry(); 2176 }; 2177 }); 2178 $('#vhsjs_hide_22_1707793551_3252032').hide(); 2179 $('#vhsjs_view_22_1707793551_3252032').show();" 2180 style="display: none" 2181 ><i class="fa fa-caret-down"></i> 2182 <span class="hover_link">Abstract</span></a 2183 > 2184 <div 2185 data-display-control="22_1707793551_3252032" 2186 id="21_1707793551_325195" 2187 style="display: none" 2188 > 2189 <div class="arrow-slidedown"> 2190 <blockquote> 2191 Agent-based modelling (ABM) is becoming a 2192 popular policy tool by modelling the reasoning 2193 processes and interactive behaviors of 2194 individuals against external environments. 2195 However, the presence of heterogeneous agents, 2196 non-linear interactions and complex emergent 2197 patterns raised by even simple behavior rules 2198 pose challenges in the model explanation 2199 process. In this work, we propose a novel 2200 iterative analysis method that leverages causal 2201 discovery algorithms to facilitate policy 2202 formulation and evaluation based on a causal 2203 understanding of the model. It strengthens the 2204 explanation power of ABM by elucidating causal 2205 relations among modelled components. We applied 2206 the method to an agent-based simulator that 2207 models passengers' routing behaviors in a 2208 virtual airport terminal. By discovering the 2209 causal relations among passengers' goals, 2210 actions, and an airport terminal environment 2211 under different COVID-19 regulations, we showed 2212 that the method can inform more effective 2213 indirect-control policies leading to positive 2214 passenger experiences, compared with a 2215 conventional ABM analysis method. 2216 </blockquote> 2217 </div> 2218 </div> 2219 </div> 2220 </div> 2221 <div class="slot-urls"></div> 2222 <a href="/wsc23papers/011.pdf" target="_blank">pdf</a 2223 ><br /> 2224 </div> 2225 <div class="slot-entry"> 2226 <a name="con158" tabindex="-1"></a> 2227 <div class="slot-title-line"> 2228 <span class="slot-title" 2229 >A Review of Agent-based Modeling Applications in 2230 Substance Abuse Policy Research</span 2231 > 2232 </div> 2233 <div class="slot-authors"> 2234 Xiang Zhong (University of Florida), Xuanjing Li 2235 (Tsinghua University), and Samantha Mangoni (University 2236 of Florida) 2237 </div> 2238 <div class="slot-abstract"> 2239 <div> 2240 <a 2241 class="clickable no-decoration" 2242 id="vhsjs_view_24_1707793551_327568" 2243 onclick="$('#vhsjs_view_24_1707793551_327568').hide(); 2244 $('#vhsjs_hide_24_1707793551_327568').show(); 2245 $('#23_1707793551_3275602').slideDown(function() { 2246 if (typeof Masonry === 'function') { 2247 $('.use_masonry').masonry(); 2248 }; 2249 2250 });" 2251 ><i class="fa fa-caret-right"></i>
2252 <span class="hover_link">Abstract</span></a 2253 ><a 2254 class="clickable no-decoration" 2255 id="vhsjs_hide_24_1707793551_327568" 2256 onclick="$('#23_1707793551_3275602').hide(function() { 2257 if (typeof Masonry === 'function') { 2258 $('.use_masonry').masonry(); 2259 }; 2260 }); 2261 $('#vhsjs_hide_24_1707793551_327568').hide(); 2262 $('#vhsjs_view_24_1707793551_327568').show();" 2263 style="display: none" 2264 ><i class="fa fa-caret-down"></i> 2265 <span class="hover_link">Abstract</span></a 2266 > 2267 <div 2268 data-display-control="24_1707793551_327568" 2269 id="23_1707793551_3275602" 2270 style="display: none" 2271 > 2272 <div class="arrow-slidedown"> 2273 <blockquote> 2274 This study provides a systematic review of 2275 existing studies that used agent-based modeling 2276 (ABM) to inform substance abuse policies and 2277 identifies future research directions. The 2278 detailed review included 20 articles, among 2279 which, tobacco, alcohol, cannabis, opioids, and 2280 heroin substance abuse were studied. These 2281 studies examined substance abuse interventions 2282 and the associations between substance use and 2283 social behavior, such as peer interaction and 2284 selection. Effective interventions included 2285 retailer density reduction policies, restriction 2286 of trading hours of licensed venues, ecstasy 2287 pill-testing and passive-alert detection dogs by 2288 police at public venues, and a mass-media drug 2289 prevention education policy. ABM can capture the 2290 dynamic interactions among and between agents 2291 and environments, making it appropriate to model 2292 complex substance abuse behaviors. Limitations 2293 in current studies include a lack of ABM 2294 validation efforts and generalizable data. 2295 Future studies should use generalizable and 2296 abundant information to inform their ABM, as 2297 well as have an explicit validation method. 2298 </blockquote> 2299 </div> 2300 </div> 2301 </div> 2302 </div> 2303 <div class="slot-urls"></div> 2304 <a href="/wsc23papers/012.pdf" target="_blank">pdf</a 2305 ><br /> 2306 </div> 2307 <div class="slot-entry"> 2308 <a name="con230" tabindex="-1"></a> 2309 <div class="slot-title-line"> 2310 <span class="slot-title" 2311 >Supporting Emergency Department Risk Mitigation with 2312 a Modular and Reusable Agent-Based Simulation 2313 Infrastructure</span 2314 > 2315 </div> 2316 <div class="slot-authors"> 2317 Thomas Godfrey (King's College London); Rahul Batra, Sam 2318 Douthwaite, and Jonathan Edgeworth (Guy's and St Thomas' 2319 NHS Foundation Trust); Matthew Edwards (King's College 2320 Hospital NHS Foundation Trust); Simon Miles (Aerogility 2321 Ltd); and Steffen Zschaler (King's College London) 2322 </div> 2323 <div class="slot-abstract"> 2324 <div> 2325 <a 2326 class="clickable no-decoration" 2327 id="vhsjs_view_26_1707793551_330257" 2328 onclick="$('#vhsjs_view_26_1707793551_330257').hide(); 2329 $('#vhsjs_hide_26_1707793551_330257').show(); 2330 $('#25_1707793551_3302484').slideDown(function() { 2331 if (typeof Masonry === 'function') { 2332 $('.use_masonry').masonry(); 2333 }; 2334 2335 });" 2336 ><i class="fa fa-caret-right"></i>
2337 <span class="hover_link">Abstract</span></a 2338 ><a 2339 class="clickable no-decoration" 2340 id="vhsjs_hide_26_1707793551_330257" 2341 onclick="$('#25_1707793551_3302484').hide(function() { 2342 if (typeof Masonry === 'function') { 2343 $('.use_masonry').masonry(); 2344 }; 2345 }); 2346 $('#vhsjs_hide_26_1707793551_330257').hide(); 2347 $('#vhsjs_view_26_1707793551_330257').show();" 2348 style="display: none" 2349 ><i class="fa fa-caret-down"></i> 2350 <span class="hover_link">Abstract</span></a 2351 > 2352 <div 2353 data-display-control="26_1707793551_330257" 2354 id="25_1707793551_3302484" 2355 style="display: none" 2356 > 2357 <div class="arrow-slidedown"> 2358 <blockquote> 2359 For emergency departments (EDs) to maintain 2360 sustainable care of patients, hospital 2361 management must continually explore potential 2362 interventions to clinical practice. Agent-based 2363 modelling (ABM) can be a valuable tool to 2364 support this planning in a controlled 2365 environment. Existing approaches to ABM 2366 development are best suited for one-off models. 2367 However, conditions in EDs can change 2368 frequently, making the use of one-off models 2369 infeasible. Decision-makers must be able to 2370 trust simulations appropriately for them to be 2371 effective in intervention exploration. 2372 Domain-specific modelling languages (DSMLs) can 2373 address these challenges by offering a reusable 2374 library of appropriately-abstract, 2375 domain-familiar, modelling concepts across case 2376 studies and automatic translation of these 2377 concepts into executable models. In this paper 2378 we present a DSML to support repeated modelling 2379 exercises in the ED domain and illustrate the 2380 use and reuse of this DSML across two concrete 2381 case studies in London-based NHS emergency 2382 departments. 2383 </blockquote> 2384 </div> 2385 </div> 2386 </div> 2387 </div> 2388 <div class="slot-urls"></div> 2389 <a href="/wsc23papers/013.pdf" target="_blank">pdf</a 2390 ><br /> 2391 </div> 2392 </div> 2393 <div class="session-entry"> 2394 <span class="session-event-type">Technical Session</span 2395 ><span class="type-track-spacer"> · </span 2396 ><span class="program-track">Agent-based Simulation</span 2397 ><br /> 2398 <div class="session-title"> 2399 Sustainable Transportation Agent-based Modeling 2400 </div> 2401 <div class="session-chair"> 2402 Chair: Xiang Zhong (University of Florida)<br /> 2403 </div> 2404 <div class="slot-entry"> 2405 <a name="con115" tabindex="-1"></a> 2406 <div class="slot-title-line"> 2407 <span class="slot-title" 2408 >Simulating Interaction Behaviors in Bi-directional 2409 Shared Corridor with Real Case Study</span 2410 > 2411 </div> 2412 <div class="slot-authors"> 2413 Yun-Pang Flötteröd, Jakob Erdmann, and Daniel 2414 Krajzewicz (German Aerospace Center (DLR)) and Johan 2415 Olstam (The Swedish National Road and Transport Research 2416 Institute) 2417 </div> 2418 <div class="slot-abstract"> 2419 <div> 2420 <a 2421 class="clickable no-decoration" 2422 id="vhsjs_view_28_1707793551_3365562" 2423 onclick="$('#vhsjs_view_28_1707793551_3365562').hide(); 2424 $('#vhsjs_hide_28_1707793551_3365562').show(); 2425 $('#27_1707793551_3365483').slideDown(function() { 2426 if (typeof Masonry === 'function') { 2427 $('.use_masonry').masonry(); 2428 }; 2429 2430 });" 2431 ><i class="fa fa-caret-right"></i>
2432 <span class="hover_link">Abstract</span></a 2433 ><a 2434 class="clickable no-decoration" 2435 id="vhsjs_hide_28_1707793551_3365562" 2436 onclick="$('#27_1707793551_3365483').hide(function() { 2437 if (typeof Masonry === 'function') { 2438 $('.use_masonry').masonry(); 2439 }; 2440 }); 2441 $('#vhsjs_hide_28_1707793551_3365562').hide(); 2442 $('#vhsjs_view_28_1707793551_3365562').show();" 2443 style="display: none" 2444 ><i class="fa fa-caret-down"></i> 2445 <span class="hover_link">Abstract</span></a 2446 > 2447 <div 2448 data-display-control="28_1707793551_3365562" 2449 id="27_1707793551_3365483" 2450 style="display: none" 2451 > 2452 <div class="arrow-slidedown"> 2453 <blockquote> 2454 Microscopic traffic simulation tools are able to 2455 evaluate possible impacts induced by automated 2456 shuttles under various conditions. However, 2457 automated shuttles operate more and more often 2458 in shared space areas and few microscopic 2459 traffic simulation tools are able to handle 2460 networks with shared space infrastructure. 2461 Interaction behaviors between road users and 2462 automated shuttles are addressed only seldom as 2463 well. In this paper, we propose the concept of 2464 bi-directional edges in the open source 2465 microscopic traffic simulation suite SUMO to 2466 simulate road users’ interactions in a 2467 bi-directional shared-space corridor. A c
2467ase 2468 study, where automated shuttles and cyclists 2469 share the bike path, and the related data 2470 collection were conducted to examine the 2471 performance of the proposed concept and 2472 understand the usage of the shared corridor. The 2473 simulation results are promising. Further 2474 refinement of the proposed concept is planned 2475 for properly reflecting complex interaction 2476 behaviors among diverse road users, and their 2477 surrounding environment. 2478 </blockquote> 2479 </div> 2480 </div> 2481 </div> 2482 </div> 2483 <div class="slot-urls"></div> 2484 <a href="/wsc23papers/014.pdf" target="_blank">pdf</a 2485 ><br /> 2486 </div> 2487 <div class="slot-entry"> 2488 <a name="con188" tabindex="-1"></a> 2489 <div class="slot-title-line"> 2490 <span class="slot-title" 2491 >Rebalancing Integrated, Demand-responsive Passenger 2492 and Freight Transport – An Agent-based 2493 Simulation Approach</span 2494 > 2495 </div> 2496 <div class="slot-authors"> 2497 Johannes Staritz, Julia Kütemeier, Helen Sand, 2498 Christoph von Viebahn, and Maylin Wartenberg (Hochschule 2499 Hannover) 2500 </div> 2501 <div class="slot-abstract"> 2502 <div> 2503 <a 2504 class="clickable no-decoration" 2505 id="vhsjs_view_30_1707793551_339218" 2506 onclick="$('#vhsjs_view_30_1707793551_339218').hide(); 2507 $('#vhsjs_hide_30_1707793551_339218').show(); 2508 $('#29_1707793551_33921').slideDown(function() { 2509 if (typeof Masonry === 'function') { 2510 $('.use_masonry').masonry(); 2511 }; 2512 2513 });" 2514 ><i class="fa fa-caret-right"></i> 2515 <span class="hover_link">Abstract</span></a 2516 ><a 2517 class="clickable no-decoration" 2518 id="vhsjs_hide_30_1707793551_339218" 2519 onclick="$('#29_1707793551_33921').hide(function() { 2520 if (typeof Masonry === 'function') { 2521 $('.use_masonry').masonry(); 2522 }; 2523 }); 2524 $('#vhsjs_hide_30_1707793551_339218').hide(); 2525 $('#vhsjs_view_30_1707793551_339218').show();" 2526 style="display: none" 2527 ><i class="fa fa-caret-down"></i> 2528 <span class="hover_link">Abstract</span></a 2529 > 2530 <div 2531 data-display-control="30_1707793551_339218" 2532 id="29_1707793551_33921" 2533 style="display: none" 2534 > 2535 <div class="arrow-slidedown"> 2536 <blockquote> 2537 Integrated, demand-responsive passenger and 2538 freight transport (IDRT) potentially provides 2539 flexibility and higher service frequency in 2540 areas of low demand due to economies of scale, 2541 while reducing negative traffic-related 2542 externalities such as pollutant emissions, noise 2543 emissions or accidents. However, to allow for 2544 efficient operations in terms of minimum travel 2545 distances, short customer waiting times, and 2546 high vehicle utilization rates, IDRT requires 2547 effective rebalancing strategies that balance 2548 supply and demand capacities by strategically 2549 positioning vehicle resources in the operational 2550 area. Therefore, we propose a rebalancing 2551 strategy for IDRT and measure its effectiveness 2552 through an agent-based simulation model. To 2553 evaluate our approach, we compare the rebalanced
2554 IDRT with a static scenario with backhauls to a 2555 central depot. Our results indicate that the 2556 proposed rebalancing approach can outperform a 2557 system without rebalancing by up to 15.1% in 2558 terms of total fleet kilometers and 30% in terms 2559 of passenger waiting time. 2560 </blockquote> 2561 </div> 2562 </div> 2563 </div> 2564 </div> 2565 <div class="slot-urls"></div> 2566 <a href="/wsc23papers/015.pdf" target="_blank">pdf</a 2567 ><br /> 2568 </div> 2569 <div class="slot-entry"> 2570 <a name="con203" tabindex="-1"></a> 2571 <div class="slot-title-line"> 2572 <span class="slot-title" 2573 >A Simulation Model for Bio-Inspired Charging 2574 Strategies for Electric Vehicles in Industrial 2575 Areas</span 2576 > 2577 </div> 2578 <div class="slot-authors"> 2579 Berry Gerrits and Martijn Mes (University of Twente) and 2580 Robert Andringa (Distribute) 2581 </div> 2582 <div class="slot-abstract"> 2583 <div> 2584 <a 2585 class="clickable no-decoration" 2586 id="vhsjs_view_32_1707793551_3414814" 2587 onclick="$('#vhsjs_view_32_1707793551_3414814').hide(); 2588 $('#vhsjs_hide_32_1707793551_3414814').show(); 2589 $('#31_1707793551_341474').slideDown(function() { 2590 if (typeof Masonry === 'function') { 2591 $('.use_masonry').masonry(); 2592 }; 2593 2594 });" 2595 ><i class="fa fa-caret-right"></i> 2596 <span class="hover_link">Abstract</span></a 2597 ><a 2598 class="clickable no-decoration" 2599 id="vhsjs_hide_32_1707793551_3414814" 2600 onclick="$('#31_1707793551_341474').hide(function() { 2601 if (typeof Masonry === 'function') { 2602 $('.use_masonry').masonry(); 2603 }; 2604 }); 2605 $('#vhsjs_hide_32_1707793551_3414814').hide(); 2606 $('#vhsjs_view_32_1707793551_3414814').show();" 2607 style="display: none" 2608 ><i class="fa fa-caret-down"></i> 2609 <span class="hover_link">Abstract</span></a 2610 > 2611 <div 2612 data-display-control="32_1707793551_3414814" 2613 id="31_1707793551_341474" 2614 style="display: none" 2615 > 2616 <div class="arrow-slidedown"> 2617 <blockquote> 2618 This paper presents an open-source agent-based 2619 simulation model to study bio-inspired charging 2620 policies for local sustainable energy systems in 2621 an industrial setting where electric vehicles 2622 (EVs) perform transportation jobs. Within this 2623 context, we focus on a system that allows to 2624 control the charging-schemes of individual EVs. 2625 To this end, we develop an agent-based 2626 simulation model in NetLogo. We present and 2627 implement a bio-inspired approach based on the 2628 foraging behavior of honeybees and our approach 2629 results in simple, yet effective decision-making 2630 logic. Our approach provides the necessary 2631 parameters to control and balance sustainable 2632 energy systems in terms of EV productivity and 2633 the consumption of locally generated energy. Our 2634 simulation results look promising: the balance 2635 between EV productivity and the use of 2636 sustainable energy can be efficiently tweaked in 2637 a predictable manner using the parameters and 2638 thresholds of the model, yielding 2639 close-to-optimal performance. 2640 </blockquote> 2641 </div> 2642 </div> 2643 </div> 2644 </div> 2645 <div class="slot-urls"></div> 2646 <a href="/wsc23papers/016.pdf" target="_blank">pdf</a 2647 ><br /> 2648 </div> 2649 </div> 2650 <div class="session-entry">
2651 <span class="session-event-type">Technical Session</span 2652 ><span class="type-track-spacer"> · </span 2653 ><span class="program-track">Agent-based Simulation</span 2654 ><br /> 2655 <div class="session-title"> 2656 Games and Agent-based Modeling 2657 </div> 2658 <div class="session-chair"> 2659 Chair: Haibei Zhu (J.P. Morgan)<br /> 2660 </div> 2661 <div class="slot-entry"> 2662 <a name="con351" tabindex="-1"></a> 2663 <div class="slot-title-line"> 2664 <span class="slot-title" 2665 >Modeling Reactive Game Agents Using the Cell-DEVS 2666 Modeling Formalism</span 2667 > 2668 </div> 2669 <div class="slot-authors"> 2670 Alvi Jawad, Cristina Ruiz-Martín, and Gabriel 2671 Wainer (Carleton University) 2672 </div> 2673 <div class="slot-abstract"> 2674 <div> 2675 <a 2676 class="clickable no-decoration" 2677 id="vhsjs_view_34_1707793551_347341" 2678 onclick="$('#vhsjs_view_34_1707793551_347341').hide(); 2679 $('#vhsjs_hide_34_1707793551_347341').show(); 2680 $('#33_1707793551_3473327').slideDown(function() { 2681 if (typeof Masonry === 'function') { 2682 $('.use_masonry').masonry(); 2683 }; 2684 2685 });" 2686 ><i class="fa fa-caret-right"></i> 2687 <span class="hover_link">Abstract</span></a 2688 ><a 2689 class="clickable no-decoration" 2690 id="vhsjs_hide_34_1707793551_347341" 2691 onclick="$('#33_1707793551_3473327').hide(function() { 2692 if (typeof Masonry === 'function') { 2693 $('.use_masonry').masonry(); 2694 }; 2695 }); 2696 $('#vhsjs_hide_34_1707793551_347341').hide(); 2697 $('#vhsjs_view_34_1707793551_347341').show();" 2698 style="display: none" 2699 ><i class="fa fa-caret-down"></i> 2700 <span class="hover_link">Abstract</span></a 2701 > 2702 <div 2703 data-display-control="34_1707793551_347341" 2704 id="33_1707793551_3473327" 2705 style="display: none" 2706 > 2707 <div class="arrow-slidedown"> 2708 <blockquote> 2709 Intelligent game agents are a vital part of 2710 modern games as they add life, story, and 2711 immersion to the game environment. The requests 2712 in the gaming industry for more realism have 2713 made intelligent agents more important than ever 2714 before. Modeling and simulation of game agents 2715 and their surrounding environment provide an 2716 alternate setting to study dynamic agent 2717 behavior before integration into the game 2718 engine. The Cell-DEVS formalism, an extension of 2719 Cellular Automata, allows modeling such 2720 behaviors using the rigorously formalized 2721 Discrete Event Systems Specification (DEVS) 2722 formalism. In this paper, we explain how to 2723 model and test reactive game agents using the 2724 Cell-DEVS formalism and the CD++ toolkit. To 2725 analyze the dynamic behavior of such agents, we 2726 perform several experiments in varying system 2727 configurations. Our experimental results confirm 2728 the versatility of Cell-DEVS and the 2729 functionalities in the CD++ toolkit to model 2730 comfort-driven, exploratory, and desire-driven 2731 game agents. 2732 </blockquote> 2733 </div> 2734 </div> 2735 </div> 2736 </div> 2737 <div class="slot-urls"></div> 2738 <a href="/wsc23papers/017.pdf" target="_blank">pdf</a 2739 ><br /> 2740 </div> 2741 <div class="slot-entry"> 2742 <a name="inv145" tabindex="-1"></a> 2743 <div class="slot-title-line">
2744 <span class="slot-title" 2745 >A Calibration Model for Bot-Like Behaviors in 2746 Agent-Based Anagram Game Simulation</span 2747 > 2748 </div> 2749 <div class="slot-authors"> 2750 Xueying Liu, Zhihao Hu, and Xinwei Deng (Virginia Tech) 2751 and Chris Kuhlman (University of Virginia) 2752 </div> 2753 <div class="slot-abstract"> 2754 <div> 2755 <a 2756 class="clickable no-decoration" 2757 id="vhsjs_view_36_1707793551_3496974" 2758 onclick="$('#vhsjs_view_36_1707793551_3496974').hide(); 2759 $('#vhsjs_hide_36_1707793551_3496974').show(); 2760 $('#35_1707793551_3496895').slideDown(function() { 2761 if (typeof Masonry === 'function') { 2762 $('.use_masonry').masonry(); 2763 }; 2764 2765 });" 2766 ><i class="fa fa-caret-right"></i> 2767 <span class="hover_link">Abstract</span></a 2768 ><a 2769 class="clickable no-decoration" 2770 id="vhsjs_hide_36_1707793551_3496974" 2771 onclick="$('#35_1707793551_3496895').hide(function() { 2772 if (typeof Masonry === 'function') { 2773 $('.use_masonry').masonry(); 2774 }; 2775 }); 2776 $('#vhsjs_hide_36_1707793551_3496974').hide(); 2777 $('#vhsjs_view_36_1707793551_3496974').show();" 2778 style="display: none" 2779 ><i class="fa fa-caret-down"></i> 2780 <span class="hover_link">Abstract</span></a 2781 > 2782 <div 2783 data-display-control="36_1707793551_3496974" 2784 id="35_1707793551_3496895" 2785 style="display: none" 2786 > 2787 <div class="arrow-slidedown"> 2788 <blockquote> 2789 Experiments that are games played among a 2790 network of players are widely used to study 2791 human behavior. Furthermore, bots or intelligent 2792 systems can be used in these games to produce 2793 contexts that elicit particular types of human 2794 responses. Bot behaviors could be specified 2795 solely based on experimental data. In this work, 2796 we take a different perspective, called the 2797 Probability Calibration (PC) approach, to 2798 simulate networked group anagram games with 2799 certain players having bot-like behaviors. The 2800 proposed method starts with data-driven models 2801 and calibrates in principled ways the parameters 2802 that alter player behaviors. It can alter the 2803 performance of each type of agent (e.g., bot) in 2804 group anagram games. Further, statistical 2805 methods are used to test whether the PC models 2806 produce results that are statistically different 2807 from those of the original models. Case studies 2808 demonstrate the merits of the proposed method. 2809 </blockquote> 2810 </div> 2811 </div> 2812 </div> 2813 </div> 2814 <div class="slot-urls"></div> 2815 <a href="/wsc23papers/018.pdf" target="_blank">pdf</a 2816 ><br /> 2817 </div> 2818 <div class="slot-entry"> 2819 <a name="inv163" tabindex="-1"></a> 2820 <div class="slot-title-line">
2821 <span class="slot-title" 2822 >Feature Importance for Uncertainty Quantification in 2823 Agent-based Modeling</span 2824 > 2825 </div> 2826 <div class="slot-authors"> 2827 Gayane Grigoryan and Andrew J. Collins (Old Dominion 2828 University) 2829 </div> 2830 <div class="slot-abstract"> 2831 <div> 2832 <a 2833 class="clickable no-decoration" 2834 id="vhsjs_view_38_1707793551_351896" 2835 onclick="$('#vhsjs_view_38_1707793551_351896').hide(); 2836 $('#vhsjs_hide_38_1707793551_351896').show(); 2837 $('#37_1707793551_3518882').slideDown(function() { 2838 if (typeof Masonry === 'function') { 2839 $('.use_masonry').masonry(); 2840 }; 2841 2842 });" 2843 ><i class="fa fa-caret-right"></i> 2844 <span class="hover_link">Abstract</span></a 2845 ><a 2846 class="clickable no-decoration" 2847 id="vhsjs_hide_38_1707793551_351896" 2848 onclick="$('#37_1707793551_3518882').hide(function() { 2849 if (typeof Masonry === 'function') { 2850 $('.use_masonry').masonry(); 2851 }; 2852 }); 2853 $('#vhsjs_hide_38_1707793551_351896').hide(); 2854 $('#vhsjs_view_38_1707793551_351896').show();" 2855 style="display: none" 2856 ><i class="fa fa-caret-down"></i> 2857 <span class="hover_link">Abstract</span></a 2858 > 2859 <div 2860 data-display-control="38_1707793551_351896" 2861 id="37_1707793551_3518882" 2862 style="display: none" 2863 > 2864 <div class="arrow-slidedown"> 2865 <blockquote> 2866 Simulation models are subject to uncertainty and 2867 sensitivity, meaning that even small variations 2868 of input can cause considerable fluctuations in 2869 the output results. Consequently, this can 2870 amplify the uncertainty associated with the 2871 simulation, thereby limiting the confidence one 2872 can have in its outcomes. To mitigate these 2873 effects, this paper suggests using a cooperative 2874 game theory-based feature importance method, 2875 which can identify uncertainty in a dataset, and 2876 provide additional insights that could be used 2877 in the development or analysis of a simulation 2878 model. A predator-prey scenario was considered, 2879 demonstrating its usefulness in identifying 2880 important parameters or features. By identifying 2881 the most influential parameters or features, 2882 this approach can help improve the accuracy, 2883 explainability, and reliability of simulation 2884 models as well as other models with highly 2885 variable input parameters. 2886 </blockquote> 2887 </div> 2888 </div> 2889 </div> 2890 </div> 2891 <div class="slot-urls"></div> 2892 <a href="/wsc23papers/019.pdf" target="_blank">pdf</a 2893 ><br /> 2894 </div> 2895 </div> 2896 <div class="session-entry">
2897 <span class="session-event-type">Technical Session</span 2898 ><span class="type-track-spacer"> · </span 2899 ><span class="program-track">Agent-based Simulation</span 2900 ><br /> 2901 <div class="session-title"> 2902 Transportation Agent-based Modeling 2903 </div> 2904 <div class="session-chair"> 2905 Chair: Kshama Dwarakanath (J.P. Morgan AI Research)<br /> 2906 </div> 2907 <div class="slot-entry"> 2908 <a name="con222" tabindex="-1"></a> 2909 <div class="slot-title-line"> 2910 <span class="slot-title" 2911 >A Simulation-Based Method for Analyzing Supply Chain 2912 Vulnerability Under Pandemic: A Special Focus on the 2913 Covid-19</span 2914 > 2915 </div> 2916 <div class="slot-authors"> 2917 Xinglu Xu and Bochi Liu (Dalian University of 2918 Technology) and Weihong Grace Guo (Rutgers, The State 2919 University of New Jersey) 2920 </div> 2921 <div class="slot-abstract"> 2922 <div> 2923 <a 2924 class="clickable no-decoration" 2925 id="vhsjs_view_40_1707793551_3844788" 2926 onclick="$('#vhsjs_view_40_1707793551_3844788').hide(); 2927 $('#vhsjs_hide_40_1707793551_3844788').show(); 2928 $('#39_1707793551_384469').slideDown(function() { 2929 if (typeof Masonry === 'function') { 2930 $('.use_masonry').masonry(); 2931 }; 2932 2933 });" 2934 ><i class="fa fa-caret-right"></i> 2935 <span class="hover_link">Abstract</span></a 2936 ><a 2937 class="clickable no-decoration" 2938 id="vhsjs_hide_40_1707793551_3844788" 2939 onclick="$('#39_1707793551_384469').hide(function() { 2940 if (typeof Masonry === 'function') { 2941 $('.use_masonry').masonry(); 2942 }; 2943 }); 2944 $('#vhsjs_hide_40_1707793551_3844788').hide(); 2945 $('#vhsjs_view_40_1707793551_3844788').show();" 2946 style="display: none" 2947 ><i class="fa fa-caret-down"></i> 2948 <span class="hover_link">Abstract</span></a 2949 > 2950 <div 2951 data-display-control="40_1707793551_3844788" 2952 id="39_1707793551_384469" 2953 style="display: none" 2954 > 2955 <div class="arrow-slidedown"> 2956 <blockquote> 2957 This paper develops a simulation-based 2958 quantitative method to investigate the joint 2959 impact of multiple risks on the supply chain 2960 system during the pandemic. A hybrid simulation 2961 method that combines the 2962 susceptible-infected-recovered (SIR) model and 2963 the agent-based simulation method is proposed to 2964 simulate the risk propagation along the supply 2965 chain and the interactions between distribution 2966 centers and retailers. By analyzing the results 2967 of scenarios with different interventions under 2968 COVID-19, results show that the impact of 2969 interventions is diminishing along the supply 2970 chain. For intervention deployment, adding 2971 testing capacity is of great importance. For 2972 stakeholder management strategies, diversifying 2973 the upstream partners is helpful. Against the 2974 backdrop of a multi-wave global pandemic, this 2975 paper takes the COVID-19 pandemic as an example 2976 to provide a paradigm for modeling the risk 2977 propagation in supply chain systems. Also, the 2978 study demonstrates how to estimate possible 2979 time-varying risk scenarios in face of the data 2980 shortage challenge. 2981 </blockquote> 2982 </div> 2983 </div> 2984 </div> 2985 </div> 2986 <div class="slot-urls"></div> 2987 <a href="/wsc23papers/020.pdf" target="_blank">pdf</a 2988 ><br /> 2989 </div> 2990 <div class="slot-entry"> 2991 <a name="con236" tabindex="-1"></a> 2992 <div class="slot-title-line">
2993 <span class="slot-title" 2994 >System Simulation and Machine Learning-Based 2995 Maintenance Optimization for an Inland Waterway 2996 Transportation System</span 2997 > 2998 </div> 2999 <div class="slot-authors"> 3000 Maryam Aghamohammadghasem, Jose Azucena, Farid 3001 Hashemian, Haitao Liao, Shengfan Zhang, and Heather 3002 Nachtmann (University of Arkansas) 3003 </div> 3004 <div class="slot-abstract"> 3005 <div> 3006 <a 3007 class="clickable no-decoration" 3008 id="vhsjs_view_42_1707793551_3870106" 3009 onclick="$('#vhsjs_view_42_1707793551_3870106').hide(); 3010 $('#vhsjs_hide_42_1707793551_3870106').show(); 3011 $('#41_1707793551_3870022').slideDown(function() { 3012 if (typeof Masonry === 'function') { 3013 $('.use_masonry').masonry(); 3014 }; 3015 3016 });" 3017 ><i class="fa fa-caret-right"></i> 3018 <span class="hover_link">Abstract</span></a 3019 ><a 3020 class="clickable no-decoration" 3021 id="vhsjs_hide_42_1707793551_3870106" 3022 onclick="$('#41_1707793551_3870022').hide(function() { 3023 if (typeof Masonry === 'function') { 3024 $('.use_masonry').masonry(); 3025 }; 3026 }); 3027 $('#vhsjs_hide_42_1707793551_3870106').hide(); 3028 $('#vhsjs_view_42_1707793551_3870106').show();" 3029 style="display: none" 3030 ><i class="fa fa-caret-down"></i> 3031 <span class="hover_link">Abstract</span></a 3032 > 3033 <div 3034 data-display-control="42_1707793551_3870106" 3035 id="41_1707793551_3870022" 3036 style="display: none" 3037 > 3038 <div class="arrow-slidedown"> 3039 <blockquote> 3040 To continue operations of the inland waterway 3041 transportation system (IWTS), the interconnected 3042 infrastructure, such as locks and dam systems, 3043 must remain in good operating condition. 3044 However, as the IWTS ages, unexpected 3045 disruptions increase, causing significant 3046 transportation delays and economic losses. To 3047 evaluate the impacts of IWTS disruptions, a 3048 Python-enhanced NetLogo simulation tool is 3049 developed, where the extreme natural events are 3050 considered and represented by a spatiotemporal 3051 model. Utilizing this tool, optimal maintenance 3052 strategies that maximize cargo throughput on the 3053 IWTS are determined via deep reinforcement 3054 learning. A case study of the lower Mississippi 3055 River system and the McClellan-Kerr Arkansas 3056 River Navigation System is conducted to 3057 illustrate the capability of the developed 3058 simulation and machine learning-based method for 3059 IWTS maintenance optimization. 3060 </blockquote> 3061 </div> 3062 </div> 3063 </div> 3064 </div> 3065 <div class="slot-urls"></div> 3066 <a href="/wsc23papers/022.pdf" target="_blank">pdf</a 3067 ><br /> 3068 </div> 3069 <div class="slot-entry"> 3070 <a name="con232" tabindex="-1"></a> 3071 <div class="slot-title-line">
3072 <span class="slot-title" 3073 >Four Years of Not-Using a Simulator: The Agent-Based 3074 Template</span 3075 > 3076 </div> 3077 <div class="slot-authors"> 3078 Dominik Brunmeir and Martin Bicher (TU Wien); Matthias 3079 Rößler, Christoph Urach, Claire Rippinger, and 3080 Matthias Wastian (dwh GmbH); and Niki Popper (TU Wien) 3081 </div> 3082 <div class="slot-abstract"> 3083 <div> 3084 <a 3085 class="clickable no-decoration" 3086 id="vhsjs_view_44_1707793551_3898673" 3087 onclick="$('#vhsjs_view_44_1707793551_3898673').hide(); 3088 $('#vhsjs_hide_44_1707793551_3898673').show(); 3089 $('#43_1707793551_389859').slideDown(function() { 3090 if (typeof Masonry === 'function') { 3091 $('.use_masonry').masonry(); 3092 }; 3093 3094 });" 3095 ><i class="fa fa-caret-right"></i> 3096 <span class="hover_link">Abstract</span></a 3097 ><a 3098 class="clickable no-decoration" 3099 id="vhsjs_hide_44_1707793551_3898673" 3100 onclick="$('#43_1707793551_389859').hide(function() { 3101 if (typeof Masonry === 'function') { 3102 $('.use_masonry').masonry(); 3103 }; 3104 }); 3105 $('#vhsjs_hide_44_1707793551_3898673').hide(); 3106 $('#vhsjs_view_44_1707793551_3898673').show();" 3107 style="display: none" 3108 ><i class="fa fa-caret-down"></i> 3109 <span class="hover_link">Abstract</span></a 3110 > 3111 <div 3112 data-display-control="44_1707793551_3898673" 3113 id="43_1707793551_389859" 3114 style="display: none" 3115 > 3116 <div class="arrow-slidedown"> 3117 <blockquote> 3118 With steadily increasing performance of 3119 computers, agent-based modeling has evolved from 3120 an analysis method for qualitative phenomena to 3121 strategy for quantitative decision support. With 3122 this orientation, however, the modeler faces new 3123 challenges during implementation. In particular, 3124 an appropriate simulation tool must feature the 3125 combination of data and model flexibility, 3126 process reproducibility, performance and 3127 portability. While existing simulators often do 3128 not sufficiently cover these features, it is 3129 also not sustainable to generally implement 3130 models from scratch. In this work, we want to 3131 present the idea of simulation templates as a 3132 compromise between the two strategies. We show, 3133 on the example of our Agent-Based Template and 3134 two use cases, the importance of the described 3135 challenges and how the simulation template 3136 concept supports solving them. We aim to 3137 generally promote the idea of developing a 3138 customized template, which, as a layer between 3139 simulator and from-the-scratch implementation, 3140 combines the advantages of both approaches. 3141 </blockquote> 3142 </div> 3143 </div> 3144 </div> 3145 </div> 3146 <div class="slot-urls"></div> 3147 <a href="/wsc23papers/021.pdf" target="_blank">pdf</a 3148 ><br /> 3149 </div> 3150 </div> 3151 <div class="session-entry">
3152 <span class="session-event-type">Technical Session</span 3153 ><span class="type-track-spacer"> · </span 3154 ><span class="program-track">Agent-based Simulation</span 3155 ><br /> 3156 <div class="session-title">Agent-based Modeling Design</div> 3157 <div class="session-chair"> 3158 Chair: Gayane Grigoryan (Old Dominion University)<br /> 3159 </div> 3160 <div class="slot-entry"> 3161 <a name="con201" tabindex="-1"></a> 3162 <div class="slot-title-line"> 3163 <span class="slot-title" 3164 >Transparency as Delayed Observability in Multi-Agent 3165 Systems</span 3166 > 3167 </div> 3168 <div class="slot-authors"> 3169 Kshama Dwarakanath and Svitlana Vyetrenko (J.P. Morgan 3170 AI Research), Toks Oyebode (J.P. Morgan Regulatory 3171 Affairs), and Tucker Balch (J.P. Morgan AI Research) 3172 </div> 3173 <div class="slot-abstract"> 3174 <div> 3175 <a 3176 class="clickable no-decoration" 3177 id="vhsjs_view_46_1707793551_394942" 3178 onclick="$('#vhsjs_view_46_1707793551_394942').hide(); 3179 $('#vhsjs_hide_46_1707793551_394942').show(); 3180 $('#45_1707793551_3949335').slideDown(function() { 3181 if (typeof Masonry === 'function') { 3182 $('.use_masonry').masonry(); 3183 }; 3184 3185 });" 3186 ><i class="fa fa-caret-right"></i> 3187 <span class="hover_link">Abstract</span></a 3188 ><a 3189 class="clickable no-decoration" 3190 id="vhsjs_hide_46_1707793551_394942" 3191 onclick="$('#45_1707793551_3949335').hide(function() { 3192 if (typeof Masonry === 'function') { 3193 $('.use_masonry').masonry(); 3194 }; 3195 }); 3196 $('#vhsjs_hide_46_1707793551_394942').hide(); 3197 $('#vhsjs_view_46_1707793551_394942').show();" 3198 style="display: none" 3199 ><i class="fa fa-caret-down"></i> 3200 <span class="hover_link">Abstract</span></a 3201 > 3202 <div 3203 data-display-control="46_1707793551_394942" 3204 id="45_1707793551_3949335" 3205 style="display: none" 3206 > 3207 <div class="arrow-slidedown"> 3208 <blockquote> 3209 Is transparency always beneficial in complex 3210 systems such as traffic networks and stock 3211 markets? How is transparency defined in 3212 multi-agent systems, and what is its optimal 3213 degree at which social welfare is highest? We 3214 take an agent-based view to define transparency 3215 (or its lacking) as delay in agent observability 3216 of environment states, and utilize simulations 3217 to analyze the impact of delay on social 3218 welfare. To model the adaptation of agent 3219 strategies with varying delays, we model agents 3220 as learners maximizing the same objectives under 3221 different delays in a simulated environment. 3222 Focusing on two agent types - constrained and 3223 unconstrained, we use multi-agent reinfor
3223cement 3224 learning to evaluate the impact of delay on 3225 agent outcomes and social welfare. Empirical 3226 demonstration of our framework in simulated 3227 financial markets shows opposing trends in 3228 outcomes of the constrained and unconstrained 3229 agents with delay, with an optimal partial 3230 transparency regime at which social welfare is 3231 maximal. 3232 </blockquote> 3233 </div> 3234 </div> 3235 </div> 3236 </div> 3237 <div class="slot-urls"></div> 3238 <a href="/wsc23papers/023.pdf" target="_blank">pdf</a 3239 ><br /> 3240 </div> 3241 <div class="slot-entry"> 3242 <a name="con275" tabindex="-1"></a> 3243 <div class="slot-title-line"> 3244 <span class="slot-title" 3245 >Once Burned, Twice Shy? The Effect of Stock Market 3246 Bubbles on Traders that Learn by Experience</span 3247 > 3248 </div> 3249 <div class="slot-authors"> 3250 Haibei Zhu and Svitlana Vyetrenko (J.P. Morgan), Serafin 3251 Grundl (Federal Reserve Board), David Byrd (Bowdoin 3252 College), and Kshama Dwarakanath and Tucker Balch (J.P. 3253 Morgan) 3254 </div> 3255 <div class="slot-abstract"> 3256 <div> 3257 <a 3258 class="clickable no-decoration" 3259 id="vhsjs_view_48_1707793551_397632" 3260 onclick="$('#vhsjs_view_48_1707793551_397632').hide(); 3261 $('#vhsjs_hide_48_1707793551_397632').show(); 3262 $('#47_1707793551_397624').slideDown(function() { 3263 if (typeof Masonry === 'function') { 3264 $('.use_masonry').masonry(); 3265 }; 3266 3267 });" 3268 ><i class="fa fa-caret-right"></i> 3269 <span class="hover_link">Abstract</span></a 3270 ><a 3271 class="clickable no-decoration" 3272 id="vhsjs_hide_48_1707793551_397632" 3273 onclick="$('#47_1707793551_397624').hide(function() { 3274 if (typeof Masonry === 'function') { 3275 $('.use_masonry').masonry(); 3276 }; 3277 }); 3278 $('#vhsjs_hide_48_1707793551_397632').hide(); 3279 $('#vhsjs_view_48_1707793551_397632').show();" 3280 style="display: none" 3281 ><i class="fa fa-caret-down"></i> 3282 <span class="hover_link">Abstract</span></a 3283 > 3284 <div 3285 data-display-control="48_1707793551_397632" 3286 id="47_1707793551_397624" 3287 style="display: none" 3288 > 3289 <div class="arrow-slidedown"> 3290 <blockquote> 3291 We study how experience with asset price bubbles 3292 changes the trading strategies of reinforcement 3293 learning (RL) traders and ask whether the change 3294 in trading strategies helps to prevent future 3295 bubbles. We train the RL traders in a 3296 multi-agent market simulation platform, ABIDES, 3297 and compare the strategies of traders trained 3298 with and without bubble experience. We find that 3299 RL traders without bubble experience behave like 3300 short-term momentum traders, whereas traders 3301 with bubble experience behave like value 3302 traders. Therefore, RL traders without bubble 3303 experience amplify bubbles, whereas RL traders 3304 with bubble experience tend to suppress and 3305 sometimes prevent them. This finding suggests 3306 that learning from experience is a mechanism for 3307 a boom and bust cycle where the experience of a 3308 collapsing bubble makes future bubbles less 3309 likely for a period of time until the memory 3310 fades and bubbles become more likely to form 3311 again. 3312 </blockquote> 3313 </div> 3314 </div> 3315 </div> 3316 </div> 3317 <div class="slot-urls"></div> 3318 <a href="/wsc23papers/024.pdf" target="_blank">pdf</a 3319 ><br /> 3320 </div> 3321 <div class="slot-entry"> 3322 <a name="con304" tabindex="-1"></a> 3323 <div class="slot-title-line">
3324 <span class="slot-title" 3325 >Matchmaking in Crowd-shipping Platforms: The Effects 3326 of Mediator Control</span 3327 > 3328 </div> 3329 <div class="slot-authors"> 3330 Preetam Kulkarni and Caroline C. Krejci (University of 3331 Texas at Arlington) 3332 </div> 3333 <div class="slot-abstract"> 3334 <div> 3335 <a 3336 class="clickable no-decoration" 3337 id="vhsjs_view_50_1707793551_399937" 3338 onclick="$('#vhsjs_view_50_1707793551_399937').hide(); 3339 $('#vhsjs_hide_50_1707793551_399937').show(); 3340 $('#49_1707793551_399929').slideDown(function() { 3341 if (typeof Masonry === 'function') { 3342 $('.use_masonry').masonry(); 3343 }; 3344 3345 });" 3346 ><i class="fa fa-caret-right"></i> 3347 <span class="hover_link">Abstract</span></a 3348 ><a 3349 class="clickable no-decoration" 3350 id="vhsjs_hide_50_1707793551_399937" 3351 onclick="$('#49_1707793551_399929').hide(function() { 3352 if (typeof Masonry === 'function') { 3353 $('.use_masonry').masonry(); 3354 }; 3355 }); 3356 $('#vhsjs_hide_50_1707793551_399937').hide(); 3357 $('#vhsjs_view_50_1707793551_399937').show();" 3358 style="display: none" 3359 ><i class="fa fa-caret-down"></i> 3360 <span class="hover_link">Abstract</span></a 3361 > 3362 <div 3363 data-display-control="50_1707793551_399937" 3364 id="49_1707793551_399929" 3365 style="display: none" 3366 > 3367 <div class="arrow-slidedown"> 3368 <blockquote> 3369 A critical design decision for crowdsourcing 3370 platforms is the degree to which the platform 3371 mediator controls participant interactions. 3372 Platforms having a centralized model of 3373 mediation optimize for convenience, speed, and 3374 security in participant interactions, while 3375 platforms operating under decentralized control 3376 require greater user effort but offer them 3377 greater control and agency. The research 3378 described in this paper is a preliminary study 3379 using agent-based modeling to evaluate and 3380 compare the performance of crowd-shipping 3381 platforms with centralized/decentralized control 3382 over matchmaking of carriers and senders. 3383 Results indicate that centralized matchmaking 3384 protects the platform from premature failure 3385 when initial carrier/sender participation is 3386 low. Furthermore, when the platform’s 3387 assignment algorithm is designed to maximize 3388 platform revenue, subject to meeting 3389 carriers’ profit expectations, centralized 3390 matchmaking will tend to outperform 3391 decentralized matchmaking for both the mediator 3392 and the carriers. 3393 </blockquote> 3394 </div> 3395 </div> 3396 </div> 3397 </div> 3398 <div class="slot-urls"></div> 3399 <a href="/wsc23papers/025.pdf" target="_blank">pdf</a 3400 ><br /> 3401 </div> 3402 </div> 3403 </div> 3404 <div class="centered"> 3405 <div class="top-link"><a href="#top">Return to Top</a></div> 3406 </div> 3407 <hr /> 3408 </div> 3409 <div class="area-section"> 3410 <div class="centered"> 3411 <a name="ptrack103" tabindex="-1"></a> 3412 <div class="section-title">Analysis Methodology</div> 3413 </div> 3414 <div class="centered track-chair">
3415 <span class="track-chair-role" 3416 >Track Coordinator - Analysis Methodology: </span 3417 ><span class="track-chair-names" 3418 >Ben Feng (University of Waterloo), Sara Shashaani (North 3419 Carolina State University)</span 3420 > 3421 </div> 3422 <div class="section-entry"> 3423 <div class="session-entry"> 3424 <span class="session-event-type">Technical Session</span 3425 ><span class="type-track-spacer"> · </span 3426 ><span class="program-track">Analysis Methodology</span 3427 ><br /> 3428 <div class="session-title"> 3429 Simulation in Queueing Systems 3430 </div> 3431 <div class="session-chair"> 3432 Chair: Jun Luo (Shanghai Jiao Tong University)<br /> 3433 </div> 3434 <div class="slot-entry"> 3435 <a name="con267" tabindex="-1"></a> 3436 <div class="slot-title-line"> 3437 <span class="slot-title" 3438 >Real-Time Estimations for the Waiting-Time 3439 Distribution in Time-Varying Queues</span 3440 > 3441 </div> 3442 <div class="slot-authors"> 3443 Kurtis Konrad and Yunan Liu (North Carolina State 3444 University) 3445 </div> 3446 <div class="slot-abstract"> 3447 <div> 3448 <a 3449 class="clickable no-decoration" 3450 id="vhsjs_view_52_1707793551_4081285" 3451 onclick="$('#vhsjs_view_52_1707793551_4081285').hide(); 3452 $('#vhsjs_hide_52_1707793551_4081285').show(); 3453 $('#51_1707793551_4081204').slideDown(function() { 3454 if (typeof Masonry === 'function') { 3455 $('.use_masonry').masonry(); 3456 }; 3457 3458 });" 3459 ><i class="fa fa-caret-right"></i> 3460 <span class="hover_link">Abstract</span></a 3461 ><a 3462 class="clickable no-decoration" 3463 id="vhsjs_hide_52_1707793551_4081285" 3464 onclick="$('#51_1707793551_4081204').hide(function() { 3465 if (typeof Masonry === 'function') { 3466 $('.use_masonry').masonry(); 3467 }; 3468 }); 3469 $('#vhsjs_hide_52_1707793551_4081285').hide(); 3470 $('#vhsjs_view_52_1707793551_4081285').show();" 3471 style="display: none" 3472 ><i class="fa fa-caret-down"></i> 3473 <span class="hover_link">Abstract</span></a 3474 > 3475 <div 3476 data-display-control="52_1707793551_4081285" 3477 id="51_1707793551_4081204" 3478 style="display: none" 3479 > 3480 <div class="arrow-slidedown"> 3481 <blockquote> 3482 Customers’ waiting times are the most 3483 commonly used performance data to measure the 3484 quality of service in service systems such as 3485 call centers and healthcare. Unlike stationary 3486 queueing models where customers’ waiting 3487 times are statistically similar, the prediction 3488 of waiting times is far less straightforward in 3489 time-varying queues having nonstationary demand 3490 (i.e., arrival rate) and supply (i.e., number of 3491 servers). In this paper, we develop a novel 3492 methodology for more accurately computing the 3493 wait time distribution in a time-varying 3494 queueing system. We design extensive simulation 3495 experiments to evaluate our prediction methods. 3496 In addition, we discover that the waiting-time 3497 prediction is highly sensitive to the 3498 work-releasing policy of the staffing plan, 3499 i.e., the rule under which the number of servers 3500 changes in time. 3501 </blockquote> 3502 </div> 3503 </div> 3504 </div> 3505 </div> 3506 <div class="slot-urls"></div> 3507 <a href="/wsc23papers/026.pdf" target="_blank">pdf</a 3508 ><br /> 3509 </div> 3510 <div class="slot-entry"> 3511 <a name="con268" tabindex="-1"></a> 3512 <div class="slot-title-line">
3513 <span class="slot-title" 3514 >Achieving Stable Service-Level Targets in 3515 Time-Varying Queueing Systems: A Simulation-Based 3516 Offline Learning Staffing Algorithm</span 3517 > 3518 </div> 3519 <div class="slot-authors"> 3520 Kurtis Konrad and Yunan Liu (North Carolina State 3521 University) 3522 </div> 3523 <div class="slot-abstract"> 3524 <div> 3525 <a 3526 class="clickable no-decoration" 3527 id="vhsjs_view_54_1707793551_4103327" 3528 onclick="$('#vhsjs_view_54_1707793551_4103327').hide(); 3529 $('#vhsjs_hide_54_1707793551_4103327').show(); 3530 $('#53_1707793551_4103243').slideDown(function() { 3531 if (typeof Masonry === 'function') { 3532 $('.use_masonry').masonry(); 3533 }; 3534 3535 });" 3536 ><i class="fa fa-caret-right"></i> 3537 <span class="hover_link">Abstract</span></a 3538 ><a 3539 class="clickable no-decoration" 3540 id="vhsjs_hide_54_1707793551_4103327" 3541 onclick="$('#53_1707793551_4103243').hide(function() { 3542 if (typeof Masonry === 'function') { 3543 $('.use_masonry').masonry(); 3544 }; 3545 }); 3546 $('#vhsjs_hide_54_1707793551_4103327').hide(); 3547 $('#vhsjs_view_54_1707793551_4103327').show();" 3548 style="display: none" 3549 ><i class="fa fa-caret-down"></i> 3550 <span class="hover_link">Abstract</span></a 3551 > 3552 <div 3553 data-display-control="54_1707793551_4103327" 3554 id="53_1707793551_4103243" 3555 style="display: none" 3556 > 3557 <div class="arrow-slidedown"> 3558 <blockquote> 3559 In this paper, we develop a new staffing 3560 algorithm for achieving stable service-level 3561 targets in queues with time-varying arrivals. 3562 Specifically, we aim to stabilize the tail 3563 probability of delay, which is the probability 3564 that the waiting time exceeds a designated 3565 target τ > 0. We integrate reinforcement 3566 learning into the decision making in queueing 3567 models; our new method recursively evolve the 3568 staffing decision by alternating between two 3569 phases: (i) we generate simulated queueing data 3570 by operating the system under the present 3571 staffing function (exploration), and (ii) we 3572 utilize the newly generated data to devise 3573 improved staffing decision (exploitation). We 3574 demonstrate the effectiveness of our new method 3575 using various numerical examples. 3576 </blockquote> 3577 </div> 3578 </div> 3579 </div> 3580 </div> 3581 <div class="slot-urls"></div> 3582 <a href="/wsc23papers/027.pdf" target="_blank">pdf</a 3583 ><br /> 3584 </div> 3585 <div class="slot-entry"> 3586 <a name="con182" tabindex="-1"></a> 3587 <div class="slot-title-line">
3588 <span class="slot-title" 3589 >Estimating Spline-based Nonhomogeneous Poisson 3590 Intensities Using Constrained Quadratic 3591 Programming</span 3592 > 3593 </div> 3594 <div class="slot-authors"> 3595 Siqi Chen, Jing Yang (Sunny) Xi, and Wai Kin (Victor) 3596 Chan (Tsinghua-Berkeley Shenzhen Institute, Shenzhen 3597 International Graduate School, Tsinghua University) 3598 </div> 3599 <div class="slot-abstract"> 3600 <div> 3601 <a 3602 class="clickable no-decoration" 3603 id="vhsjs_view_56_1707793551_412643" 3604 onclick="$('#vhsjs_view_56_1707793551_412643').hide(); 3605 $('#vhsjs_hide_56_1707793551_412643').show(); 3606 $('#55_1707793551_4126348').slideDown(function() { 3607 if (typeof Masonry === 'function') { 3608 $('.use_masonry').masonry(); 3609 }; 3610 3611 });" 3612 ><i class="fa fa-caret-right"></i> 3613 <span class="hover_link">Abstract</span></a 3614 ><a 3615 class="clickable no-decoration" 3616 id="vhsjs_hide_56_1707793551_412643" 3617 onclick="$('#55_1707793551_4126348').hide(function() { 3618 if (typeof Masonry === 'function') { 3619 $('.use_masonry').masonry(); 3620 }; 3621 }); 3622 $('#vhsjs_hide_56_1707793551_412643').hide(); 3623 $('#vhsjs_view_56_1707793551_412643').show();" 3624 style="display: none" 3625 ><i class="fa fa-caret-down"></i> 3626 <span class="hover_link">Abstract</span></a 3627 > 3628 <div 3629 data-display-control="56_1707793551_412643" 3630 id="55_1707793551_4126348" 3631 style="display: none" 3632 > 3633 <div class="arrow-slidedown"> 3634 <blockquote> 3635 This paper estimates the intensity function of a 3636 nonhomogeneous Poisson process (NHPP) using a 3637 spline-based method with constrained quadratic 3638 programming (CQP). Based on the property of 3639 B-splines, we transform the estimation problem 3640 into an optimization problem and apply CQP to 3641 obtain the estimated intensity function with low 3642 computational expense. Numerical experiments are 3643 conducted to verify the performance of our 3644 method. In addition, the impacts of the number 3645 of intervals from event-count data and the 3646 number of knots in B-splines are also discussed 3647 to explore the properties of spline-based 3648 models. 3649 </blockquote> 3650 </div> 3651 </div> 3652 </div> 3653 </div> 3654 <div class="slot-urls"></div> 3655 <a href="/wsc23papers/028.pdf" target="_blank">pdf</a 3656 ><br /> 3657 </div> 3658 </div> 3659 <div class="session-entry"> 3660 <span class="session-event-type">Technical Session</span 3661 ><span class="type-track-spacer"> · </span 3662 ><span class="program-track">Analysis Methodology</span 3663 ><br /> 3664 <div class="session-title"> 3665 Advances in Rare-event Simulation 3666 </div> 3667 <div class="session-chair"> 3668 Chair: Linyun He (Georgia Institute of Technology)<br /> 3669 </div> 3670 <div class="slot-entry"> 3671 <a name="inv199" tabindex="-1"></a> 3672 <div class="slot-title-line">
3673 <span class="slot-title" 3674 >Efficiency of Estimating Functions of Means in 3675 Rare-Event Contexts</span 3676 > 3677 </div> 3678 <div class="slot-authors"> 3679 Marvin Nakayama (New Jersey Institute of Technology) and 3680 Bruno Tuffin (INRIA, University of Rennes) 3681 </div> 3682 <div class="slot-abstract"> 3683 <div> 3684 <a 3685 class="clickable no-decoration" 3686 id="vhsjs_view_58_1707793551_4199407" 3687 onclick="$('#vhsjs_view_58_1707793551_4199407').hide(); 3688 $('#vhsjs_hide_58_1707793551_4199407').show(); 3689 $('#57_1707793551_4199326').slideDown(function() { 3690 if (typeof Masonry === 'function') { 3691 $('.use_masonry').masonry(); 3692 }; 3693 3694 });" 3695 ><i class="fa fa-caret-right"></i> 3696 <span class="hover_link">Abstract</span></a 3697 ><a 3698 class="clickable no-decoration" 3699 id="vhsjs_hide_58_1707793551_4199407" 3700 onclick="$('#57_1707793551_4199326').hide(function() { 3701 if (typeof Masonry === 'function') { 3702 $('.use_masonry').masonry(); 3703 }; 3704 }); 3705 $('#vhsjs_hide_58_1707793551_4199407').hide(); 3706 $('#vhsjs_view_58_1707793551_4199407').show();" 3707 style="display: none" 3708 ><i class="fa fa-caret-down"></i> 3709 <span class="hover_link">Abstract</span></a 3710 > 3711 <div 3712 data-display-control="58_1707793551_4199407" 3713 id="57_1707793551_4199326" 3714 style="display: none" 3715 > 3716 <div class="arrow-slidedown"> 3717 <blockquote> 3718 When estimating a function of means, where some 3719 but not necessarily all of them correspond to 3720 rare events, we provide conditions under which 3721 having efficient estimators of each individual 3722 mean leads to an efficient estimator of the 3723 function of the means. We illustrate this 3724 setting through several examples, and numerical 3725 results complement the theory. 3726 </blockquote> 3727 </div> 3728 </div> 3729 </div> 3730 </div> 3731 <div class="slot-urls"></div> 3732 <a href="/wsc23papers/029.pdf" target="_blank">pdf</a 3733 ><br /> 3734 </div> 3735 <div class="slot-entry"> 3736 <a name="con220" tabindex="-1"></a> 3737 <div class="slot-title-line"> 3738 <span class="slot-title" 3739 >Conditional Importance Sampling for Convex Rare-Event 3740 Sets</span 3741 > 3742 </div> 3743 <div class="slot-authors"> 3744 Dohyun Ahn and Lewen Zheng (The Chinese University of 3745 Hong Kong) 3746 </div> 3747 <div class="slot-abstract"> 3748 <div> 3749 <a 3750 class="clickable no-decoration" 3751 id="vhsjs_view_60_1707793551_4221368" 3752 onclick="$('#vhsjs_view_60_1707793551_4221368').hide(); 3753 $('#vhsjs_hide_60_1707793551_4221368').show(); 3754 $('#59_1707793551_4221287').slideDown(function() { 3755 if (typeof Masonry === 'function') { 3756 $('.use_masonry').masonry(); 3757 }; 3758 3759 });" 3760 ><i class="fa fa-caret-right"></i>
3761 <span class="hover_link">Abstract</span></a 3762 ><a 3763 class="clickable no-decoration" 3764 id="vhsjs_hide_60_1707793551_4221368" 3765 onclick="$('#59_1707793551_4221287').hide(function() { 3766 if (typeof Masonry === 'function') { 3767 $('.use_masonry').masonry(); 3768 }; 3769 }); 3770 $('#vhsjs_hide_60_1707793551_4221368').hide(); 3771 $('#vhsjs_view_60_1707793551_4221368').show();" 3772 style="display: none" 3773 ><i class="fa fa-caret-down"></i> 3774 <span class="hover_link">Abstract</span></a 3775 > 3776 <div 3777 data-display-control="60_1707793551_4221368" 3778 id="59_1707793551_4221287" 3779 style="display: none" 3780 > 3781 <div class="arrow-slidedown"> 3782 <blockquote> 3783 This paper studies the efficient estimation of 3784 expectations defined on convex rare-event sets 3785 using importance sampling. Classical importance 3786 sampling methods often neglect the geometry of 3787 the target set, resulting in a significant 3788 number of samples falling outside the target 3789 set. This can lead to an increase in the 3790 relative error of the estimator as the target 3791 event becomes rarer. To address this issue, we 3792 develop a conditional importance sampling scheme 3793 that achieves bounded relative error by changing 3794 the sampling distribution to ensure that a 3795 majority of samples lie inside the target set. 3796 The proposed method is easy to implement and 3797 significantly outperforms the existing 3798 approaches in various numerical experiments. 3799 </blockquote> 3800 </div> 3801 </div> 3802 </div> 3803 </div> 3804 <div class="slot-urls"></div> 3805 <a href="/wsc23papers/030.pdf" target="_blank">pdf</a 3806 ><br /> 3807 </div> 3808 <div class="slot-entry"> 3809 <a name="con359" tabindex="-1"></a> 3810 <div class="slot-title-line"> 3811 <span class="slot-title" 3812 >Curse of Dimensionality in Rare-Event 3813 Simulation</span 3814 > 3815 </div> 3816 <div> 3817 <span class="BTP award" 3818 >Best Contributed Theoretical Paper - Finalist</span 3819 > 3820 </div> 3821 <div class="slot-authors"> 3822 Yuanlu Bai, Antonius B. Dieker, and Henry Lam (Columbia 3823 University) 3824 </div> 3825 <div class="slot-abstract"> 3826 <div> 3827 <a 3828 class="clickable no-decoration" 3829 id="vhsjs_view_62_1707793551_4244947" 3830 onclick="$('#vhsjs_view_62_1707793551_4244947').hide(); 3831 $('#vhsjs_hide_62_1707793551_4244947').show(); 3832 $('#61_1707793551_424487').slideDown(function() { 3833 if (typeof Masonry === 'function') { 3834 $('.use_masonry').masonry(); 3835 }; 3836 3837 });" 3838 ><i class="fa fa-caret-right"></i>
3839 <span class="hover_link">Abstract</span></a 3840 ><a 3841 class="clickable no-decoration" 3842 id="vhsjs_hide_62_1707793551_4244947" 3843 onclick="$('#61_1707793551_424487').hide(function() { 3844 if (typeof Masonry === 'function') { 3845 $('.use_masonry').masonry(); 3846 }; 3847 }); 3848 $('#vhsjs_hide_62_1707793551_4244947').hide(); 3849 $('#vhsjs_view_62_1707793551_4244947').show();" 3850 style="display: none" 3851 ><i class="fa fa-caret-down"></i> 3852 <span class="hover_link">Abstract</span></a 3853 > 3854 <div 3855 data-display-control="62_1707793551_4244947" 3856 id="61_1707793551_424487" 3857 style="display: none" 3858 > 3859 <div class="arrow-slidedown"> 3860 <blockquote> 3861 In rare-event simulation, importance sampling 3862 (IS) is widely used to improve the efficiency of 3863 probability estimation. Asymptotic optimality is 3864 a common efficiency criterion, which requires 3865 that the relative error of the estimator only 3866 grows subexponentially in the rarity parameter. 3867 Most studies, however, consider low-dimensional 3868 problems and the effect of dimensionality is 3869 seldom analyzed. Motivated by recent AI-related 3870 applications, we take a first step towards 3871 high-dimensional rare-event simulation and 3872 demonstrate that for very simple examples, IS 3873 proposals that utilize exponential tilting, 3874 arguably the most common IS approach, can suffer 3875 from the "curse of dimensionality". That is, 3876 while the growth rate of the relative error is 3877 polynomial in the rarity parameter thus leading 3878 to asymptotic optimality, the degree of the 3879 polynomial depends on the problem 3880 dimensionality. Therefore, when the dimension is 3881 high, the relative error can be huge even in the 3882 rarity parameter regime where IS is 3883 conventionally believed to work well. 3884 </blockquote> 3885 </div> 3886 </div> 3887 </div> 3888 </div> 3889 <div class="slot-urls"></div> 3890 <a href="/wsc23papers/031.pdf" target="_blank">pdf</a 3891 ><br /> 3892 </div> 3893 </div> 3894 <div class="session-entry"> 3895 <span class="session-event-type">Technical Session</span 3896 ><span class="type-track-spacer"> · </span 3897 ><span class="program-track">Analysis Methodology</span 3898 ><br /> 3899 <div class="session-title"> 3900 Advances in Importance Sampling 3901 </div> 3902 <div class="session-chair"> 3903 Chair: Dohyun Ahn (The Chinese University of Hong Kong)<br /> 3904 </div> 3905 <div class="slot-entry"> 3906 <a name="con352" tabindex="-1"></a> 3907 <div class="slot-title-line"> 3908 <span class="slot-title" 3909 >Efficient Input Uncertainty Quantification for 3910 Regenerative Simulation</span 3911 > 3912 </div> 3913 <div>
3914 <span class="BTP award" 3915 >Best Contributed Theoretical Paper - Finalist</span 3916 > 3917 </div> 3918 <div class="slot-authors"> 3919 Linyun He (Georgia Institute of Technology), Mingbin Ben 3920 Feng (University of Waterloo), and Eunhye Song (Georgia 3921 Institute of Technology) 3922 </div> 3923 <div class="slot-abstract"> 3924 <div> 3925 <a 3926 class="clickable no-decoration" 3927 id="vhsjs_view_64_1707793551_4295719" 3928 onclick="$('#vhsjs_view_64_1707793551_4295719').hide(); 3929 $('#vhsjs_hide_64_1707793551_4295719').show(); 3930 $('#63_1707793551_4295633').slideDown(function() { 3931 if (typeof Masonry === 'function') { 3932 $('.use_masonry').masonry(); 3933 }; 3934 3935 });" 3936 ><i class="fa fa-caret-right"></i> 3937 <span class="hover_link">Abstract</span></a 3938 ><a 3939 class="clickable no-decoration" 3940 id="vhsjs_hide_64_1707793551_4295719" 3941 onclick="$('#63_1707793551_4295633').hide(function() { 3942 if (typeof Masonry === 'function') { 3943 $('.use_masonry').masonry(); 3944 }; 3945 }); 3946 $('#vhsjs_hide_64_1707793551_4295719').hide(); 3947 $('#vhsjs_view_64_1707793551_4295719').show();" 3948 style="display: none" 3949 ><i class="fa fa-caret-down"></i> 3950 <span class="hover_link">Abstract</span></a 3951 > 3952 <div 3953 data-display-control="64_1707793551_4295719" 3954 id="63_1707793551_4295633" 3955 style="display: none" 3956 > 3957 <div class="arrow-slidedown"> 3958 <blockquote> 3959 The initial bias in steady-state simulation can 3960 be characterized as the bias of a ratio 3961 estimator if the simulation model has a 3962 regenerative structure. This work tackles input 3963 uncertainty quantification for a regenerative 3964 simulation model when its input distributions 3965 are estimated from finite data. Our aim is to 3966 construct a bootstrap-based confidence interval 3967 (CI) for the true simulation output mean 3968 performance that provides a correct coverage 3969 with significantly less computational cost than 3970 the traditional methods. Exploiting the 3971 regenerative structure, we propose a $k$-nearest 3972 neighbor ($k$NN) ratio estimator for the 3973 steady-state performance measure at each set of 3974 bootstrapped input models and construct a 3975 bootstrap CI from the computed estimators. 3976 Asymptotically optimal choices for $k$ and 3977 bootstrap sample size are discussed. We further 3978 improve the CI by combining the $k$NN and 3979 likelihood ratio methods. We empirically compare 3980 the efficiency of the proposed estimators with 3981 the standard estimator using queueing examples. 3982 </blockquote> 3983 </div> 3984 </div> 3985 </div> 3986 </div> 3987 <div class="slot-urls"></div> 3988 <a href="/wsc23papers/032.pdf" target="_blank">pdf</a 3989 ><br /> 3990 </div> 3991 <div class="slot-entry"> 3992 <a name="inv146" tabindex="-1"></a> 3993 <div class="slot-title-line">
3994 <span class="slot-title" 3995 >Robust Importance Sampling for Stochastic Simulations 3996 with Uncertain Parametric Input Model</span 3997 > 3998 </div> 3999 <div class="slot-authors"> 4000 Seung Min Baik and Young Myoung Ko (Pohang University of 4001 Science and Technology (POSTECH)) and Eunshin Byon 4002 (University of Michigan) 4003 </div> 4004 <div class="slot-abstract"> 4005 <div> 4006 <a 4007 class="clickable no-decoration" 4008 id="vhsjs_view_66_1707793551_4319546" 4009 onclick="$('#vhsjs_view_66_1707793551_4319546').hide(); 4010 $('#vhsjs_hide_66_1707793551_4319546').show(); 4011 $('#65_1707793551_4319468').slideDown(function() { 4012 if (typeof Masonry === 'function') { 4013 $('.use_masonry').masonry(); 4014 }; 4015 4016 });" 4017 ><i class="fa fa-caret-right"></i> 4018 <span class="hover_link">Abstract</span></a 4019 ><a 4020 class="clickable no-decoration" 4021 id="vhsjs_hide_66_1707793551_4319546" 4022 onclick="$('#65_1707793551_4319468').hide(function() { 4023 if (typeof Masonry === 'function') { 4024 $('.use_masonry').masonry(); 4025 }; 4026 }); 4027 $('#vhsjs_hide_66_1707793551_4319546').hide(); 4028 $('#vhsjs_view_66_1707793551_4319546').show();" 4029 style="display: none" 4030 ><i class="fa fa-caret-down"></i> 4031 <span class="hover_link">Abstract</span></a 4032 > 4033 <div 4034 data-display-control="66_1707793551_4319546" 4035 id="65_1707793551_4319468" 4036 style="display: none" 4037 > 4038 <div class="arrow-slidedown"> 4039 <blockquote> 4040 In stochastic simulations, input model 4041 uncertainty may significantly impact output 4042 estimation accuracy. Although variance reduction 4043 techniques alleviate the computational burden, 4044 input model uncertainty remains unaddressed. 4045 Among several variance reduction techniques, we 4046 propose a robust version of the importance 4047 sampling method. We formulate a min-max 4048 optimization problem for finding a robust 4049 sampling density for simulation inputs 4050 considering a parametric uncertainty set that 4051 represents candidates of the true input 4052 distribution. We utilize the Bayesian 4053 optimization framework for solving the outer 4054 problem and the barrier method for tackling the 4055 inner problem. By incorporating input model 4056 uncertainty in the sampling stage, our method 4057 effectively allocates simulation effort to 4058 improve estimation robustness. Numerical 4059 experiments demonstrate the advantages of the 4060 proposed method over a benchmark model assuming 4061 a precisely known input model. Our approach 4062 produces more accurate output estimation (i.e., 4063 an estimator with lower variance), highlighting 4064 its robustness and potential applicability in a 4065 variety of situations. 4066 </blockquote> 4067 </div> 4068 </div> 4069 </div> 4070 </div> 4071 <div class="slot-urls"></div> 4072 <a href="/wsc23papers/033.pdf" target="_blank">pdf</a 4073 ><br /> 4074 </div> 4075 <div class="slot-entry"> 4076 <a name="con349" tabindex="-1"></a> 4077 <div class="slot-title-line">
4078 <span class="slot-title" 4079 >Generalized Importance Sampling for Nested 4080 Simulation</span 4081 > 4082 </div> 4083 <div class="slot-authors"> 4084 Qingyuan Chen (Cornell University) and Mingbin Ben Feng 4085 (University of Waterloo) 4086 </div> 4087 <div class="slot-abstract"> 4088 <div> 4089 <a 4090 class="clickable no-decoration" 4091 id="vhsjs_view_68_1707793551_434392" 4092 onclick="$('#vhsjs_view_68_1707793551_434392').hide(); 4093 $('#vhsjs_hide_68_1707793551_434392').show(); 4094 $('#67_1707793551_4343836').slideDown(function() { 4095 if (typeof Masonry === 'function') { 4096 $('.use_masonry').masonry(); 4097 }; 4098 4099 });" 4100 ><i class="fa fa-caret-right"></i> 4101 <span class="hover_link">Abstract</span></a 4102 ><a 4103 class="clickable no-decoration" 4104 id="vhsjs_hide_68_1707793551_434392" 4105 onclick="$('#67_1707793551_4343836').hide(function() { 4106 if (typeof Masonry === 'function') { 4107 $('.use_masonry').masonry(); 4108 }; 4109 }); 4110 $('#vhsjs_hide_68_1707793551_434392').hide(); 4111 $('#vhsjs_view_68_1707793551_434392').show();" 4112 style="display: none" 4113 ><i class="fa fa-caret-down"></i> 4114 <span class="hover_link">Abstract</span></a 4115 > 4116 <div 4117 data-display-control="68_1707793551_434392" 4118 id="67_1707793551_4343836" 4119 style="display: none" 4120 > 4121 <div class="arrow-slidedown"> 4122 <blockquote> 4123 Importance sampling (IS) is a classical variance 4124 reduction technique. Under mild conditions, an 4125 IS estimator is unbiased, so one often seeks 4126 variance-minimizing optimal sampling 4127 distribution. IS has remarkable success in many 4128 applications such as engineering, operations 4129 research, and finance. In some applications such 4130 as enterprise risk management and input 4131 uncertainty quantification, complex simulation 4132 designs such as nested simulation arises 4133 naturally: The outer-level simulation generates 4134 a set of risk factors, i.e., the scenarios, 4135 which are used as inputs for inner-level 4136 simulations. Nested simulation leads to wasteful 4137 use of computations as inner simulation outputs 4138 in each scenario are isolated from other 4139 scenarios. In this study, we propose, analyze, 4140 and test a generalized importance sampling 4141 technique for nested simulation. Our generalized 4142 IS approach reuses one set of inner simulation 4143 outputs across different outer scenarios. 4144 Numerical experiments show that our proposal is 4145 orders of magnitudes more efficient than the 4146 standard procedure. 4147 </blockquote> 4148 </div> 4149 </div> 4150 </div> 4151 </div> 4152 <div class="slot-urls"></div> 4153 <a href="/wsc23papers/034.pdf" target="_blank">pdf</a 4154 ><br /> 4155 </div> 4156 </div> 4157 <div class="session-entry">
4158 <span class="session-event-type">Technical Session</span 4159 ><span class="type-track-spacer"> · </span 4160 ><span class="program-track">Analysis Methodology</span 4161 ><br /> 4162 <div class="session-title">Output Analysis</div> 4163 <div class="session-chair"> 4164 Chair: Sara Shashaani (North Carolina State University)<br /> 4165 </div> 4166 <div class="slot-entry"> 4167 <a name="inv148" tabindex="-1"></a> 4168 <div class="slot-title-line"> 4169 <span class="slot-title" 4170 >Bootstrap Confidence Intervals for Simulation Output 4171 Parameters</span 4172 > 4173 </div> 4174 <div class="slot-authors"> 4175 Russell R. Barton (The Pennsylvania State University) 4176 and Luke A. Rhodes-Leader (Lancaster University) 4177 </div> 4178 <div class="slot-abstract"> 4179 <div> 4180 <a 4181 class="clickable no-decoration" 4182 id="vhsjs_view_70_1707793551_4400861" 4183 onclick="$('#vhsjs_view_70_1707793551_4400861').hide(); 4184 $('#vhsjs_hide_70_1707793551_4400861').show(); 4185 $('#69_1707793551_4400775').slideDown(function() { 4186 if (typeof Masonry === 'function') { 4187 $('.use_masonry').masonry(); 4188 }; 4189 4190 });" 4191 ><i class="fa fa-caret-right"></i> 4192 <span class="hover_link">Abstract</span></a 4193 ><a 4194 class="clickable no-decoration" 4195 id="vhsjs_hide_70_1707793551_4400861" 4196 onclick="$('#69_1707793551_4400775').hide(function() { 4197 if (typeof Masonry === 'function') { 4198 $('.use_masonry').masonry(); 4199 }; 4200 }); 4201 $('#vhsjs_hide_70_1707793551_4400861').hide(); 4202 $('#vhsjs_view_70_1707793551_4400861').show();" 4203 style="display: none" 4204 ><i class="fa fa-caret-down"></i> 4205 <span class="hover_link">Abstract</span></a 4206 > 4207 <div 4208 data-display-control="70_1707793551_4400861" 4209 id="69_1707793551_4400775" 4210 style="display: none" 4211 > 4212 <div class="arrow-slidedown"> 4213 <blockquote> 4214 Bootstrapping has been used to characterize the 4215 impact on discrete-event simulation output 4216 arising from input model uncertainty for thirty 4217 years. The distribution of simulation output 4218 statistics can be very non-normal, especially in 4219 simulation of heavily loaded queueing systems, 4220 and systems operating at a near optimal value of 4221 the output measure. This paper presents issues 4222 facing simulationists in using bootstrapping to 4223 provide confidence intervals for parameters 4224 related to the distribution of simulation output 4225 statistics, and identifies appropriate 4226 alternatives to the basic and percentile 4227 bootstrap methods. Both input uncertainty and 4228 ordinary output analysis settings are included. 4229 </blockquote> 4230 </div> 4231 </div> 4232 </div> 4233 </div> 4234 <div class="slot-urls"></div> 4235 <a href="/wsc23papers/035.pdf" target="_blank">pdf</a 4236 ><br /> 4237 </div> 4238 <div class="slot-entry"> 4239 <a name="inv170" tabindex="-1"></a> 4240 <div class="slot-title-line">
4241 <span class="slot-title" 4242 >Optimal Batching under Computation Budget</span 4243 > 4244 </div> 4245 <div class="slot-authors"> 4246 Shengyi He and Henry Lam (Columbia University) 4247 </div> 4248 <div class="slot-abstract"> 4249 <div> 4250 <a 4251 class="clickable no-decoration" 4252 id="vhsjs_view_72_1707793551_4422536" 4253 onclick="$('#vhsjs_view_72_1707793551_4422536').hide(); 4254 $('#vhsjs_hide_72_1707793551_4422536').show(); 4255 $('#71_1707793551_442245').slideDown(function() { 4256 if (typeof Masonry === 'function') { 4257 $('.use_masonry').masonry(); 4258 }; 4259 4260 });" 4261 ><i class="fa fa-caret-right"></i> 4262 <span class="hover_link">Abstract</span></a 4263 ><a 4264 class="clickable no-decoration" 4265 id="vhsjs_hide_72_1707793551_4422536" 4266 onclick="$('#71_1707793551_442245').hide(function() { 4267 if (typeof Masonry === 'function') { 4268 $('.use_masonry').masonry(); 4269 }; 4270 }); 4271 $('#vhsjs_hide_72_1707793551_4422536').hide(); 4272 $('#vhsjs_view_72_1707793551_4422536').show();" 4273 style="display: none" 4274 ><i class="fa fa-caret-down"></i> 4275 <span class="hover_link">Abstract</span></a 4276 > 4277 <div 4278 data-display-control="72_1707793551_4422536" 4279 id="71_1707793551_442245" 4280 style="display: none" 4281 > 4282 <div class="arrow-slidedown"> 4283 <blockquote> 4284 Batching methods operate by dividing data into 4285 batches and conducting inference by aggregating 4286 estimates from batched data. These methods have 4287 been used extensively in simulation output 4288 analysis and, among other strengths, an 4289 advantage is the light computation cost when 4290 using a small number of batches. However, under 4291 computation budget constraints, it is open to 4292 our knowledge which batching approach among the 4293 range of alternatives is statistically optimal, 4294 which is important in guiding procedural 4295 configuration. We show that standard batching, 4296 but also certain carefully designed schemes 4297 using uneven-size batches or overlapping 4298 batches, are large-sample optimal in the sense 4299 of so-called uniformly most accurate 4300 unbiasedness from a dual view of hypothesis 4301 testing. 4302 </blockquote> 4303 </div> 4304 </div> 4305 </div> 4306 </div> 4307 <div class="slot-urls"></div> 4308 <a href="/wsc23papers/036.pdf" target="_blank">pdf</a 4309 ><br /> 4310 </div> 4311 <div class="slot-entry"> 4312 <a name="inv129" tabindex="-1"></a> 4313 <div class="slot-title-line"> 4314 <span class="slot-title" 4315 >Confidence Intervals for Randomized Quasi-Monte Carlo 4316 Estimators</span 4317 > 4318 </div> 4319 <div class="slot-authors"> 4320 Pierre L'Ecuyer (Université de Montréal), 4321 Marvin K. Nakayama (New Jersey Institute of Technology), 4322 Art B. Owen (Stanford University), and Bruno Tuffin 4323 (Inria) 4324 </div> 4325 <div class="slot-abstract"> 4326 <div> 4327 <a 4328 class="clickable no-decoration" 4329 id="vhsjs_view_74_1707793551_4446888" 4330 onclick="$('#vhsjs_view_74_1707793551_4446888').hide(); 4331 $('#vhsjs_hide_74_1707793551_4446888').show(); 4332 $('#73_1707793551_4446805').slideDown(function() { 4333 if (typeof Masonry === 'function') { 4334 $('.use_masonry').masonry(); 4335 }; 4336 4337 });" 4338 ><i class="fa fa-caret-right"></i>
4339 <span class="hover_link">Abstract</span></a 4340 ><a 4341 class="clickable no-decoration" 4342 id="vhsjs_hide_74_1707793551_4446888" 4343 onclick="$('#73_1707793551_4446805').hide(function() { 4344 if (typeof Masonry === 'function') { 4345 $('.use_masonry').masonry(); 4346 }; 4347 }); 4348 $('#vhsjs_hide_74_1707793551_4446888').hide(); 4349 $('#vhsjs_view_74_1707793551_4446888').show();" 4350 style="display: none" 4351 ><i class="fa fa-caret-down"></i> 4352 <span class="hover_link">Abstract</span></a 4353 > 4354 <div 4355 data-display-control="74_1707793551_4446888" 4356 id="73_1707793551_4446805" 4357 style="display: none" 4358 > 4359 <div class="arrow-slidedown"> 4360 <blockquote> 4361 Randomized Quasi-Monte Carlo (RQMC) methods 4362 provide unbiased estimators whose variance often 4363 converges at a faster rate than standard Monte 4364 Carlo as a function of the sample size. However, 4365 computing valid confidence intervals is 4366 challenging because the observations from a 4367 single randomization are dependent and the 4368 central limit theorem does not ordinarily apply. 4369 A natural solution is to replicate the RQMC 4370 process independently a small number of times to 4371 estimate the variance and use a standard 4372 confidence interval based on a normal or Student 4373 t distribution. We investigate the standard 4374 Student t approach and two bootstrap methods for 4375 getting nonparametic confidence intervals for 4376 the mean using a modest number of replicates. 4377 Our main conclusion is that intervals based on 4378 the Student t distribution are more reliable 4379 than even the bootstrap t method on the 4380 integration problems arising from RQMC. 4381 </blockquote> 4382 </div> 4383 </div> 4384 </div> 4385 </div> 4386 <div class="slot-urls"></div> 4387 <a href="/wsc23papers/037.pdf" target="_blank">pdf</a 4388 ><br /> 4389 </div> 4390 </div> 4391 <div class="session-entry"> 4392 <span class="session-event-type">Technical Session</span 4393 ><span class="type-track-spacer"> · </span 4394 ><span class="program-track">Analysis Methodology</span 4395 ><br /> 4396 <div class="session-title">Steady-state Simulation</div> 4397 <div class="session-chair"> 4398 Chair: David Goldsman (Georgia Institute of Technology)<br /> 4399 </div> 4400 <div class="slot-entry"> 4401 <a name="con364" tabindex="-1"></a> 4402 <div class="slot-title-line"> 4403 <span class="slot-title" 4404 >A Fixed-Sample-Size Method for Estimating 4405 Steady-State Quantiles</span 4406 > 4407 </div> 4408 <div class="slot-authors"> 4409 Athanasios Lolos, Christos Alexopoulos, and David 4410 Goldsman (Georgia Institute of Technology); Kemal 4411 Dinçer Dingeç (Gebze Technical University); 4412 Anup C. Mokashi (Memorial Sloan Kettering Cancer 4413 Center); and James R. Wilson (North Carolina State 4414 University) 4415 </div> 4416 <div class="slot-abstract"> 4417 <div> 4418 <a 4419 class="clickable no-decoration" 4420 id="vhsjs_view_76_1707793551_4495392" 4421 onclick="$('#vhsjs_view_76_1707793551_4495392').hide(); 4422 $('#vhsjs_hide_76_1707793551_4495392').show(); 4423 $('#75_1707793551_449531').slideDown(function() { 4424 if (typeof Masonry === 'function') { 4425 $('.use_masonry').masonry(); 4426 }; 4427 4428 });" 4429 ><i class="fa fa-caret-right"></i>
4430 <span class="hover_link">Abstract</span></a 4431 ><a 4432 class="clickable no-decoration" 4433 id="vhsjs_hide_76_1707793551_4495392" 4434 onclick="$('#75_1707793551_449531').hide(function() { 4435 if (typeof Masonry === 'function') { 4436 $('.use_masonry').masonry(); 4437 }; 4438 }); 4439 $('#vhsjs_hide_76_1707793551_4495392').hide(); 4440 $('#vhsjs_view_76_1707793551_4495392').show();" 4441 style="display: none" 4442 ><i class="fa fa-caret-down"></i> 4443 <span class="hover_link">Abstract</span></a 4444 > 4445 <div 4446 data-display-control="76_1707793551_4495392" 4447 id="75_1707793551_449531" 4448 style="display: none" 4449 > 4450 <div class="arrow-slidedown"> 4451 <blockquote> 4452 We propose FQUEST, a fully automated 4453 fixed-sample-size procedure for computing 4454 confidence intervals (CIs) for steady-state 4455 quantiles. The user provides a 4456 (simulation-generated) dataset of arbitrary size 4457 and specifies the required quantile and nominal 4458 coverage probability of the anticipated CI. 4459 FQUEST incorporates the simulation analysis 4460 methods of batching, standardized time series 4461 (STS), and sectioning. Preliminary 4462 experimentation with the waiting-time process in 4463 a congested M/M/1 queueing system showed that 4464 FQUEST performed well by delivering CIs with 4465 estimated coverage probability close to the 4466 nominal level, even in unfavorable circumstances 4467 where the sample sizes were inadequate. In the 4468 latter cases and for very small samples for 4469 steady-state quantile estimation, the close 4470 conformance of the CI coverage probability 4471 typically came at the expense of loose CI 4472 precision. 4473 </blockquote> 4474 </div> 4475 </div> 4476 </div> 4477 </div> 4478 <div class="slot-urls"></div> 4479 <a href="/wsc23papers/038.pdf" target="_blank">pdf</a 4480 ><br /> 4481 </div> 4482 <div class="slot-entry"> 4483 <a name="inv209" tabindex="-1"></a> 4484 <div class="slot-title-line"> 4485 <span class="slot-title" 4486 >COSIMLA with General Regeneration Set to Compute 4487 Markov Chain Stationary Expectations</span 4488 > 4489 </div> 4490 <div class="slot-authors"> 4491 Peter W. Glynn (Stanford University) and Zeyu Zheng 4492 (University of California Berkeley) 4493 </div> 4494 <div class="slot-abstract"> 4495 <div> 4496 <a 4497 class="clickable no-decoration" 4498 id="vhsjs_view_78_1707793551_4517536" 4499 onclick="$('#vhsjs_view_78_1707793551_4517536').hide(); 4500 $('#vhsjs_hide_78_1707793551_4517536').show(); 4501 $('#77_1707793551_4517457').slideDown(function() { 4502 if (typeof Masonry === 'function') { 4503 $('.use_masonry').masonry(); 4504 }; 4505 4506 });" 4507 ><i class="fa fa-caret-right"></i>
4508 <span class="hover_link">Abstract</span></a 4509 ><a 4510 class="clickable no-decoration" 4511 id="vhsjs_hide_78_1707793551_4517536" 4512 onclick="$('#77_1707793551_4517457').hide(function() { 4513 if (typeof Masonry === 'function') { 4514 $('.use_masonry').masonry(); 4515 }; 4516 }); 4517 $('#vhsjs_hide_78_1707793551_4517536').hide(); 4518 $('#vhsjs_view_78_1707793551_4517536').show();" 4519 style="display: none" 4520 ><i class="fa fa-caret-down"></i> 4521 <span class="hover_link">Abstract</span></a 4522 > 4523 <div 4524 data-display-control="78_1707793551_4517536" 4525 id="77_1707793551_4517457" 4526 style="display: none" 4527 > 4528 <div class="arrow-slidedown"> 4529 <blockquote> 4530 We extend the COSIMLA approach (short for 4531 "COmbined SIMulation and Linear Algebra'') 4532 recently developed in Zheng, Infanger, and Glynn 4533 (2022) to compute stationary expectations for 4534 Markov chains with large or infinite discrete 4535 state space. Our work follows the idea of 4536 combing the best of linear algebra and 4537 simulation---using linear algebra to compute the 4538 "center'' of the state space and using 4539 simulation to compute the contributions from 4540 outside of the "center''. Different from Zheng, 4541 Infanger, and Glynn (2022) that needed to fix a 4542 single regeneration state, our work develops a 4543 new method that allows the use of a flexible 4544 regeneration set with a finite number of states. 4545 We show that this new method allows more 4546 efficient computation for the COSIMLA approach. 4547 </blockquote> 4548 </div> 4549 </div> 4550 </div> 4551 </div> 4552 <div class="slot-urls"></div> 4553 <a href="/wsc23papers/039.pdf" target="_blank">pdf</a 4554 ><br /> 4555 </div> 4556 <div class="slot-entry"> 4557 <a name="inv140" tabindex="-1"></a> 4558 <div class="slot-title-line"> 4559 <span class="slot-title" 4560 >Fast Approximation to Discrete-Event Simulation of 4561 Markovian Queueing Networks</span 4562 > 4563 </div> 4564 <div class="slot-authors"> 4565 Tan Wang (Fudan University), Yingda Song (Shanghai 4566 Jiaotong University), and Jeff Hong (Fudan University) 4567 </div> 4568 <div class="slot-abstract"> 4569 <div> 4570 <a 4571 class="clickable no-decoration" 4572 id="vhsjs_view_80_1707793551_4541314" 4573 onclick="$('#vhsjs_view_80_1707793551_4541314').hide(); 4574 $('#vhsjs_hide_80_1707793551_4541314').show(); 4575 $('#79_1707793551_4541235').slideDown(function() { 4576 if (typeof Masonry === 'function') { 4577 $('.use_masonry').masonry(); 4578 }; 4579 4580 });" 4581 ><i class="fa fa-caret-right"></i>
4582 <span class="hover_link">Abstract</span></a 4583 ><a 4584 class="clickable no-decoration" 4585 id="vhsjs_hide_80_1707793551_4541314" 4586 onclick="$('#79_1707793551_4541235').hide(function() { 4587 if (typeof Masonry === 'function') { 4588 $('.use_masonry').masonry(); 4589 }; 4590 }); 4591 $('#vhsjs_hide_80_1707793551_4541314').hide(); 4592 $('#vhsjs_view_80_1707793551_4541314').show();" 4593 style="display: none" 4594 ><i class="fa fa-caret-down"></i> 4595 <span class="hover_link">Abstract</span></a 4596 > 4597 <div 4598 data-display-control="80_1707793551_4541314" 4599 id="79_1707793551_4541235" 4600 style="display: none" 4601 > 4602 <div class="arrow-slidedown"> 4603 <blockquote> 4604 Simulation of queueing networks is generally 4605 carried out by discrete-event simulation (DES), 4606 in which the simulation time is driven by the 4607 occurrence of the next event. However, for 4608 large-scale queueing networks, especially when 4609 the network is very busy, keeping track of all 4610 events is computationally inefficient. Moreover, 4611 as the traditional DES is inherently sequential, 4612 it is difficult to harness the capability of 4613 parallel computing. In this paper, we propose a 4614 parallel fast simulation approximation framework 4615 for large-scale Markovian queueing networks, 4616 where the simulation horizon is discretized into 4617 small time intervals and the system state is 4618 updated according to the events happening in 4619 each time interval. The computational complexity 4620 analysis demonstrates that our method is more 4621 efficient for large-scale networks compared with 4622 traditional DES. We also show its relative error 4623 converges to zero. The experimental results show 4624 that our framework can be much faster than the 4625 state-of-the-art DES tools. 4626 </blockquote> 4627 </div> 4628 </div> 4629 </div> 4630 </div> 4631 <div class="slot-urls"></div> 4632 <a href="/wsc23papers/304.pdf" target="_blank">pdf</a 4633 ><br /> 4634 </div> 4635 </div> 4636 <div class="session-entry"> 4637 <span class="session-event-type">Technical Session</span 4638 ><span class="type-track-spacer"> · </span 4639 ><span class="program-track">Analysis Methodology</span 4640 ><br /> 4641 <div class="session-title"> 4642 Innovative Applications of Simulation Methodology 4643 </div> 4644 <div class="session-chair"> 4645 Chair: Hua Zheng (Northeastern University)<br /> 4646 </div> 4647 <div class="slot-entry"> 4648 <a name="inv178" tabindex="-1"></a> 4649 <div class="slot-title-line"> 4650 <span class="slot-title" 4651 >Structure-function Dynamics Hybrid Modeling: RNA 4652 Degradation</span 4653 > 4654 </div> 4655 <div class="slot-authors"> 4656 Hua Zheng, Wei Xie, Paul C. Whitford, Ailun Wang, 4657 Chunsheng Fang, and Wandi Xu (Northeastern
4657University) 4658 </div> 4659 <div class="slot-abstract"> 4660 <div> 4661 <a 4662 class="clickable no-decoration" 4663 id="vhsjs_view_82_1707793551_4592648" 4664 onclick="$('#vhsjs_view_82_1707793551_4592648').hide(); 4665 $('#vhsjs_hide_82_1707793551_4592648').show(); 4666 $('#81_1707793551_4592564').slideDown(function() { 4667 if (typeof Masonry === 'function') { 4668 $('.use_masonry').masonry(); 4669 }; 4670 4671 });" 4672 ><i class="fa fa-caret-right"></i> 4673 <span class="hover_link">Abstract</span></a 4674 ><a 4675 class="clickable no-decoration" 4676 id="vhsjs_hide_82_1707793551_4592648" 4677 onclick="$('#81_1707793551_4592564').hide(function() { 4678 if (typeof Masonry === 'function') { 4679 $('.use_masonry').masonry(); 4680 }; 4681 }); 4682 $('#vhsjs_hide_82_1707793551_4592648').hide(); 4683 $('#vhsjs_view_82_1707793551_4592648').show();" 4684 style="display: none" 4685 ><i class="fa fa-caret-down"></i> 4686 <span class="hover_link">Abstract</span></a 4687 > 4688 <div 4689 data-display-control="82_1707793551_4592648" 4690 id="81_1707793551_4592564" 4691 style="display: none" 4692 > 4693 <div class="arrow-slidedown"> 4694 <blockquote> 4695 RNA structure and functional dynamics play 4696 fundamental roles in controlling biological 4697 systems. Molecular dynamics simulation, which 4698 can characterize interactions at an atomistic 4699 level, can advance the understanding on new drug 4700 discovery, manufacturing, and delivery 4701 mechanisms. However, it is computationally 4702 unattainable to support the development of a 4703 digital twin for enzymatic reaction network 4704 mechanism learning, and end-to-end bioprocess 4705 design and control. Thus, we create a hybrid 4706 ("mechanistic + machine learning") model 4707 characterizing the interdependence of RNA 4708 structure and functional dynamics from atomistic 4709 to macroscopic levels. To assess the proposed 4710 modeling strategy, we consider RNA degradation 4711 which is a critical process in cellular biology 4712 that affects gene expression. The empirical 4713 study on RNA lifetime prediction demonstrates 4714 the promising performance of the proposed 4715 multi-scale bioprocess hybrid modeling strategy. 4716 </blockquote> 4717 </div> 4718 </div> 4719 </div> 4720 </div> 4721 <div class="slot-urls"></div> 4722 <a href="/wsc23papers/040.pdf" target="_blank">pdf</a 4723 ><br /> 4724 </div> 4725 <div class="slot-entry"> 4726 <a name="inv141" tabindex="-1"></a> 4727 <div class="slot-title-line"> 4728 <span class="slot-title" 4729 >Tracking and Detecting Systematic Errors in Digital 4730 Twins</span 4731 > 4732 </div> 4733 <div class="slot-authors"> 4734 Luke A. Rhodes-Leader (Lancaster University) and Barry 4735 L. Nelson (Northwestern University) 4736 </div> 4737 <div class="slot-abstract"> 4738 <div> 4739 <a 4740 class="clickable no-decoration" 4741 id="vhsjs_view_84_1707793551_4614694" 4742 onclick="$('#vhsjs_view_84_1707793551_4614694').hide(); 4743 $('#vhsjs_hide_84_1707793551_4614694').show(); 4744 $('#83_1707793551_4614613').slideDown(function() { 4745 if (typeof Masonry === 'function') { 4746 $('.use_masonry').masonry(); 4747 }; 4748 4749 });" 4750 ><i class="fa fa-caret-right"></i>
4751 <span class="hover_link">Abstract</span></a 4752 ><a 4753 class="clickable no-decoration" 4754 id="vhsjs_hide_84_1707793551_4614694" 4755 onclick="$('#83_1707793551_4614613').hide(function() { 4756 if (typeof Masonry === 'function') { 4757 $('.use_masonry').masonry(); 4758 }; 4759 }); 4760 $('#vhsjs_hide_84_1707793551_4614694').hide(); 4761 $('#vhsjs_view_84_1707793551_4614694').show();" 4762 style="display: none" 4763 ><i class="fa fa-caret-down"></i> 4764 <span class="hover_link">Abstract</span></a 4765 > 4766 <div 4767 data-display-control="84_1707793551_4614694" 4768 id="83_1707793551_4614613" 4769 style="display: none" 4770 > 4771 <div class="arrow-slidedown"> 4772 <blockquote> 4773 Digital Twins (DTs) have immense promise for 4774 exploiting the power of computer simulation to 4775 control large-scale real-world systems. The key 4776 idea is to evaluate or optimize decisions using 4777 the DT, and then implement them in the 4778 real-world system. Even with best practices, the 4779 DT and the real-world system may become 4780 misaligned over time. In this paper we provide a 4781 statistical method to detect such misalignment 4782 even though both the simulation and the 4783 real-world system are inherently stochastic. An 4784 empirical evaluation and a realistic 4785 illustration are provided. 4786 </blockquote> 4787 </div> 4788 </div> 4789 </div> 4790 </div> 4791 <div class="slot-urls"></div> 4792 <a href="/wsc23papers/041.pdf" target="_blank">pdf</a 4793 ><br /> 4794 </div> 4795 <div class="slot-entry"> 4796 <a name="inv131" tabindex="-1"></a> 4797 <div class="slot-title-line"> 4798 <span class="slot-title" 4799 >Sensitivity Analysis for Stopping Criteria with 4800 Application to Organ Transplantations</span 4801 > 4802 </div> 4803 <div class="slot-authors"> 4804 Xingyu Ren, Michael Fu, and Steven Marcus (University of 4805 Maryland) 4806 </div> 4807 <div class="slot-abstract"> 4808 <div> 4809 <a 4810 class="clickable no-decoration" 4811 id="vhsjs_view_86_1707793551_4637747" 4812 onclick="$('#vhsjs_view_86_1707793551_4637747').hide(); 4813 $('#vhsjs_hide_86_1707793551_4637747').show(); 4814 $('#85_1707793551_4637668').slideDown(function() { 4815 if (typeof Masonry === 'function') { 4816 $('.use_masonry').masonry(); 4817 }; 4818 4819 });" 4820 ><i class="fa fa-caret-right"></i> 4821 <span class="hover_link">Abstract</span></a 4822 ><a 4823 class="clickable no-decoration" 4824 id="vhsjs_hide_86_1707793551_4637747" 4825 onclick="$('#85_1707793551_4637668').hide(function() { 4826 if (typeof Masonry === 'function') { 4827 $('.use_masonry').masonry(); 4828 }; 4829 }); 4830 $('#vhsjs_hide_86_1707793551_4637747').hide(); 4831 $('#vhsjs_view_86_1707793551_4637747').show();" 4832 style="display: none" 4833 ><i class="fa fa-caret-down"></i>
4834 <span class="hover_link">Abstract</span></a 4835 > 4836 <div 4837 data-display-control="86_1707793551_4637747" 4838 id="85_1707793551_4637668" 4839 style="display: none" 4840 > 4841 <div class="arrow-slidedown"> 4842 <blockquote> 4843 We consider a stopping problem and its 4844 application to the decision-making process 4845 regarding the optimal timing of organ 4846 transplantation for individual patients. At each 4847 decision period, the patient state is inspected 4848 and a decision is made whether to transplant. If 4849 the organ is transplanted, the process 4850 terminates; otherwise, the process continues 4851 until a transplant happens or the patient dies. 4852 Under suitable conditions, we show that there 4853 exists a control limit optimal policy. We 4854 propose a smoothed perturbation analysis (SPA) 4855 estimator for the gradient of the total expected 4856 discounted reward with respect to the control 4857 limit. Moreover, we show that the SPA estimator 4858 is asymptotically unbiased. 4859 </blockquote> 4860 </div> 4861 </div> 4862 </div> 4863 </div> 4864 <div class="slot-urls"></div> 4865 <a href="/wsc23papers/042.pdf" target="_blank">pdf</a 4866 ><br /> 4867 </div> 4868 </div> 4869 <div class="session-entry"> 4870 <span class="session-event-type">Technical Session</span 4871 ><span class="type-track-spacer"> · </span 4872 ><span class="program-track">Analysis Methodology</span 4873 ><br /> 4874 <div class="session-title"> 4875 Design of Experiments and Screening 4876 </div> 4877 <div class="session-chair"> 4878 Chair: Zeyu Zheng (University of California, Berkeley)<br /> 4879 </div> 4880 <div class="slot-entry"> 4881 <a name="con124" tabindex="-1"></a> 4882 <div class="slot-title-line"> 4883 <span class="slot-title" 4884 >The Variability in Design Quality Measures for 4885 Multiple Types of Space-filling Designs Created by 4886 Leading Software Packages</span 4887 > 4888 </div> 4889 <div class="slot-authors"> 4890 Thomas W. Lucas (Naval Postgraduate School) and Jeffrey 4891 D. Parker (United States Marine Corps) 4892 </div> 4893 <div class="slot-abstract"> 4894 <div> 4895 <a 4896 class="clickable no-decoration" 4897 id="vhsjs_view_88_1707793551_4696136" 4898 onclick="$('#vhsjs_view_88_1707793551_4696136').hide(); 4899 $('#vhsjs_hide_88_1707793551_4696136').show(); 4900 $('#87_1707793551_4696057').slideDown(function() { 4901 if (typeof Masonry === 'function') { 4902 $('.use_masonry').masonry(); 4903 }; 4904 4905 });" 4906 ><i class="fa fa-caret-right"></i> 4907 <span class="hover_link">Abstract</span></a 4908 ><a 4909 class="clickable no-decoration" 4910 id="vhsjs_hide_88_1707793551_4696136" 4911 onclick="$('#87_1707793551_4696057').hide(function() { 4912 if (typeof Masonry === 'function') { 4913 $('.use_masonry').masonry(); 4914 }; 4915 }); 4916 $('#vhsjs_hide_88_1707793551_4696136').hide(); 4917 $('#vhsjs_view_88_1707793551_4696136').show();" 4918 style="display: none" 4919 ><i class="fa fa-caret-down"></i>
4920 <span class="hover_link">Abstract</span></a 4921 > 4922 <div 4923 data-display-control="88_1707793551_4696136" 4924 id="87_1707793551_4696057" 4925 style="display: none" 4926 > 4927 <div class="arrow-slidedown"> 4928 <blockquote> 4929 Space-filling designs (SFDs) underpin many 4930 large-scale simulation studies. The algorithms 4931 that construct SFDs are mostly stochastic and 4932 cannot guarantee that optimal solutions can be 4933 found within a practical amount of time. This 4934 paper uses massive experimentation to find the 4935 empirical distributions of a diverse set of 4936 design-quality measures in highly-used classes 4937 of SFDs constructed by leading software 4938 packages. The objective is to provide simulation 4939 practitioners with a better understanding of 4940 what they can expect from different SFD choices. 4941 The results show substantial variability in 4942 measures of correlation and space-fillingness in 4943 the design classes and dimensions investigated. 4944 Therefore, computer experimenters should 4945 generate and assess several candidate designs 4946 using different random-number-generator seeds to 4947 reduce the risk of using a poor design simply 4948 due to random chance. We also find that in the 4949 largest designs investigated, the uniform 4950 designs generally perform best for both our 4951 correlation and uniformity measures. 4952 </blockquote> 4953 </div> 4954 </div> 4955 </div> 4956 </div> 4957 <div class="slot-urls"></div> 4958 <a href="/wsc23papers/043.pdf" target="_blank">pdf</a 4959 ><br /> 4960 </div> 4961 <div class="slot-entry"> 4962 <a name="con212" tabindex="-1"></a> 4963 <div class="slot-title-line"> 4964 <span class="slot-title" 4965 >Top-m Factor Screening for Stochastic Simulation: 4966 Multi-Armed Bandit And Sequential Bifurcation 4967 Combined</span 4968 > 4969 </div> 4970 <div class="slot-authors"> 4971 Wen Shi (Central South University), Hong Wan (North 4972 Carolina State University), and Xiang Xie (Central South 4973 University) 4974 </div> 4975 <div class="slot-abstract"> 4976 <div> 4977 <a 4978 class="clickable no-decoration" 4979 id="vhsjs_view_90_1707793551_4720023" 4980 onclick="$('#vhsjs_view_90_1707793551_4720023').hide(); 4981 $('#vhsjs_hide_90_1707793551_4720023').show(); 4982 $('#89_1707793551_4719942').slideDown(function() { 4983 if (typeof Masonry === 'function') { 4984 $('.use_masonry').masonry(); 4985 }; 4986 4987 });" 4988 ><i class="fa fa-caret-right"></i> 4989 <span class="hover_link">Abstract</span></a 4990 ><a 4991 class="clickable no-decoration" 4992 id="vhsjs_hide_90_1707793551_4720023" 4993 onclick="$('#89_1707793551_4719942').hide(function() { 4994 if (typeof Masonry === 'function') { 4995 $('.use_masonry').masonry(); 4996 }; 4997 }); 4998 $('#vhsjs_hide_90_1707793551_4720023').hide(); 4999 $('#vhsjs_view_90_1707793551_4720023').show();" 5000 style="display: none" 5001 ><i class="fa fa-caret-down"></i>
5002 <span class="hover_link">Abstract</span></a 5003 > 5004 <div 5005 data-display-control="90_1707793551_4720023" 5006 id="89_1707793551_4719942" 5007 style="display: none" 5008 > 5009 <div class="arrow-slidedown"> 5010 <blockquote> 5011 We propose a novel screening framework 5012 (abbreviated to TopmSB) to identify the top m 5013 key factors affecting the system performance. 5014 Our framework builds on the standard SB 5015 screening mechanism but incorporates an adaptive 5016 multi-armed bandit (MAB) procedure in each stage 5017 to prioritize the largest group. Compared to SB, 5018 TopmSB avoids specifying perplexing 5019 (un)importance threshold parameters, while 5020 providing desired computational efficiency and 5021 statistical precision guarantee. Numerical 5022 experiments demonstrate the efficiency and 5023 effectiveness of the proposed method. 5024 </blockquote> 5025 </div> 5026 </div> 5027 </div> 5028 </div> 5029 <div class="slot-urls"></div> 5030 <a href="/wsc23papers/044.pdf" target="_blank">pdf</a 5031 ><br /> 5032 </div> 5033 <div class="slot-entry"> 5034 <a name="con134" tabindex="-1"></a> 5035 <div class="slot-title-line"> 5036 <span class="slot-title" 5037 >Best Arm Identification with Fairness Constraints on 5038 Subpopulations</span 5039 > 5040 </div> 5041 <div class="slot-authors"> 5042 Yuhang Wu, Zeyu Zheng, and Tingyu Zhu (University of 5043 California, Berkeley) 5044 </div> 5045 <div class="slot-abstract"> 5046 <div> 5047 <a 5048 class="clickable no-decoration" 5049 id="vhsjs_view_92_1707793551_474383" 5050 onclick="$('#vhsjs_view_92_1707793551_474383').hide(); 5051 $('#vhsjs_hide_92_1707793551_474383').show(); 5052 $('#91_1707793551_4743748').slideDown(function() { 5053 if (typeof Masonry === 'function') { 5054 $('.use_masonry').masonry(); 5055 }; 5056 5057 });" 5058 ><i class="fa fa-caret-right"></i> 5059 <span class="hover_link">Abstract</span></a 5060 ><a 5061 class="clickable no-decoration" 5062 id="vhsjs_hide_92_1707793551_474383" 5063 onclick="$('#91_1707793551_4743748').hide(function() { 5064 if (typeof Masonry === 'function') { 5065 $('.use_masonry').masonry(); 5066 }; 5067 }); 5068 $('#vhsjs_hide_92_1707793551_474383').hide(); 5069 $('#vhsjs_view_92_1707793551_474383').show();" 5070 style="display: none" 5071 ><i class="fa fa-caret-down"></i> 5072 <span class="hover_link">Abstract</span></a 5073 > 5074 <div 5075 data-display-control="92_1707793551_474383" 5076 id="91_1707793551_4743748" 5077 style="display: none" 5078 > 5079 <div class="arrow-slidedown"> 5080 <blockquote> 5081 We formulate, analyze and solve the problem of 5082 best arm identification with fairness 5083 constraints on subpopulations (BAICS). Standard 5084 best arm identification problems aim at 5085 selecting an arm that has the largest expected 5086 reward where the expectation is taken over the 5087 entire population. The BAICS problem requires 5088 that a selected arm must be fair to all 5089 subpopulations (e.g., different ethnic groups or 5090 different types of customers) by satisfying 5091 constraints that the expected reward conditional 5092 on every subpopulation needs to be larger than 5093 some thresholds. The BAICS problem aims at 5094 correctly identify, with high confidence, the 5095 arm with the largest expected reward from all 5096 arms that satisfy subpopulation constraints. We 5097 analyze the complexity of the BAICS problem by 5098 proving a best achievable lower bound on the 5099 sample complexity with closed-form 5100 representation. We then design an algorithm and 5101 prove the sample complexity to match with the 5102 lower bound in terms of order. 5103 </blockquote> 5104 </div> 5105 </div> 5106 </div> 5107 </div> 5108 <div class="slot-urls"></div> 5109 <a href="/wsc23papers/045.pdf" target="_blank">pdf</a 5110 ><br /> 5111 </div> 5112 </div> 5113 <div class="session-entry">
5114 <span class="session-event-type">Technical Session</span 5115 ><span class="type-track-spacer"> · </span 5116 ><span class="program-track">Analysis Methodology</span 5117 ><br /> 5118 <div class="session-title"> 5119 Analysis Uses in Optimization 5120 </div> 5121 <div class="session-chair"> 5122 Chair: Ilya Ryzhov (University of Maryland)<br /> 5123 </div> 5124 <div class="slot-entry"> 5125 <a name="con138" tabindex="-1"></a> 5126 <div class="slot-title-line"> 5127 <span class="slot-title" 5128 >Efficient Bandwidth Selection for Kernel Density 5129 Estimation</span 5130 > 5131 </div> 5132 <div class="slot-authors"> 5133 Haidong Li (University of Chinese Academy of Sciences), 5134 Long Wang and Yijie Peng (Peking University), and Di 5135 Wang (Shanghai Jiao Tong University) 5136 </div> 5137 <div class="slot-abstract"> 5138 <div> 5139 <a 5140 class="clickable no-decoration" 5141 id="vhsjs_view_94_1707793551_4804628" 5142 onclick="$('#vhsjs_view_94_1707793551_4804628').hide(); 5143 $('#vhsjs_hide_94_1707793551_4804628').show(); 5144 $('#93_1707793551_480455').slideDown(function() { 5145 if (typeof Masonry === 'function') { 5146 $('.use_masonry').masonry(); 5147 }; 5148 5149 });" 5150 ><i class="fa fa-caret-right"></i> 5151 <span class="hover_link">Abstract</span></a 5152 ><a 5153 class="clickable no-decoration" 5154 id="vhsjs_hide_94_1707793551_4804628" 5155 onclick="$('#93_1707793551_480455').hide(function() { 5156 if (typeof Masonry === 'function') { 5157 $('.use_masonry').masonry(); 5158 }; 5159 }); 5160 $('#vhsjs_hide_94_1707793551_4804628').hide(); 5161 $('#vhsjs_view_94_1707793551_4804628').show();" 5162 style="display: none" 5163 ><i class="fa fa-caret-down"></i> 5164 <span class="hover_link">Abstract</span></a 5165 > 5166 <div 5167 data-display-control="94_1707793551_4804628" 5168 id="93_1707793551_480455" 5169 style="display: none" 5170 > 5171 <div class="arrow-slidedown"> 5172 <blockquote> 5173 We consider bandwidth selection for kernel 5174 density estimation. The performance of kernel 5175 density estimator heavily relies on the quality 5176 of the bandwidth. In this paper, we propose an 5177 efficient plug-in kernel density estimator which 5178 first perturbs the bandwidth to estimate the 5179 optimal bandwidth, followed by applying a kernel 5180 density estimator with the estimated optimal 5181 bandwidth. The proposed method utilizes the 5182 zeroth-order information of kernel function and 5183 has a faster convergence rate than other plug-in 5184 methods in existing literature. Simulation 5185 results demonstrate superior finite sample 5186 performance and robustness of the proposed 5187 method. 5188 </blockquote> 5189 </div> 5190 </div> 5191 </div> 5192 </div> 5193 <div class="slot-urls"></div> 5194 <a href="/wsc23papers/046.pdf" target="_blank">pdf</a 5195 ><br /> 5196 </div> 5197 <div class="slot-entry"> 5198 <a name="con303" tabindex="-1"></a> 5199 <div class="slot-title-line">
5200 <span class="slot-title" 5201 >CGPT: A Conditional Gaussian Process Tree for 5202 Grey-Box Bayesian Optimization</span 5203 > 5204 </div> 5205 <div class="slot-authors"> 5206 Mengrui (Mina) Jiang, Tanmay Khandait, and Giulia 5207 Pedrielli (Arizona State University) 5208 </div> 5209 <div class="slot-abstract"> 5210 <div> 5211 <a 5212 class="clickable no-decoration" 5213 id="vhsjs_view_96_1707793551_482723" 5214 onclick="$('#vhsjs_view_96_1707793551_482723').hide(); 5215 $('#vhsjs_hide_96_1707793551_482723').show(); 5216 $('#95_1707793551_4827154').slideDown(function() { 5217 if (typeof Masonry === 'function') { 5218 $('.use_masonry').masonry(); 5219 }; 5220 5221 });" 5222 ><i class="fa fa-caret-right"></i> 5223 <span class="hover_link">Abstract</span></a 5224 ><a 5225 class="clickable no-decoration" 5226 id="vhsjs_hide_96_1707793551_482723" 5227 onclick="$('#95_1707793551_4827154').hide(function() { 5228 if (typeof Masonry === 'function') { 5229 $('.use_masonry').masonry(); 5230 }; 5231 }); 5232 $('#vhsjs_hide_96_1707793551_482723').hide(); 5233 $('#vhsjs_view_96_1707793551_482723').show();" 5234 style="display: none" 5235 ><i class="fa fa-caret-down"></i> 5236 <span class="hover_link">Abstract</span></a 5237 > 5238 <div 5239 data-display-control="96_1707793551_482723" 5240 id="95_1707793551_4827154" 5241 style="display: none" 5242 > 5243 <div class="arrow-slidedown"> 5244 <blockquote> 5245 In black-box optimization problems, Bayesian 5246 optimization algorithms are often applied by 5247 generating inputs and measure values to discover 5248 hidden structure and determine where to sample 5249 sequentially. However, information about system 5250 properties can be available. In different 5251 learning tasks, we may know that the objective 5252 is the minimum of functions, or a network. In 5253 this paper we consider the case where the 5254 structure of the objective function can be 5255 encoded as a tree. We propose the new 5256 Conditional Gaussian Process tree (CGPT) model 5257 for "tree functions'' to embed the function 5258 structure and improving the prediction power of 5259 the Gaussian process. We utilize the 5260 intermediate information at the tree nodes, to 5261 formulate a novel likelihood for the estimation 5262 of the CGPT parameters. We formulate the 5263 learning and investigate the performance of the 5264 proposed approach. Our study shows that CGPT 5265 always outperforms a single Gaussian process 5266 model. 5267 </blockquote> 5268 </div> 5269 </div> 5270 </div> 5271 </div> 5272 <div class="slot-urls"></div> 5273 <a href="/wsc23papers/047.pdf" target="_blank">pdf</a 5274 ><br /> 5275 </div> 5276 <div class="slot-entry"> 5277 <a name="inv133" tabindex="-1"></a> 5278 <div class="slot-title-line">
5279 <span class="slot-title" 5280 >Mean-Variance Portfolio Optimization with Nonlinear 5281 Derivative Securities</span 5282 > 5283 </div> 5284 <div class="slot-authors"> 5285 Shiyu Wang and Guowei Cai (Lingnan College, Sun Yat-sen 5286 University); Peiwen Yu (Soochow University); Guangwu Liu 5287 (City University of Hong Kong); and Jun Luo (Shanghai 5288 Jiao Tong University) 5289 </div> 5290 <div class="slot-abstract"> 5291 <div> 5292 <a 5293 class="clickable no-decoration" 5294 id="vhsjs_view_98_1707793551_485117" 5295 onclick="$('#vhsjs_view_98_1707793551_485117').hide(); 5296 $('#vhsjs_hide_98_1707793551_485117').show(); 5297 $('#97_1707793551_4851089').slideDown(function() { 5298 if (typeof Masonry === 'function') { 5299 $('.use_masonry').masonry(); 5300 }; 5301 5302 });" 5303 ><i class="fa fa-caret-right"></i> 5304 <span class="hover_link">Abstract</span></a 5305 ><a 5306 class="clickable no-decoration" 5307 id="vhsjs_hide_98_1707793551_485117" 5308 onclick="$('#97_1707793551_4851089').hide(function() { 5309 if (typeof Masonry === 'function') { 5310 $('.use_masonry').masonry(); 5311 }; 5312 }); 5313 $('#vhsjs_hide_98_1707793551_485117').hide(); 5314 $('#vhsjs_view_98_1707793551_485117').show();" 5315 style="display: none" 5316 ><i class="fa fa-caret-down"></i> 5317 <span class="hover_link">Abstract</span></a 5318 > 5319 <div 5320 data-display-control="98_1707793551_485117" 5321 id="97_1707793551_4851089" 5322 style="display: none" 5323 > 5324 <div class="arrow-slidedown"> 5325 <blockquote> 5326 In this paper, we propose a simulation approach 5327 to mean-variance optimization for portfolios 5328 comprised of derivative securities. The key of 5329 the proposed method is on the development of an 5330 unbiased and consistent estimator of the 5331 covariance matrix of asset returns which do not 5332 admit closed-form formulas but require Monte 5333 Carlo estimation, leading to a sample-based 5334 optimization problem that is easy to solve. We 5335 characterize the asymptotic properties of the 5336 proposed covariance estimator, and the solution 5337 to and the objective value of the sample-based 5338 optimization problem. Performance of the 5339 proposed approach is demonstrated via numerical 5340 experiments. 5341 </blockquote> 5342 </div> 5343 </div> 5344 </div> 5345 </div> 5346 <div class="slot-urls"></div> 5347 <a href="/wsc23papers/048.pdf" target="_blank">pdf</a 5348 ><br /> 5349 </div> 5350 </div> 5351 </div> 5352 <div class="centered"> 5353 <div class="top-link"><a href="#top">Return to Top</a></div> 5354 </div> 5355 <hr /> 5356 </div> 5357 <div class="area-section"> 5358 <div class="centered"> 5359 <a name="ptrack110" tabindex="-1"></a> 5360 <div class="section-title"> 5361 Aviation Modeling and Analysis 5362 </div> 5363 </div> 5364 <div class="centered track-chair">
5365 <span class="track-chair-role" 5366 >Track Coordinator - Aviation Modeling and Analysis: </span 5367 ><span class="track-chair-names" 5368 >Sameer Alam (Nanyang Technological University), Miguel 5369 Mujica Mota (Amsterdam University of Applied Sciences), 5370 Michael Schultz (Bundeswehr University Munich)</span 5371 > 5372 </div> 5373 <div class="section-entry"> 5374 <div class="session-entry"> 5375 <span class="session-event-type">Technical Session</span 5376 ><span class="type-track-spacer"> · </span 5377 ><span class="program-track" 5378 >Aviation Modeling and Analysis</span 5379 ><br /> 5380 <div class="session-title"> 5381 Airport and Airspace Operations 5382 </div> 5383 <div class="slot-entry"> 5384 <a name="con251" tabindex="-1"></a> 5385 <div class="slot-title-line"> 5386 <span class="slot-title" 5387 >Tactical Minimization of the Environmental Impact of 5388 Holding in the Terminal Airspace and an Associated 5389 Economic Model</span 5390 > 5391 </div> 5392 <div class="slot-authors"> 5393 Aditya Paranjape and Anwesha Basu (Tata Consultancy 5394 Services Ltd) 5395 </div> 5396 <div class="slot-abstract"> 5397 <div> 5398 <a 5399 class="clickable no-decoration" 5400 id="vhsjs_view_100_1707793551_4944856" 5401 onclick="$('#vhsjs_view_100_1707793551_4944856').hide(); 5402 $('#vhsjs_hide_100_1707793551_4944856').show(); 5403 $('#99_1707793551_4944773').slideDown(function() { 5404 if (typeof Masonry === 'function') { 5405 $('.use_masonry').masonry(); 5406 }; 5407 5408 });" 5409 ><i class="fa fa-caret-right"></i> 5410 <span class="hover_link">Abstract</span></a 5411 ><a 5412 class="clickable no-decoration" 5413 id="vhsjs_hide_100_1707793551_4944856" 5414 onclick="$('#99_1707793551_4944773').hide(function() { 5415 if (typeof Masonry === 'function') { 5416 $('.use_masonry').masonry(); 5417 }; 5418 }); 5419 $('#vhsjs_hide_100_1707793551_4944856').hide(); 5420 $('#vhsjs_view_100_1707793551_4944856').show();" 5421 style="display: none" 5422 ><i class="fa fa-caret-down"></i> 5423 <span class="hover_link">Abstract</span></a 5424 > 5425 <div 5426 data-display-control="100_1707793551_4944856" 5427 id="99_1707793551_4944773" 5428 style="display: none" 5429 > 5430 <div class="arrow-slidedown"> 5431 <blockquote> 5432 Minimization of the carbon footprint of aviation 5433 is an active area of interest to the industry 5434 and policy makers alike. Optimization of the 5435 individual flight phases is an important step in 5436 that direction. This paper considers the holding 5437 phase, wherein aircraft hold in the terminal 5438 airspace of airports prior to approach and 5439 landing during times of busy operation or when 5440 the arrival capacity is reduced due to factors 5441 such as bad weather. We propose a tactical 5442 method to allocate landing slots while 5443 minimizing the environmental impact of holds. An 5444 environmentally-driven policy can be perceived 5445 as unfair, particularly by airlines whose 5446 environmentally friendly aircraft which might 5447 need to hold longer than they would under a fair 5448 first-come-first-served policy. To alleviate 5449 this challenge, we propose a number of economic 5450 reward schemes, including one based on a linear 5451 programming problem obtained by applying 5452 complementary slackness to the dual of the 5453 assignment problem. 5454 </blockquote> 5455 </div> 5456 </div> 5457 </div> 5458 </div> 5459 <div class="slot-urls"></div> 5460 <a href="/wsc23papers/049.pdf" target="_blank">pdf</a 5461 ><br /> 5462 </div> 5463 <div class="slot-entry"> 5464 <a name="cea145" tabindex="-1"></a> 5465 <div class="slot-title-line">
5466 <span class="slot-title" 5467 >Use of Variable Sized Entities to Model Airport 5468 Passenger Flow with Pedestrian Dynamics</span 5469 > 5470 </div> 5471 <div class="slot-authors"> 5472 Erich Deines and Tanuj Babele (TransSolutions LLC) and 5473 Gary Gardner (InControl) 5474 </div> 5475 <div class="slot-abstract"> 5476 <div> 5477 <a 5478 class="clickable no-decoration" 5479 id="vhsjs_view_102_1707793551_501582" 5480 onclick="$('#vhsjs_view_102_1707793551_501582').hide(); 5481 $('#vhsjs_hide_102_1707793551_501582').show(); 5482 $('#101_1707793551_5015733').slideDown(function() { 5483 if (typeof Masonry === 'function') { 5484 $('.use_masonry').masonry(); 5485 }; 5486 5487 });" 5488 ><i class="fa fa-caret-right"></i> 5489 <span class="hover_link">Abstract</span></a 5490 ><a 5491 class="clickable no-decoration" 5492 id="vhsjs_hide_102_1707793551_501582" 5493 onclick="$('#101_1707793551_5015733').hide(function() { 5494 if (typeof Masonry === 'function') { 5495 $('.use_masonry').masonry(); 5496 }; 5497 }); 5498 $('#vhsjs_hide_102_1707793551_501582').hide(); 5499 $('#vhsjs_view_102_1707793551_501582').show();" 5500 style="display: none" 5501 ><i class="fa fa-caret-down"></i> 5502 <span class="hover_link">Abstract</span></a 5503 > 5504 <div 5505 data-display-control="102_1707793551_501582" 5506 id="101_1707793551_5015733" 5507 style="display: none" 5508 > 5509 <div class="arrow-slidedown"> 5510 <blockquote> 5511 This paper describes the use of variable-sized 5512 entities within the framework of the InControl 5513 simulation software product Pedestrian Dynamics 5514 to rapidly model passenger flow and congestion 5515 for a series of check-in hall lobby designs for 5516 a US domestic airline terminal. Note that the 5517 airline and airport will remain anonymous for 5518 this presentation due to confidentiality. 5519 </blockquote> 5520 </div> 5521 </div> 5522 </div> 5523 </div> 5524 <div class="slot-urls"></div> 5525 <a href="/wsc23papers/cea145.pdf" target="_blank">pdf</a 5526 ><br /> 5527 </div> 5528 </div> 5529 <div class="session-entry"> 5530 <span class="session-event-type">Technical Session</span 5531 ><span class="type-track-spacer"> · </span 5532 ><span class="program-track" 5533 >Aviation Modeling and Analysis</span 5534 ><br /> 5535 <div class="session-title"> 5536 Machine Learning Applications in Aviation 5537 </div> 5538 <div class="session-chair"> 5539 Chair: John Shortle (George Mason University)<br /> 5540 </div> 5541 <div class="slot-entry"> 5542 <a name="con218" tabindex="-1"></a> 5543 <div class="slot-title-line">
5544 <span class="slot-title" 5545 >Aircraft Line Maintenance Scheduling using Simulation 5546 and Reinforcement Learning</span 5547 > 5548 </div> 5549 <div class="slot-authors"> 5550 Simon Widmer, Syed Shaukat, and Cheng-Lung Wu (UNSW) 5551 </div> 5552 <div class="slot-abstract"> 5553 <div> 5554 <a 5555 class="clickable no-decoration" 5556 id="vhsjs_view_104_1707793551_5245051" 5557 onclick="$('#vhsjs_view_104_1707793551_5245051').hide(); 5558 $('#vhsjs_hide_104_1707793551_5245051').show(); 5559 $('#103_1707793551_5244968').slideDown(function() { 5560 if (typeof Masonry === 'function') { 5561 $('.use_masonry').masonry(); 5562 }; 5563 5564 });" 5565 ><i class="fa fa-caret-right"></i> 5566 <span class="hover_link">Abstract</span></a 5567 ><a 5568 class="clickable no-decoration" 5569 id="vhsjs_hide_104_1707793551_5245051" 5570 onclick="$('#103_1707793551_5244968').hide(function() { 5571 if (typeof Masonry === 'function') { 5572 $('.use_masonry').masonry(); 5573 }; 5574 }); 5575 $('#vhsjs_hide_104_1707793551_5245051').hide(); 5576 $('#vhsjs_view_104_1707793551_5245051').show();" 5577 style="display: none" 5578 ><i class="fa fa-caret-down"></i> 5579 <span class="hover_link">Abstract</span></a 5580 > 5581 <div 5582 data-display-control="104_1707793551_5245051" 5583 id="103_1707793551_5244968" 5584 style="display: none" 5585 > 5586 <div class="arrow-slidedown"> 5587 <blockquote> 5588 This paper presents a reinforcement learning 5589 (RL) algorithm prototype to solve the aircraft 5590 line maintenance scheduling problem. The Line 5591 Maintenance Scheduling Problem (LMSP) is 5592 concerned with scheduling a set of maintenance 5593 tasks during an aircraft's ground time. To 5594 address this problem, we introduce a novel LMSP 5595 method combining a hybrid simulation model and 5596 reinforcement learning to schedule maintenance 5597 tasks at multiple airports. Initially, this 5598 paper briefly reviews the existing literature on 5599 optimization-based and AI-enhanced aircraft 5600 maintenance scheduling. Secondly, the novel 5601 reinforcement learning LMSP method is 5602 introduced, evaluated using industry data, and 5603 compared with optimization-based LMSP solutions. 5604 Our experiments demonstrate that the LMSP method 5605 using reinforcement learning is capable of 5606 identifying near-optimal policies for scheduling 5607 line maintenance jobs when compared to the exact 5608 and heuristics-based methods. The proposed model 5609 provides an excellent foundation for future 5610 studies on AI-enhanced scheduling problems. 5611 </blockquote> 5612 </div> 5613 </div> 5614 </div> 5615 </div> 5616 <div class="slot-urls"></div> 5617 <a href="/wsc23papers/050.pdf" target="_blank">pdf</a 5618 ><br /> 5619 </div> 5620 <div class="slot-entry"> 5621 <a name="con112" tabindex="-1"></a> 5622 <div class="slot-title-line">
5623 <span class="slot-title" 5624 >Neural Networks for GNSS Matrix Attitude 5625 Determination in Aerospace Transportation</span 5626 > 5627 </div> 5628 <div class="slot-authors"> 5629 Raul de Celis, Jose Gonzalez-Barroso, Pablo 5630 Solano-Lopez, and Luis Cadarso (Rey Juan Carlos 5631 University) 5632 </div> 5633 <div class="slot-abstract"> 5634 <div> 5635 <a 5636 class="clickable no-decoration" 5637 id="vhsjs_view_106_1707793551_5268106" 5638 onclick="$('#vhsjs_view_106_1707793551_5268106').hide(); 5639 $('#vhsjs_hide_106_1707793551_5268106').show(); 5640 $('#105_1707793551_5268028').slideDown(function() { 5641 if (typeof Masonry === 'function') { 5642 $('.use_masonry').masonry(); 5643 }; 5644 5645 });" 5646 ><i class="fa fa-caret-right"></i> 5647 <span class="hover_link">Abstract</span></a 5648 ><a 5649 class="clickable no-decoration" 5650 id="vhsjs_hide_106_1707793551_5268106" 5651 onclick="$('#105_1707793551_5268028').hide(function() { 5652 if (typeof Masonry === 'function') { 5653 $('.use_masonry').masonry(); 5654 }; 5655 }); 5656 $('#vhsjs_hide_106_1707793551_5268106').hide(); 5657 $('#vhsjs_view_106_1707793551_5268106').show();" 5658 style="display: none" 5659 ><i class="fa fa-caret-down"></i> 5660 <span class="hover_link">Abstract</span></a 5661 > 5662 <div 5663 data-display-control="106_1707793551_5268106" 5664 id="105_1707793551_5268028" 5665 style="display: none" 5666 > 5667 <div class="arrow-slidedown"> 5668 <blockquote> 5669 Accurate navigation and control of Aerial 5670 Vehicles requires precise estimations of their 5671 position and attitude. Measuring an aircraft's 5672 rotation involves comparing two vectors in 5673 different reference frames, such as inertial and 5674 body axes. Typically, a GNSS sensor-based matrix 5675 with at least three sensors is utilized for this 5676 purpose, taking advantage of the carrier phase 5677 measurements. However, factors such as 5678 multipath, frequency lock loss, cycle slips, and 5679 severe clock drifts can impede accurate integer 5680 ambiguity resolution. To address these 5681 challenges, a new neural network-based technique 5682 has been developed to optimize the management of 5683 large amounts of data and increase carrier phase 5684 ambiguity resolution reliability. By using 5685 carrier phase difference and pseudorange 5686 information, various neural network 5687 configurations can be trained to solve the 5688 ambiguity and estimate the precise attitude of 5689 the GNSS sensor matrix. The provided solution 5690 can be used alone or hybridized with other 5691 attitude sensor such as gyroscope information. 5692 </blockquote> 5693 </div> 5694 </div> 5695 </div> 5696 </div> 5697 <div class="slot-urls"></div> 5698 <a href="/wsc23papers/051.pdf" target="_blank">pdf</a 5699 ><br /> 5700 </div> 5701 </div> 5702 </div> 5703 <div class="centered"> 5704 <div class="top-link"><a href="#top">Return to Top</a></div> 5705 </div> 5706 <hr /> 5707 </div> 5708 <div class="area-section"> 5709 <div class="centered"> 5710 <a name="ptrack112" tabindex="-1"></a> 5711 <div class="section-title">Complex and Resilient Systems</div> 5712 </div> 5713 <div class="centered track-chair">
5714 <span class="track-chair-role" 5715 >Track Coordinator - Complex and Resilient Systems: </span 5716 ><span class="track-chair-names" 5717 >Saurabh Mittal (MITRE Corporation), Claudia Szabo (The 5718 University of Adelaide, University of Adelaide)</span 5719 > 5720 </div> 5721 <div class="section-entry"> 5722 <div class="session-entry"> 5723 <span class="session-event-type">Technical Session</span 5724 ><span class="type-track-spacer"> · </span 5725 ><span class="program-track" 5726 >Complex and Resilient Systems</span 5727 ><br /> 5728 <div class="session-title"> 5729 Cyber Resilience in Complex Systems 5730 </div> 5731 <div class="session-chair"> 5732 Chair: Claudia Szabo (University of Adelaide, The 5733 University of Adelaide)<br /> 5734 </div> 5735 <div class="slot-entry"> 5736 <a name="con102" tabindex="-1"></a> 5737 <div class="slot-title-line"> 5738 <span class="slot-title" 5739 >A Mathematical Theory to Quantify Cyber-Resilience in 5740 IT/OT Networks</span 5741 > 5742 </div> 5743 <div class="slot-authors"> 5744 Ranjan Pal (Massachusetts Institute of Technology), 5745 Rohan Sequeira (University of Southern California), and 5746 Michael Siegel (Massachusetts Institute of Technology) 5747 </div> 5748 <div class="slot-abstract"> 5749 <div> 5750 <a 5751 class="clickable no-decoration" 5752 id="vhsjs_view_108_1707793551_5373607" 5753 onclick="$('#vhsjs_view_108_1707793551_5373607').hide(); 5754 $('#vhsjs_hide_108_1707793551_5373607').show(); 5755 $('#107_1707793551_5373526').slideDown(function() { 5756 if (typeof Masonry === 'function') { 5757 $('.use_masonry').masonry(); 5758 }; 5759 5760 });" 5761 ><i class="fa fa-caret-right"></i> 5762 <span class="hover_link">Abstract</span></a 5763 ><a 5764 class="clickable no-decoration" 5765 id="vhsjs_hide_108_1707793551_5373607" 5766 onclick="$('#107_1707793551_5373526').hide(function() { 5767 if (typeof Masonry === 'function') { 5768 $('.use_masonry').masonry(); 5769 }; 5770 }); 5771 $('#vhsjs_hide_108_1707793551_5373607').hide(); 5772 $('#vhsjs_view_108_1707793551_5373607').show();" 5773 style="display: none" 5774 ><i class="fa fa-caret-down"></i> 5775 <span class="hover_link">Abstract</span></a 5776 > 5777 <div 5778 data-display-control="108_1707793551_5373607" 5779 id="107_1707793551_5373526" 5780 style="display: none" 5781 > 5782 <div class="arrow-slidedown"> 5783 <blockquote> 5784 Modern enterprise infrastructures (EIs) 5785 including those of industrial control systems 5786 (ICSs) are becoming increasingly crucial to 5787 businesses in a wide range of sectors spanning 5788 multiple end-user verticals (e.g., energy, 5789 chemical, manufacturing, biotechnology). These 5790 EIs improve the (real-time) decision support,
5791 productivity, and efficiency of business 5792 processes, but necessarily reliant upon the 5793 cyber-resilience of complex infrastructures for 5794 sustainable business continuity. We are 5795 interested in the long-standing open question in 5796 the cyber-resilience domain: how can managers 5797 formally quantify cyber-resilience for any 5798 complex networked EI (sub-)system in the event 5799 of a cyber-attack affecting its multiple 5800 (inter-dependent) components? We propose a 5801 simulation-backed framework derived from 5802 probabilistic graph theory to answer this 5803 question. We pioneer the derivation and analysis 5804 of a quantifiable, closed-form manager friendly 5805 expression exhibiting the degree of 5806 cyber-resilience (dependent upon individual EI 5807 component functionality quality and the varying 5808 extents of functional dependencies across 5809 networked components) within the (sub-)system 5810 post cyber-attack(s) affecting an EI. 5811 </blockquote> 5812 </div> 5813 </div> 5814 </div> 5815 </div> 5816 <div class="slot-urls"></div> 5817 <a href="/wsc23papers/052.pdf" target="_blank">pdf</a 5818 ><br /> 5819 </div> 5820 <div class="slot-entry"> 5821 <a name="inv155" tabindex="-1"></a> 5822 <div class="slot-title-line"> 5823 <span class="slot-title" 5824 >Trustworthy Artificial Intelligence Framework for 5825 Proactive Detection and Risk Explanation of Cyber 5826 Attacks in Smart Grid</span 5827 > 5828 </div> 5829 <div class="slot-authors"> 5830 Shirajum Munir and Sachin Shetty (Old Dominion 5831 University) 5832 </div> 5833 <div class="slot-abstract"> 5834 <div> 5835 <a 5836 class="clickable no-decoration" 5837 id="vhsjs_view_110_1707793551_5394645" 5838 onclick="$('#vhsjs_view_110_1707793551_5394645').hide(); 5839 $('#vhsjs_hide_110_1707793551_5394645').show(); 5840 $('#109_1707793551_5394566').slideDown(function() { 5841 if (typeof Masonry === 'function') { 5842 $('.use_masonry').masonry(); 5843 }; 5844 5845 });" 5846 ><i class="fa fa-caret-right"></i> 5847 <span class="hover_link">Abstract</span></a 5848 ><a 5849 class="clickable no-decoration" 5850 id="vhsjs_hide_110_1707793551_5394645" 5851 onclick="$('#109_1707793551_5394566').hide(function() { 5852 if (typeof Masonry === 'function') { 5853 $('.use_masonry').masonry(); 5854 }; 5855 }); 5856 $('#vhsjs_hide_110_1707793551_5394645').hide(); 5857 $('#vhsjs_view_110_1707793551_5394645').show();" 5858 style="display: none" 5859 ><i class="fa fa-caret-down"></i> 5860 <span class="hover_link">Abstract</span></a 5861 > 5862 <div
5863 data-display-control="110_1707793551_5394645" 5864 id="109_1707793551_5394566" 5865 style="display: none" 5866 > 5867 <div class="arrow-slidedown"> 5868 <blockquote> 5869 The rapid growth of distributed energy resources 5870 (DERs), such as renewable energy sources, 5871 generators, consumers, and prosumers in the 5872 smart grid infrastructure, poses significant 5873 cybersecurity and trust challenges to the grid 5874 controller. Consequently, it is crucial to 5875 identify adversarial tactics and measure the 5876 strength of the attacker’s DER. To enable 5877 a trustworthy smart grid controller, this work 5878 investigates a trustworthy artificial 5879 intelligence (AI) mechanism for proactive 5880 identification and explanation of the cyber risk 5881 caused by the control/status message of DERs. 5882 Thus, proposing and developing a trustworthy AI 5883 framework to facilitate the deployment of any AI 5884 algorithms for detecting potential cyber threats 5885 and analyzing root causes based on Shapley value 5886 interpretation while dynamically quantifying the 5887 risk of an attack based on Ward’s minimum 5888 variance formula. The experiment with a 5889 state-of-the-art dataset establishes the 5890 proposed framework as a trustworthy AI by 5891 fulfilling the capabilities of reliability, 5892 fairness, explainability, transparency, 5893 reproducibility, and accountability. 5894 </blockquote> 5895 </div> 5896 </div> 5897 </div> 5898 </div> 5899 <div class="slot-urls"></div> 5900 <a href="/wsc23papers/053.pdf" target="_blank">pdf</a 5901 ><br /> 5902 </div> 5903 <div class="slot-entry"> 5904 <a name="inv125" tabindex="-1"></a> 5905 <div class="slot-title-line"> 5906 <span class="slot-title" 5907 >A Mathematical Theory to Price Cyber-Cat Bonds 5908 Boosting IT/OT Security</span 5909 > 5910 </div> 5911 <div class="slot-authors"> 5912 Ranjan Pal (MIT Sloan School of Management) and 5913 Bodhibrata Nag (Indian Institute of Management Calcutta) 5914 </div> 5915 <div class="slot-abstract"> 5916 <div> 5917 <a 5918 class="clickable no-decoration" 5919 id="vhsjs_view_112_1707793551_5416245" 5920 onclick="$('#vhsjs_view_112_1707793551_5416245').hide(); 5921 $('#vhsjs_hide_112_1707793551_5416245').show(); 5922 $('#111_1707793551_541617').slideDown(function() { 5923 if (typeof Masonry === 'function') { 5924 $('.use_masonry').masonry(); 5925 }; 5926 5927 });" 5928 ><i class="fa fa-caret-right"></i> 5929 <span class="hover_link">Abstract</span></a 5930 ><a 5931 class="clickable no-decoration" 5932 id="vhsjs_hide_112_1707793551_5416245" 5933 onclick="$('#111_1707793551_541617').hide(function() { 5934 if (typeof Masonry === 'function') { 5935 $('.use_masonry').masonry(); 5936 }; 5937 }); 5938 $('#vhsjs_hide_112_1707793551_5416245').hide(); 5939 $('#vhsjs_view_112_1707793551_5416245').show();" 5940 style="display: none" 5941 ><i class="fa fa-caret-down"></i>
5942 <span class="hover_link">Abstract</span></a 5943 > 5944 <div 5945 data-display-control="112_1707793551_5416245" 5946 id="111_1707793551_541617" 5947 style="display: none" 5948 > 5949 <div class="arrow-slidedown"> 5950 <blockquote> 5951 The density of enterprise cyber (re-)insurance 5952 markets to manage (aggregate) enterprise 5953 cyber-risk has been low enough to realize their 5954 potential to significantly improve 5955 cyber-security and consequently the 5956 cyber-reliability of (ICS) enterprise 5957 ecosystems. In this paper, we propose the use of 5958 catastrophic (CAT) bonds as a radical and 5959 alternative residual cyber-risk management 5960 methodology to alleviate the big supply demand 5961 gap in the current cyber (re-)insurance 5962 industry, by boosting capital injection in the 5963 latter industry. Two important follow up 5964 questions arise: (i) when is it feasible for 5965 cyber (re-)insurers to invest in CAT bonds? and 5966 (ii) how can we price cyber-CAT bonds 5967 conditioned on the feasibility condition(s)? We 5968 focus on answering the second question pivoted 5969 upon an existential answer to the first. We 5970 propose a novel practically motivated 5971 information asymmetry (IA) driven cyber-CAT bond 5972 pricing model, built upon theories of financial 5973 stochastic processes and Monte Carlo 5974 simulations, in realistic arbitraged incomplete 5975 markets. 5976 </blockquote> 5977 </div> 5978 </div> 5979 </div> 5980 </div> 5981 <div class="slot-urls"></div> 5982 <a href="/wsc23papers/054.pdf" target="_blank">pdf</a 5983 ><br /> 5984 </div> 5985 </div> 5986 <div class="session-entry"> 5987 <span class="session-event-type">Technical Session</span 5988 ><span class="type-track-spacer"> · </span 5989 ><span class="program-track" 5990 >Complex and Resilient Systems</span 5991 ><br /> 5992 <div class="session-title"> 5993 Panel: Resilience and Complexity in Socio-cyber-physical 5994 Systems 5995 </div> 5996 <div class="session-chair"> 5997 Chair: Claudia Szabo (University of Adelaide, The 5998 University of Adelaide)<br /> 5999 </div> 6000 <div class="slot-entry"> 6001 <a name="inv138" tabindex="-1"></a> 6002 <div class="slot-title-line"> 6003 <span class="slot-title" 6004 >Resilience and Complexity in Socio-Cyber-Physical 6005 Systems</span 6006 > 6007 </div> 6008 <div class="slot-authors"> 6009 Claudia Szabo (University of Adelaide), Rodrigo Castro 6010 (CIFASIS-CONICET), Joachim Denil (University of 6011 Antwerp), and Susan M. Sanchez (Naval Postgraduate 6012 School) 6013 </div> 6014 <div class="slot-abstract"> 6015 <div> 6016 <a 6017 class="clickable no-decoration" 6018 id="vhsjs_view_114_1707793551_5469108" 6019 onclick="$('#vhsjs_view_114_1707793551_5469108').hide(); 6020 $('#vhsjs_hide_114_1707793551_5469108').show(); 6021 $('#113_1707793551_546903').slideDown(function() { 6022 if (typeof Masonry === 'function') { 6023 $('.use_masonry').masonry(); 6024 }; 6025 6026 });" 6027 ><i class="fa fa-caret-right"></i>
6028 <span class="hover_link">Abstract</span></a 6029 ><a 6030 class="clickable no-decoration" 6031 id="vhsjs_hide_114_1707793551_5469108" 6032 onclick="$('#113_1707793551_546903').hide(function() { 6033 if (typeof Masonry === 'function') { 6034 $('.use_masonry').masonry(); 6035 }; 6036 }); 6037 $('#vhsjs_hide_114_1707793551_5469108').hide(); 6038 $('#vhsjs_view_114_1707793551_5469108').show();" 6039 style="display: none" 6040 ><i class="fa fa-caret-down"></i> 6041 <span class="hover_link">Abstract</span></a 6042 > 6043 <div 6044 data-display-control="114_1707793551_5469108" 6045 id="113_1707793551_546903" 6046 style="display: none" 6047 > 6048 <div class="arrow-slidedown"> 6049 <blockquote> 6050 Socio-Cyber-Physical Systems are ubiquitous in 6051 today’s world. They are inherently complex 6052 systems built out of many large-scale systems 6053 that encompass different perspectives and 6054 numerous stakeholders. This leads to several 6055 challenges in managing their complexity and 6056 emergent behavior. In addition, these systems 6057 tend to include many adaptive and autonomous 6058 systems with different goals and different 6059 adaptations to environment changes or failures. 6060 The design, analysis, and testing of such 6061 systems is inherently challenging but is 6062 becoming critical due to their wide adoption. In 6063 this panel, we aim to discuss some of these 6064 challenges and potential solutions. 6065 </blockquote> 6066 </div> 6067 </div> 6068 </div> 6069 </div> 6070 <div class="slot-urls"></div> 6071 <a href="/wsc23papers/055.pdf" target="_blank">pdf</a 6072 ><br /> 6073 </div> 6074 </div> 6075 <div class="session-entry"> 6076 <span class="session-event-type">Technical Session</span 6077 ><span class="type-track-spacer"> · </span 6078 ><span class="program-track" 6079 >Complex and Resilient Systems</span 6080 ><br /> 6081 <div class="session-title"> 6082 Panel: Using Simulation to Improve Trust and Autonomy 6083 Adoption 6084 </div> 6085 <div class="session-chair"> 6086 Chair: Kelly Neville (MITRE Corporation)<br /> 6087 </div> 6088 <div class="slot-entry"> 6089 <a name="inv149" tabindex="-1"></a> 6090 <div class="slot-title-line"> 6091 <span class="slot-title" 6092 >The Use of Simulation to Improve Trust and Adoption 6093 of Autonomy and AI in High-Consequence Work 6094 Systems</span 6095 > 6096 </div> 6097 <div class="slot-authors"> 6098 Emily Barrett, Lisa Billman, Theresa Fersch, Valerie 6099 Gawron, and Kelly Neville (MITRE Corporation); Emily 6100 Patterson (The Ohio State University); and Eric Vorm 6101 (Naval Air Warfare Center) 6102 </div> 6103 <div class="slot-abstract"> 6104 <div> 6105 <a 6106 class="clickable no-decoration" 6107 id="vhsjs_view_116_1707793551_5515285" 6108 onclick="$('#vhsjs_view_116_1707793551_5515285').hide(); 6109 $('#vhsjs_hide_116_1707793551_5515285').show(); 6110 $('#115_1707793551_5515203').slideDown(function() { 6111 if (typeof Masonry === 'function') { 6112 $('.use_masonry').masonry(); 6113 }; 6114 6115 });" 6116 ><i class="fa fa-caret-right"></i>
6117 <span class="hover_link">Abstract</span></a 6118 ><a 6119 class="clickable no-decoration" 6120 id="vhsjs_hide_116_1707793551_5515285" 6121 onclick="$('#115_1707793551_5515203').hide(function() { 6122 if (typeof Masonry === 'function') { 6123 $('.use_masonry').masonry(); 6124 }; 6125 }); 6126 $('#vhsjs_hide_116_1707793551_5515285').hide(); 6127 $('#vhsjs_view_116_1707793551_5515285').show();" 6128 style="display: none" 6129 ><i class="fa fa-caret-down"></i> 6130 <span class="hover_link">Abstract</span></a 6131 > 6132 <div 6133 data-display-control="116_1707793551_5515285" 6134 id="115_1707793551_5515203" 6135 style="display: none" 6136 > 6137 <div class="arrow-slidedown"> 6138 <blockquote> 6139 We assert that simulation should be an integral 6140 part of technology development and acquisition. 6141 Its use to iteratively evaluate new technology 6142 across the development timeline can help ensure 6143 technologies contribute to resilience in work 6144 operations. This, in turn, benefits trust and 6145 likelihood of adoption. Potential hindrances to 6146 simulation in technology development are the 6147 time and complexity simulation can introduce. 6148 Time may be needed to model entities and 6149 dynamics to be simulated, plan and conduct 6150 simulation-based tests and experiments, and 6151 translate the results into requirements, user 6152 stories, or other inputs to the 6153 technology’s design and implementation 6154 plan. Complexity is increased when simulation 6155 results suggest new or changed requirements, 6156 identify technology design and implementation 6157 improvements, or produce conflicting feedback 6158 from potential users. We will discuss these 6159 challenges, methods and tools that minimize 6160 their disruptive effects, varieties of 6161 simulation we have used to support technology 6162 development, and benefits of using simulation in 6163 development. 6164 </blockquote> 6165 </div> 6166 </div> 6167 </div> 6168 </div> 6169 <div class="slot-urls"></div> 6170 <a href="/wsc23papers/056.pdf" target="_blank">pdf</a 6171 ><br /> 6172 </div> 6173 </div> 6174 <div class="session-entry"> 6175 <span class="session-event-type">Technical Session</span 6176 ><span class="type-track-spacer"> · </span 6177 ><span class="program-track" 6178 >Complex and Resilient Systems</span 6179 ><br /> 6180 <div class="session-title"> 6181 Resilient Enterprise and Services 6182 </div> 6183 <div class="session-chair"> 6184 Chair: Claudia Szabo (University of Adelaide, The 6185 University of Adelaide)<br /> 6186 </div> 6187 <div class="slot-entry"> 6188 <a name="con346" tabindex="-1"></a> 6189 <div class="slot-title-line">
6190 <span class="slot-title" 6191 >Symbiotic Use of Digital Twin, Simulation and Design 6192 Thinking Approach for Resilient Enterprise</span 6193 > 6194 </div> 6195 <div class="slot-authors"> 6196 Souvik Barat, Sylvan Lobo, Reshma Korabu, Himabindu 6197 Thogaru, and Ravi Mahamuni (Tata Consultancy Services 6198 Research) 6199 </div> 6200 <div class="slot-abstract"> 6201 <div> 6202 <a 6203 class="clickable no-decoration" 6204 id="vhsjs_view_118_1707793551_5579693" 6205 onclick="$('#vhsjs_view_118_1707793551_5579693').hide(); 6206 $('#vhsjs_hide_118_1707793551_5579693').show(); 6207 $('#117_1707793551_5579612').slideDown(function() { 6208 if (typeof Masonry === 'function') { 6209 $('.use_masonry').masonry(); 6210 }; 6211 6212 });" 6213 ><i class="fa fa-caret-right"></i> 6214 <span class="hover_link">Abstract</span></a 6215 ><a 6216 class="clickable no-decoration" 6217 id="vhsjs_hide_118_1707793551_5579693" 6218 onclick="$('#117_1707793551_5579612').hide(function() { 6219 if (typeof Masonry === 'function') { 6220 $('.use_masonry').masonry(); 6221 }; 6222 }); 6223 $('#vhsjs_hide_118_1707793551_5579693').hide(); 6224 $('#vhsjs_view_118_1707793551_5579693').show();" 6225 style="display: none" 6226 ><i class="fa fa-caret-down"></i> 6227 <span class="hover_link">Abstract</span></a 6228 > 6229 <div 6230 data-display-control="118_1707793551_5579693" 6231 id="117_1707793551_5579612" 6232 style="display: none" 6233 > 6234 <div class="arrow-slidedown"> 6235 <blockquote> 6236 Enterprises are increasingly facing the need to 6237 be resilient in the face of uncertainty and 6238 dynamism. Simulatable digital twins have become 6239 critical aids for analyzing and adapting complex 6240 systems. Design thinking and service design 6241 methodologies, in contrast, are gaining momentum 6242 for ideation, subjective evaluation, and 6243 innovation. A systematic application of these 6244 methodologies to explore innovative ideas and a 6245 faithful virtual environment to test and 6246 fine-tune those ideas without impacting real 6247 systems could be transformational. This paper 6248 presents an approach that establishes a 6249 symbiotic relationship between these two 6250 approaches to introduce precision and 6251 innovativeness to make enterprises resilient. We 6252 describe the key characteristics of resilient 6253 enterprises, present our approach, and 6254 illustrate its effectiveness with a case study 6255 focusing on a transformation toward a new normal 6256 to address the Covid-19 pandemic induced 6257 disruptions in the IT industry. 6258 </blockquote> 6259 </div> 6260 </div> 6261 </div> 6262 </div> 6263 <div class="slot-urls"></div> 6264 <a href="/wsc23papers/057.pdf" target="_blank">pdf</a 6265 ><br /> 6266 </div> 6267 <div class="slot-entry"> 6268 <a name="con194" tabindex="-1"></a> 6269 <div class="slot-title-line">
6270 <span class="slot-title" 6271 >Markov Process Simulations of Service Systems with 6272 Concurrent Hawkes Service Interactions</span 6273 > 6274 </div> 6275 <div class="slot-authors"> 6276 Andrew Daw (University of Southern California) and Galit 6277 B. Yom-Tov (Technion - Israel Institute of Technology) 6278 </div> 6279 <div class="slot-abstract"> 6280 <div> 6281 <a 6282 class="clickable no-decoration" 6283 id="vhsjs_view_120_1707793551_5602062" 6284 onclick="$('#vhsjs_view_120_1707793551_5602062').hide(); 6285 $('#vhsjs_hide_120_1707793551_5602062').show(); 6286 $('#119_1707793551_5601976').slideDown(function() { 6287 if (typeof Masonry === 'function') { 6288 $('.use_masonry').masonry(); 6289 }; 6290 6291 });" 6292 ><i class="fa fa-caret-right"></i> 6293 <span class="hover_link">Abstract</span></a 6294 ><a 6295 class="clickable no-decoration" 6296 id="vhsjs_hide_120_1707793551_5602062" 6297 onclick="$('#119_1707793551_5601976').hide(function() { 6298 if (typeof Masonry === 'function') { 6299 $('.use_masonry').masonry(); 6300 }; 6301 }); 6302 $('#vhsjs_hide_120_1707793551_5602062').hide(); 6303 $('#vhsjs_view_120_1707793551_5602062').show();" 6304 style="display: none" 6305 ><i class="fa fa-caret-down"></i> 6306 <span class="hover_link">Abstract</span></a 6307 > 6308 <div 6309 data-display-control="120_1707793551_5602062" 6310 id="119_1707793551_5601976" 6311 style="display: none" 6312 > 6313 <div class="arrow-slidedown"> 6314 <blockquote>
6315 In multi-tasked services such as in 6316 messaging-based contact centers, parallel 6317 service interactions share a mutual dependence 6318 through the agent's concurrency. Here, we 6319 introduce Markov process simulation methods for 6320 bivariate Hawkes cluster service models that are 6321 not Markovian by default due to their 6322 concurrency dependence. To do so, we propose an 6323 alternate construction that maintains extra 6324 "shadow" variables for how the process would be 6325 under other concurrency levels. We prove that 6326 this construction yields an equivalent Markov 6327 process, and we show through numerical 6328 experiments that its corresponding simulation 6329 algorithm is significantly more efficient than 6330 the non-Markovian alternatives. 6331 </blockquote> 6332 </div> 6333 </div> 6334 </div> 6335 </div> 6336 <div class="slot-urls"></div> 6337 <a href="/wsc23papers/058.pdf" target="_blank">pdf</a 6338 ><br /> 6339 </div> 6340 <div class="slot-entry"> 6341 <a name="cea136" tabindex="-1"></a> 6342 <div class="slot-title-line"> 6343 <span class="slot-title" 6344 >Stochastic Climate Simulation for Power Grid Net 6345 Demand Risk Assessment</span 6346 > 6347 </div> 6348 <div class="slot-authors">Rob Cirincione (Sunairio)</div> 6349 <div class="slot-abstract"> 6350 <div> 6351 <a 6352 class="clickable no-decoration" 6353 id="vhsjs_view_122_1707793551_5623624" 6354 onclick="$('#vhsjs_view_122_1707793551_5623624').hide(); 6355 $('#vhsjs_hide_122_1707793551_5623624').show(); 6356 $('#121_1707793551_5623536').slideDown(function() { 6357 if (typeof Masonry === 'function') { 6358 $('.use_masonry').masonry(); 6359 }; 6360 6361 });" 6362 ><i class="fa fa-caret-right"></i> 6363 <span class="hover_link">Abstract</span></a 6364 ><a 6365 class="clickable no-decoration" 6366 id="vhsjs_hide_122_1707793551_5623624" 6367 onclick="$('#121_1707793551_5623536').hide(function() { 6368 if (typeof Masonry === 'function') { 6369 $('.use_masonry').masonry(); 6370 }; 6371 }); 6372 $('#vhsjs_hide_122_1707793551_5623624').hide(); 6373 $('#vhsjs_view_122_1707793551_5623624').show();" 6374 style="display: none" 6375 ><i class="fa fa-caret-down"></i> 6376 <span class="hover_link">Abstract</span></a 6377 > 6378 <div 6379 data-display-control="122_1707793551_5623624" 6380 id="121_1707793551_5623536" 6381 style="display: none" 6382 > 6383 <div class="arrow-slidedown"> 6384 <blockquote> 6385 Power grid planners and power portfolio managers 6386 are increasingly concerned with anticipating 6387 “net demand” risks, which is defined 6388 as customer demand minus renewables for a 6389 particular time period. Net demand is a better 6390 predictor of grid stress than peak demand in a 6391 grid with significant renewables penetration. 6392 For Holy Cross Energy, Sunairio simulated 1,000 6393 probabilistic outcomes of hourly weather across 6394 a geographic region that encompassed the 6395 locations of customers and renewable energy 6396 resources (wind, solar), for 15 years. The 6397 hourly weather simulations were transformed to 6398 hourly energy simulations of customer demand, 6399 wind generation, and solar generation via 6400 machine learning models, creating a broad, 6401 climate-change-aware, coincident data set from 6402 which to quantify concurrent risks to net 6403 demand. Net demand paths of particular interest 6404 for grid planning were curated via statistical 6405 processing. 6406 </blockquote> 6407 </div> 6408 </div> 6409 </div> 6410 </div> 6411 <div class="slot-urls"></div> 6412 <a href="/wsc23papers/cea136.pdf" target="_blank">pdf</a 6413 ><br /> 6414 </div> 6415 </div> 6416 <div class="session-entry">
6417 <span class="session-event-type">Technical Session</span 6418 ><span class="type-track-spacer"> · </span 6419 ><span class="program-track" 6420 >Complex and Resilient Systems</span 6421 ><br /> 6422 <div class="session-title"> 6423 Handling Uncertainty in Complex and Resilient Systems 6424 </div> 6425 <div class="session-chair"> 6426 Chair: Souvik Barat (TCS)<br /> 6427 </div> 6428 <div class="slot-entry"> 6429 <a name="con266" tabindex="-1"></a> 6430 <div class="slot-title-line"> 6431 <span class="slot-title" 6432 >Effects of Timing of Agents' Reactions in 6433 Pharmaceutical Supply Chains under Disruption</span 6434 > 6435 </div> 6436 <div class="slot-authors"> 6437 Rozhin Doroudi, Ozlem Ergun, Jacqueline Griffin, and 6438 Stacy Marsella (Northeastern University) 6439 </div> 6440 <div class="slot-abstract"> 6441 <div> 6442 <a 6443 class="clickable no-decoration" 6444 id="vhsjs_view_124_1707793551_5676293" 6445 onclick="$('#vhsjs_view_124_1707793551_5676293').hide(); 6446 $('#vhsjs_hide_124_1707793551_5676293').show(); 6447 $('#123_1707793551_5676208').slideDown(function() { 6448 if (typeof Masonry === 'function') { 6449 $('.use_masonry').masonry(); 6450 }; 6451 6452 });" 6453 ><i class="fa fa-caret-right"></i> 6454 <span class="hover_link">Abstract</span></a 6455 ><a 6456 class="clickable no-decoration" 6457 id="vhsjs_hide_124_1707793551_5676293" 6458 onclick="$('#123_1707793551_5676208').hide(function() { 6459 if (typeof Masonry === 'function') { 6460 $('.use_masonry').masonry(); 6461 }; 6462 }); 6463 $('#vhsjs_hide_124_1707793551_5676293').hide(); 6464 $('#vhsjs_view_124_1707793551_5676293').show();" 6465 style="display: none" 6466 ><i class="fa fa-caret-down"></i> 6467 <span class="hover_link">Abstract</span></a 6468 > 6469 <div 6470 data-display-control="124_1707793551_5676293" 6471 id="123_1707793551_5676208" 6472 style="display: none" 6473 > 6474 <div class="arrow-slidedown"> 6475 <blockquote> 6476 Disruptions in the supply chain network can have 6477 significant and far-reaching consequences, 6478 especially in pharmaceutical supply chains that 6479 affect health and financial outcomes and raise 6480 equity concerns. To inform strategies that can 6481 address this critical global problem, we study 6482 disruptions in pharmaceutical supply chains 6483 using multiagent simulations. These simulations 6484 include decision-theoretic agents with a theory 6485 of mind reasoning that allows them to reason 6486 about the other agents in the supply chain, 6487 including their trustworthiness. The simulations 6488 reveal how supplier-buyer interactions have 6489 non-local effects which can exacerbate and 6490 extend disruption impacts. In addition, a 6491 distributor’
6491s focus on its own short-term 6492 profit can lower its long-term profit and damage 6493 equity in healthcenters. We also demonstrate how 6494 agents adapt to changes in the environment and 6495 changes in other agents’ behavior and how 6496 in the absence of explicit communication and 6497 coordination, the timing of these adaptations 6498 inhibits disruption mitigation efforts from 6499 transpiring. 6500 </blockquote> 6501 </div> 6502 </div> 6503 </div> 6504 </div> 6505 <div class="slot-urls"></div> 6506 <a href="/wsc23papers/059.pdf" target="_blank">pdf</a 6507 ><br /> 6508 </div> 6509 <div class="slot-entry"> 6510 <a name="con263" tabindex="-1"></a> 6511 <div class="slot-title-line"> 6512 <span class="slot-title" 6513 >Model Predictive Control in Optimal Intervention of 6514 COVID-19 with Mixed Epistemic-Aleatoric 6515 Uncertainty</span 6516 > 6517 </div> 6518 <div class="slot-authors"> 6519 Jinming Wan, Saeideh Mirghorbani, N. Eva Wu, and 6520 Changqing Cheng (Binghamton University) 6521 </div> 6522 <div class="slot-abstract"> 6523 <div> 6524 <a 6525 class="clickable no-decoration" 6526 id="vhsjs_view_126_1707793551_5699885" 6527 onclick="$('#vhsjs_view_126_1707793551_5699885').hide(); 6528 $('#vhsjs_hide_126_1707793551_5699885').show(); 6529 $('#125_1707793551_5699809').slideDown(function() { 6530 if (typeof Masonry === 'function') { 6531 $('.use_masonry').masonry(); 6532 }; 6533 6534 });" 6535 ><i class="fa fa-caret-right"></i> 6536 <span class="hover_link">Abstract</span></a 6537 ><a 6538 class="clickable no-decoration" 6539 id="vhsjs_hide_126_1707793551_5699885" 6540 onclick="$('#125_1707793551_5699809').hide(function() { 6541 if (typeof Masonry === 'function') { 6542 $('.use_masonry').masonry(); 6543 }; 6544 }); 6545 $('#vhsjs_hide_126_1707793551_5699885').hide(); 6546 $('#vhsjs_view_126_1707793551_5699885').show();" 6547 style="display: none" 6548 ><i class="fa fa-caret-down"></i> 6549 <span class="hover_link">Abstract</span></a 6550 > 6551 <div 6552 data-display-control="126_1707793551_5699885" 6553 id="125_1707793551_5699809" 6554 style="display: none" 6555 > 6556 <div class="arrow-slidedown"> 6557 <blockquote> 6558 Non-pharmaceutical interventions (NPI) have been 6559 proven vital in the fight against the COVID-19 6560 pandemic before the massive rollout of 6561 vaccinations. Considering the inherent 6562 epistemic-aleatoric uncertainty of parameters, 6563 accurate simulation and modeling of the 6564 interplay between the NPI and contagion dynamics 6565 are critical to the optimal design of 6566 intervention policies. We propose a modified 6567 SIRD-MPC model that combines a modified 6568 stochastic 6569 Susceptible-Infected-Recovered-Deceased (SIRD) 6570 compartment model with mixed epistemic-aleatoric 6571 parameters and Model Predictive
6571Control (MPC), 6572 to develop robust NPI control policies to 6573 contain the infection of the COVID-19 pandemic 6574 with minimum economic impact. The simulation 6575 result indicates that our proposed model can 6576 significantly decrease the infection rate 6577 compared to the practical results under the same 6578 initial conditions. 6579 </blockquote> 6580 </div> 6581 </div> 6582 </div> 6583 </div> 6584 <div class="slot-urls"></div> 6585 <a href="/wsc23papers/060.pdf" target="_blank">pdf</a 6586 ><br /> 6587 </div> 6588 </div> 6589 <div class="session-entry"> 6590 <span class="session-event-type">Technical Session</span 6591 ><span class="type-track-spacer"> · </span 6592 ><span class="program-track" 6593 >Complex and Resilient Systems</span 6594 ><br /> 6595 <div class="session-title"> 6596 Reliability in Power Systems 6597 </div> 6598 <div class="session-chair"> 6599 Chair: Jinming Wan (Binghamton University)<br /> 6600 </div> 6601 <div class="slot-entry"> 6602 <a name="con290" tabindex="-1"></a> 6603 <div class="slot-title-line"> 6604 <span class="slot-title" 6605 >Cascading Transformer Failure Probability Model Under 6606 Geomagnetic Disturbances</span 6607 > 6608 </div> 6609 <div class="slot-authors"> 6610 Pratishtha Shukla, James Nutaro, and Srikanth Yoginath 6611 (Oak Ridge National Laboratory) 6612 </div> 6613 <div class="slot-abstract"> 6614 <div> 6615 <a 6616 class="clickable no-decoration" 6617 id="vhsjs_view_128_1707793551_576973" 6618 onclick="$('#vhsjs_view_128_1707793551_576973').hide(); 6619 $('#vhsjs_hide_128_1707793551_576973').show(); 6620 $('#127_1707793551_5769649').slideDown(function() { 6621 if (typeof Masonry === 'function') { 6622 $('.use_masonry').masonry(); 6623 }; 6624 6625 });" 6626 ><i class="fa fa-caret-right"></i> 6627 <span class="hover_link">Abstract</span></a 6628 ><a 6629 class="clickable no-decoration" 6630 id="vhsjs_hide_128_1707793551_576973" 6631 onclick="$('#127_1707793551_5769649').hide(function() { 6632 if (typeof Masonry === 'function') { 6633 $('.use_masonry').masonry(); 6634 }; 6635 }); 6636 $('#vhsjs_hide_128_1707793551_576973').hide(); 6637 $('#vhsjs_view_128_1707793551_576973').show();" 6638 style="display: none" 6639 ><i class="fa fa-caret-down"></i> 6640 <span class="hover_link">Abstract</span></a 6641 > 6642 <div 6643 data-display-control="128_1707793551_576973" 6644 id="127_1707793551_5769649" 6645 style="display: none" 6646 > 6647 <div class="arrow-slidedown"> 6648 <blockquote> 6649 This paper develops a probabilistic model to 6650 assess the cascading failure of transformers in 6651 an electric power grid experiencing geomagnetic 6652 disturbances caused by a solar storm. We propose
6653 a model in which the probability of failure is a 6654 function of the intensity of the solar storm, 6655 the physical properties of the transformer, the 6656 geographical location of the transformer, and 6657 the flow of electrical power. We demonstrate the 6658 proposed model using the IEEE 14-bus system and 6659 several notional solar storms. The model quickly 6660 computes the initial and cascading failure 6661 probabilities of the transformers in the system 6662 as a first step towards quantifying the risks 6663 posed by future solar storms. 6664 </blockquote> 6665 </div> 6666 </div> 6667 </div> 6668 </div> 6669 <div class="slot-urls"></div> 6670 <a href="/wsc23papers/061.pdf" target="_blank">pdf</a 6671 ><br /> 6672 </div> 6673 <div class="slot-entry"> 6674 <a name="cea144" tabindex="-1"></a> 6675 <div class="slot-title-line"> 6676 <span class="slot-title" 6677 >Impact of Salt-To-Steam Heat Exchanger Failure Rates 6678 on Lifetime Production of Concentrating Solar Power 6679 Tower Plants</span 6680 > 6681 </div> 6682 <div class="slot-authors"> 6683 Karoline Hood (US Army, Colorado School of Mines) and 6684 Alex Zolan (National Renewable Energy Laboratory) 6685 </div> 6686 <div class="slot-abstract"> 6687 <div> 6688 <a 6689 class="clickable no-decoration" 6690 id="vhsjs_view_130_1707793551_5791068" 6691 onclick="$('#vhsjs_view_130_1707793551_5791068').hide(); 6692 $('#vhsjs_hide_130_1707793551_5791068').show(); 6693 $('#129_1707793551_5790987').slideDown(function() { 6694 if (typeof Masonry === 'function') { 6695 $('.use_masonry').masonry(); 6696 }; 6697 6698 });" 6699 ><i class="fa fa-caret-right"></i> 6700 <span class="hover_link">Abstract</span></a 6701 ><a 6702 class="clickable no-decoration" 6703 id="vhsjs_hide_130_1707793551_5791068" 6704 onclick="$('#129_1707793551_5790987').hide(function() { 6705 if (typeof Masonry === 'function') { 6706 $('.use_masonry').masonry(); 6707 }; 6708 }); 6709 $('#vhsjs_hide_130_1707793551_5791068').hide(); 6710 $('#vhsjs_view_130_1707793551_5791068').show();" 6711 style="display: none" 6712 ><i class="fa fa-caret-down"></i> 6713 <span class="hover_link">Abstract</span></a 6714 > 6715 <div 6716 data-display-control="130_1707793551_5791068" 6717 id="129_1707793551_5790987" 6718 style="display: none" 6719 > 6720 <div class="arrow-slidedown"> 6721 <blockquote> 6722 Heat exchangers in the steam generation system 6723 (SGS) of concentrated solar power (CSP) plants 6724 are unique in their functionality. Consequently, 6725 equipment replacements have long lead times. A 6726 typical CSP plant using an organic Rankine cycle
6727 has one or two salt-to-steam trains (SSTs) 6728 within the SGS. When one heat exchanger in the 6729 SGS fails, the individual SGS fails. We use an 6730 existing framework that combines simulation and 6731 optimization models to assess the impacts of 6732 irrecoverable failures on long-term production. 6733 The methodology provides an optimized dispatch 6734 with the integration of unplanned simulated 6735 failures over a thirty-year period. Our work 6736 shows a system of two trains provides resiliency 6737 and reduces downtime of a plant by six to eight 6738 times compared to a single train. The gross 6739 revenue increases by 31% and 11% for single and 6740 two trains, respectively, when the expected 6741 lifetime increases from five to 10 years. 6742 </blockquote> 6743 </div> 6744 </div> 6745 </div> 6746 </div> 6747 <div class="slot-urls"></div> 6748 <a href="/wsc23papers/cea144.pdf" target="_blank">pdf</a 6749 ><br /> 6750 </div> 6751 </div> 6752 </div> 6753 <div class="centered"> 6754 <div class="top-link"><a href="#top">Return to Top</a></div> 6755 </div> 6756 <hr /> 6757 </div> 6758 <div class="area-section"> 6759 <div class="centered"> 6760 <a name="ptrack104" tabindex="-1"></a> 6761 <div class="section-title">Data Science for Simulation</div> 6762 </div> 6763 <div class="centered track-chair"> 6764 <span class="track-chair-role" 6765 >Track Coordinator - Data Science for Simulation: </span 6766 ><span class="track-chair-names" 6767 >Abdolreza Abhari (Ryerson University), Hamdi Kavak (George 6768 Mason University)</span 6769 > 6770 </div> 6771 <div class="section-entry"> 6772 <div class="session-entry"> 6773 <span class="session-event-type">Technical Session</span 6774 ><span class="type-track-spacer"> · </span 6775 ><span class="program-track" 6776 >Data Science for Simulation</span 6777 ><br /> 6778 <div class="session-title"> 6779 Machine Learning for Simulation 6780 </div> 6781 <div class="session-chair"> 6782 Chair: Hamdi Kavak (George Mason University)<br /> 6783 </div> 6784 <div class="slot-entry"> 6785 <a name="con207" tabindex="-1"></a> 6786 <div class="slot-title-line"> 6787 <span class="slot-title" 6788 >Causal Dynamic Bayesian Networks for Simulation 6789 Metamodeling</span 6790 > 6791 </div> 6792 <div> 6793 <span class="BTP award" 6794 >Best Contributed Theoretical Paper - Finalist</span 6795 > 6796 </div> 6797 <div class="slot-authors"> 6798 Pracheta Boddavaram Amaranath (University of 6799 Massachusetts Amherst), Sam Witty (Basis Research 6800 Institute), and Peter J. Haas and David Jensen 6801 (University of Massachusetts Amherst) 6802 </div> 6803 <div class="slot-abstract"> 6804 <div> 6805 <a 6806 class="clickable no-decoration" 6807 id="vhsjs_view_132_1707793551_651884" 6808 onclick="$('#vhsjs_view_132_1707793551_651884').hide(); 6809 $('#vhsjs_hide_132_1707793551_651884').show(); 6810 $('#131_1707793551_651871').slideDown(function() { 6811 if (typeof Masonry === 'function') { 6812 $('.use_masonry').masonry(); 6813 }; 6814 6815 });" 6816 ><i class="fa fa-caret-right"></i>
6817 <span class="hover_link">Abstract</span></a 6818 ><a 6819 class="clickable no-decoration" 6820 id="vhsjs_hide_132_1707793551_651884" 6821 onclick="$('#131_1707793551_651871').hide(function() { 6822 if (typeof Masonry === 'function') { 6823 $('.use_masonry').masonry(); 6824 }; 6825 }); 6826 $('#vhsjs_hide_132_1707793551_651884').hide(); 6827 $('#vhsjs_view_132_1707793551_651884').show();" 6828 style="display: none" 6829 ><i class="fa fa-caret-down"></i> 6830 <span class="hover_link">Abstract</span></a 6831 > 6832 <div 6833 data-display-control="132_1707793551_651884" 6834 id="131_1707793551_651871" 6835 style="display: none" 6836 > 6837 <div class="arrow-slidedown"> 6838 <blockquote> 6839 A traditional metamodel for a discrete-event 6840 simulation approximates a real-valued 6841 performance measure as a function of the 6842 input-parameter values. We introduce a novel 6843 class of metamodels based on modular dynamic 6844 Bayesian networks (MDBNs), a subclass of 6845 probabilistic graphical models which can be used 6846 to efficiently answer a rich class of 6847 probabilistic and causal queries (PCQs). Such 6848 queries represent the joint probability 6849 distribution of the system state at multiple 6850 time points, given observations of, and 6851 interventions on, other state variables and 6852 input parameters. This paper is a first 6853 demonstration of how the extensive theory and 6854 technology of causal graphical models can be 6855 used to enhance simulation metamodeling. We 6856 demonstrate this potential by showing how a 6857 single MDBN for an M/M/1 queue can be learned 6858 from simulation data and then be used to quickly 6859 and accurately answer a variety of PCQs, most of 6860 which are out-of-scope for existing metamodels. 6861 </blockquote> 6862 </div> 6863 </div> 6864 </div> 6865 </div> 6866 <div class="slot-urls"></div> 6867 <a href="/wsc23papers/062.pdf" target="_blank">pdf</a 6868 ><br /> 6869 </div> 6870 <div class="slot-entry"> 6871 <a name="con301" tabindex="-1"></a> 6872 <div class="slot-title-line"> 6873 <span class="slot-title" 6874 >Deep-learning-assisted Cardiac Electrophysiology 6875 Simulation</span 6876 > 6877 </div> 6878 <div class="slot-authors"> 6879 Weixuan Dong, Yifu Li, and Rui Zhu (The University of 6880 Oklahoma) 6881 </div> 6882 <div class="slot-abstract"> 6883 <div> 6884 <a 6885 class="clickable no-decoration" 6886 id="vhsjs_view_134_1707793551_6543565" 6887 onclick="$('#vhsjs_view_134_1707793551_6543565').hide(); 6888 $('#vhsjs_hide_134_1707793551_6543565').show(); 6889 $('#133_1707793551_6543484').slideDown(function() { 6890 if (typeof Masonry === 'function') { 6891 $('.use_masonry').masonry(); 6892 }; 6893 6894 });" 6895 ><i class="fa fa-caret-right"></i>
6896 <span class="hover_link">Abstract</span></a 6897 ><a 6898 class="clickable no-decoration" 6899 id="vhsjs_hide_134_1707793551_6543565" 6900 onclick="$('#133_1707793551_6543484').hide(function() { 6901 if (typeof Masonry === 'function') { 6902 $('.use_masonry').masonry(); 6903 }; 6904 }); 6905 $('#vhsjs_hide_134_1707793551_6543565').hide(); 6906 $('#vhsjs_view_134_1707793551_6543565').show();" 6907 style="display: none" 6908 ><i class="fa fa-caret-down"></i> 6909 <span class="hover_link">Abstract</span></a 6910 > 6911 <div 6912 data-display-control="134_1707793551_6543565" 6913 id="133_1707793551_6543484" 6914 style="display: none" 6915 > 6916 <div class="arrow-slidedown"> 6917 <blockquote> 6918 Simulation built upon partial and ordinary 6919 differential equations has been a classic 6920 approach to modeling cardiac 6921 electrophysiological dynamics. However, 6922 mitigating the computational burden of 6923 differential equations is still a challenging 6924 problem. This paper provides a novel alternative 6925 utilizing data-driven recurrent neural networks 6926 for cardiac electrophysiological dynamic 6927 simulation. Specifically, we develop a long 6928 short-term memory (LSTM)-assisted simulation to 6929 capture the underlying dynamics of cardiac 6930 electrophysiology while preserving computational 6931 efficiency. Experimental results demonstrate the 6932 efficiency and effectiveness of the proposed 6933 method, which outperforms the differential 6934 equation-based simulation approach while 6935 significantly reducing the computational cost. 6936 The proposed method offers a promising 6937 alternative to traditional simulation and may 6938 contribute to the development of more efficient 6939 and accurate approaches for simulating cardiac 6940 electrophysiology. 6941 </blockquote> 6942 </div> 6943 </div> 6944 </div> 6945 </div> 6946 <div class="slot-urls"></div> 6947 <a href="/wsc23papers/063.pdf" target="_blank">pdf</a 6948 ><br /> 6949 </div> 6950 <div class="slot-entry"> 6951 <a name="con360" tabindex="-1"></a> 6952 <div class="slot-title-line"> 6953 <span class="slot-title" 6954 >Inferring Epidemic Dynamics Using Gaussian Process 6955 Emulation of Agent-Based Simulations</span 6956 > 6957 </div> 6958 <div class="slot-authors"> 6959 Abdulrahman Ahmed, M. Amin Rahimian, and Mark Roberts 6960 (University of Pittsburgh) 6961 </div> 6962 <div class="slot-abstract"> 6963 <div> 6964 <a 6965 class="clickable no-decoration" 6966 id="vhsjs_view_136_1707793551_6566803" 6967 onclick="$('#vhsjs_view_136_1707793551_6566803').hide(); 6968 $('#vhsjs_hide_136_1707793551_6566803').show(); 6969 $('#135_1707793551_6566727').slideDown(function() { 6970 if (typeof Masonry === 'function') { 6971 $('.use_masonry').masonry(); 6972 }; 6973 6974 });" 6975 ><i class="fa fa-caret-right"></i>
6976 <span class="hover_link">Abstract</span></a 6977 ><a 6978 class="clickable no-decoration" 6979 id="vhsjs_hide_136_1707793551_6566803" 6980 onclick="$('#135_1707793551_6566727').hide(function() { 6981 if (typeof Masonry === 'function') { 6982 $('.use_masonry').masonry(); 6983 }; 6984 }); 6985 $('#vhsjs_hide_136_1707793551_6566803').hide(); 6986 $('#vhsjs_view_136_1707793551_6566803').show();" 6987 style="display: none" 6988 ><i class="fa fa-caret-down"></i> 6989 <span class="hover_link">Abstract</span></a 6990 > 6991 <div 6992 data-display-control="136_1707793551_6566803" 6993 id="135_1707793551_6566727" 6994 style="display: none" 6995 > 6996 <div class="arrow-slidedown"> 6997 <blockquote> 6998 Computational models help decision makers 6999 understand epidemic dynamics to optimize public 7000 health interventions. Agent-based simulation of 7001 disease spread in synthetic populations allows 7002 us to compare and contrast different effects 7003 across identical populations or to investigate 7004 the effect of interventions keeping every other 7005 factor constant between "digital twins." FRED (A 7006 Framework for Reconstructing Epidemiological 7007 Dynamics) is an agent-based modeling system with 7008 a geo-spatial perspective using a synthetic 7009 population that is constructed based on the U.S. 7010 Census data. In this paper, we show how Gaussian 7011 process regression can be used on 7012 FRED-synthesized data to infer the differing 7013 spatial dispersion of the epidemic dynamics for 7014 two disease conditions that start from the same 7015 initial conditions and spread among identical 7016 populations. Our results showcase the utility of 7017 agent-based simulation frameworks such as FRED 7018 for inferring differences between conditions 7019 where controlling for all confounding factors 7020 for such comparisons is next to impossible 7021 without synthetic data. 7022 </blockquote> 7023 </div> 7024 </div> 7025 </div> 7026 </div> 7027 <div class="slot-urls"></div> 7028 <a href="/wsc23papers/064.pdf" target="_blank">pdf</a 7029 ><br /> 7030 </div> 7031 </div> 7032 <div class="session-entry"> 7033 <span class="session-event-type">Technical Session</span 7034 ><span class="type-track-spacer"> · </span 7035 ><span class="program-track" 7036 >Data Science for Simulation</span 7037 ><br /> 7038 <div class="session-title"> 7039 Data Analytics for Simulation 7040 </div> 7041 <div class="session-chair"> 7042 Chair: Abdolreza Abhari (Toronto Metropolitan 7043 University)<br /> 7044 </div> 7045 <div class="slot-entry"> 7046 <a name="con254" tabindex="-1"></a> 7047 <div class="slot-title-line">
7048 <span class="slot-title" 7049 >Autonomic Orchestration of In-Situ and In-Transit 7050 Data Analytics for Simulation Studies</span 7051 > 7052 </div> 7053 <div class="slot-authors"> 7054 Xiaorui Du (Technical University of Munich); Adriano 7055 Pimpini (Sapienza, University of Rome); Andrea Piccione 7056 (Huawei Munich Research Center); Zhuoxiao Meng and 7057 Anibal Siguenza-Torres (Technical University of Munich); 7058 Stefano Bortoli (Huawei Munich Research Center); Alois 7059 Knoll (Technical University of Munich); and Alessandro 7060 Pellegrini (University of Rome Tor Vergata) 7061 </div> 7062 <div class="slot-abstract"> 7063 <div> 7064 <a 7065 class="clickable no-decoration" 7066 id="vhsjs_view_138_1707793551_663725" 7067 onclick="$('#vhsjs_view_138_1707793551_663725').hide(); 7068 $('#vhsjs_hide_138_1707793551_663725').show(); 7069 $('#137_1707793551_6637168').slideDown(function() { 7070 if (typeof Masonry === 'function') { 7071 $('.use_masonry').masonry(); 7072 }; 7073 7074 });" 7075 ><i class="fa fa-caret-right"></i> 7076 <span class="hover_link">Abstract</span></a 7077 ><a 7078 class="clickable no-decoration" 7079 id="vhsjs_hide_138_1707793551_663725" 7080 onclick="$('#137_1707793551_6637168').hide(function() { 7081 if (typeof Masonry === 'function') { 7082 $('.use_masonry').masonry(); 7083 }; 7084 }); 7085 $('#vhsjs_hide_138_1707793551_663725').hide(); 7086 $('#vhsjs_view_138_1707793551_663725').show();" 7087 style="display: none" 7088 ><i class="fa fa-caret-down"></i> 7089 <span class="hover_link">Abstract</span></a 7090 > 7091 <div 7092 data-display-control="138_1707793551_663725" 7093 id="137_1707793551_6637168" 7094 style="display: none" 7095 > 7096 <div class="arrow-slidedown"> 7097 <blockquote> 7098 Modern parallel/distributed simulations can 7099 produce large amounts of data. The historical 7100 approach of performing analyses at the end of 7101 the simulation is unlikely to cope with modern, 7102 extremely large-scale analytics jobs. Indeed, 7103 the I/O subsystem can quickly become the global 7104 bottleneck. Similarly, processing on-the-fly the 7105 data produced by simulations can significantly 7106 impair the performance in terms of computational 7107 capacity and network load. We present a 7108 methodology and reference architecture for 7109 constructing an autonomic control system to 7110 determine at runtime the best placement for data 7111 processing (on simulation nodes or a set of 7112 external nodes). This allows for a good tradeoff 7113 between the load on the simulation's critical 7114 path and the data communication system. Our 7115 preliminary experimentation shows that autonomic 7116 orchestration is crucial to improve the global 7117 performance of a data analysis system, 7118 especially when the simulation node's rate of 7119 data production varies during simulation. 7120 </blockquote> 7121 </div> 7122 </div> 7123 </div> 7124 </div> 7125 <div class="slot-urls"></div> 7126 <a href="/wsc23papers/065.pdf" target="_blank">pdf</a 7127 ><br /> 7128 </div> 7129 <div class="slot-entry"> 7130 <a name="cea104" tabindex="-1"></a> 7131 <div class="slot-title-line">
7132 <span class="slot-title" 7133 >Scaling Cross-Relations with Larger Dataset</span 7134 > 7135 </div> 7136 <div class="slot-authors"> 7137 Victor Diakov (Simfoni Ltd.) 7138 </div> 7139 <div class="slot-abstract"> 7140 <div> 7141 <a 7142 class="clickable no-decoration" 7143 id="vhsjs_view_140_1707793551_6658053" 7144 onclick="$('#vhsjs_view_140_1707793551_6658053').hide(); 7145 $('#vhsjs_hide_140_1707793551_6658053').show(); 7146 $('#139_1707793551_665798').slideDown(function() { 7147 if (typeof Masonry === 'function') { 7148 $('.use_masonry').masonry(); 7149 }; 7150 7151 });" 7152 ><i class="fa fa-caret-right"></i> 7153 <span class="hover_link">Abstract</span></a 7154 ><a 7155 class="clickable no-decoration" 7156 id="vhsjs_hide_140_1707793551_6658053" 7157 onclick="$('#139_1707793551_665798').hide(function() { 7158 if (typeof Masonry === 'function') { 7159 $('.use_masonry').masonry(); 7160 }; 7161 }); 7162 $('#vhsjs_hide_140_1707793551_6658053').hide(); 7163 $('#vhsjs_view_140_1707793551_6658053').show();" 7164 style="display: none" 7165 ><i class="fa fa-caret-down"></i> 7166 <span class="hover_link">Abstract</span></a 7167 > 7168 <div 7169 data-display-control="140_1707793551_6658053" 7170 id="139_1707793551_665798" 7171 style="display: none" 7172 > 7173 <div class="arrow-slidedown"> 7174 <blockquote> 7175 Simulation and optimization of procurements 7176 might employ clustering dataset elements to 7177 exclude possible duplicates and improve 7178 processing resiliency. This study presents a 7179 case of applying scaling methods to reduce 7180 computation time of clustering between a smaller 7181 and a larger dataset. In this example (of 7182 selecting close supplier names), computation 7183 time scales as square of N (the number of 7184 elements), and the presented approach in effect 7185 brings computing time to be linear in N. As a 7186 result, computation time in our case is reduced 7187 by over an order of magnitude. 7188 </blockquote> 7189 </div> 7190 </div> 7191 </div> 7192 </div> 7193 <div class="slot-urls"></div> 7194 <a href="/wsc23papers/cea104.pdf" target="_blank">pdf</a 7195 ><br /> 7196 </div> 7197 <div class="slot-entry"> 7198 <a name="con361" tabindex="-1"></a> 7199 <div class="slot-title-line"> 7200 <span class="slot-title" 7201 >Uncovering Competitor Pricing Patterns in the Danish 7202 Pharmaceutical Market via Subsequence Time Series 7203 Clustering: A Case Study</span 7204 > 7205 </div> 7206 <div class="slot-authors"> 7207 Ruhollah Jamali (University of Southern Denmark) and 7208 Sanja Lazarova-Molnar (Karlsruhe Institute of 7209 Technology) 7210 </div> 7211 <div class="slot-abstract"> 7212 <div> 7213 <a 7214 class="clickable no-decoration" 7215 id="vhsjs_view_142_1707793551_6681175" 7216 onclick="$('#vhsjs_view_142_1707793551_6681175').hide(); 7217 $('#vhsjs_hide_142_1707793551_6681175').show(); 7218 $('#141_1707793551_6681097').slideDown(function() { 7219 if (typeof Masonry === 'function') { 7220 $('.use_masonry').masonry(); 7221 }; 7222 7223 });" 7224 ><i class="fa fa-caret-right"></i>
7225 <span class="hover_link">Abstract</span></a 7226 ><a 7227 class="clickable no-decoration" 7228 id="vhsjs_hide_142_1707793551_6681175" 7229 onclick="$('#141_1707793551_6681097').hide(function() { 7230 if (typeof Masonry === 'function') { 7231 $('.use_masonry').masonry(); 7232 }; 7233 }); 7234 $('#vhsjs_hide_142_1707793551_6681175').hide(); 7235 $('#vhsjs_view_142_1707793551_6681175').show();" 7236 style="display: none" 7237 ><i class="fa fa-caret-down"></i> 7238 <span class="hover_link">Abstract</span></a 7239 > 7240 <div 7241 data-display-control="142_1707793551_6681175" 7242 id="141_1707793551_6681097" 7243 style="display: none" 7244 > 7245 <div class="arrow-slidedown"> 7246 <blockquote> 7247 Adopting data-driven decision-making approaches 7248 can significantly enhance profitability and 7249 foster growth in economic situations through 7250 quantitative analysis of market dynamics. One 7251 intriguing market that warrants examination is 7252 the price competition observed within the Danish 7253 pharmaceutical sector, where numerous companies 7254 are vying for a larger market share through the 7255 offering of diverse pharmaceutical products. 7256 This paper aims to shed light on this market by 7257 employing subsequence time series clustering 7258 techniques to identify pricing patterns among 7259 the players involved in the Danish 7260 pharmaceutical industry. The data analysis 7261 pipeline performed in this study allows for the 7262 identification of price patterns for clustering 7263 and discovering different agent groups, as well 7264 as providing a foundation for expanding the 7265 current agent-based model of the European 7266 pharmaceutical parallel trade market by 7267 analyzing the pricing behavior and patterns of 7268 players, facilitating the utilization of 7269 historical data to model agent behavior and 7270 advancing research in this area. 7271 </blockquote> 7272 </div> 7273 </div> 7274 </div> 7275 </div> 7276 <div class="slot-urls"></div> 7277 <a href="/wsc23papers/066.pdf" target="_blank">pdf</a 7278 ><br /> 7279 </div> 7280 </div> 7281 <div class="session-entry"> 7282 <span class="session-event-type">Technical Session</span 7283 ><span class="type-track-spacer"> · </span 7284 ><span class="program-track" 7285 >Data Science for Simulation</span 7286 ><br /> 7287 <div class="session-title">Simulation in Action</div> 7288 <div class="session-chair"> 7289 Chair: Hamdi Kavak (George Mason University)<br /> 7290 </div> 7291 <div class="slot-entry"> 7292 <a name="con314" tabindex="-1"></a> 7293 <div class="slot-title-line">
7294 <span class="slot-title" 7295 >A Preliminary Study of Regularization Framework for 7296 Constructing Task-Specific Simulators</span 7297 > 7298 </div> 7299 <div class="slot-authors"> 7300 Dilara Aykanat (University of California, Berkeley); 7301 Nian Si (The University of Chicago); and Zeyu Zheng 7302 (University of California, Berkeley) 7303 </div> 7304 <div class="slot-abstract"> 7305 <div> 7306 <a 7307 class="clickable no-decoration" 7308 id="vhsjs_view_144_1707793551_67298" 7309 onclick="$('#vhsjs_view_144_1707793551_67298').hide(); 7310 $('#vhsjs_hide_144_1707793551_67298').show(); 7311 $('#143_1707793551_6729717').slideDown(function() { 7312 if (typeof Masonry === 'function') { 7313 $('.use_masonry').masonry(); 7314 }; 7315 7316 });" 7317 ><i class="fa fa-caret-right"></i> 7318 <span class="hover_link">Abstract</span></a 7319 ><a 7320 class="clickable no-decoration" 7321 id="vhsjs_hide_144_1707793551_67298" 7322 onclick="$('#143_1707793551_6729717').hide(function() { 7323 if (typeof Masonry === 'function') { 7324 $('.use_masonry').masonry(); 7325 }; 7326 }); 7327 $('#vhsjs_hide_144_1707793551_67298').hide(); 7328 $('#vhsjs_view_144_1707793551_67298').show();" 7329 style="display: none" 7330 ><i class="fa fa-caret-down"></i> 7331 <span class="hover_link">Abstract</span></a 7332 > 7333 <div 7334 data-display-control="144_1707793551_67298" 7335 id="143_1707793551_6729717" 7336 style="display: none" 7337 > 7338 <div class="arrow-slidedown"> 7339 <blockquote> 7340 One approach to construct or calibrate 7341 simulators, when representative real data exist, 7342 is to ensure that the synthetic data generated 7343 by the simulated match the empirical 7344 distribution of the real data. However, such 7345 approach to construct simulators does not take 7346 into consideration where the constructed 7347 simulators will be used. For some applications, 7348 there are clear tasks (such as performance 7349 evaluation of different decisions) in 7350 users’ mind where the simulated data will 7351 serve as input to the tasks. In this work, we 7352 propose an approach to use the knowledge of 7353 these tasks to guide the construction of 7354 simulators, in addition to the distribution 7355 match of simulated data and real data by 7356 regularizing the objective function with a task 7357 related penalty. We conduct a preliminary 7358 numerical study of this approach to illustrate 7359 the effectiveness compared to not taking into 7360 consideration the specific tasks of the 7361 simulators. 7362 </blockquote> 7363 </div> 7364 </div> 7365 </div> 7366 </div> 7367 <div class="slot-urls"></div> 7368 <a href="/wsc23papers/067.pdf" target="_blank">pdf</a 7369 ><br /> 7370 </div> 7371 <div class="slot-entry"> 7372 <a name="con316" tabindex="-1"></a> 7373 <div class="slot-title-line">
7374 <span class="slot-title" 7375 >Using Simulation to Assess the Reliability of 7376 Forecasts in High-tech Industry</span 7377 > 7378 </div> 7379 <div class="slot-authors"> 7380 Bhoomica Mysore Nataraja (Eindhoven University of 7381 Technology); Tanmay Aggarwal (Lambda Function Inc); and 7382 Nitish Singh, Koen Herps, and Ivo Adan (Eindhoven 7383 University of Technology) 7384 </div> 7385 <div class="slot-abstract"> 7386 <div> 7387 <a 7388 class="clickable no-decoration" 7389 id="vhsjs_view_146_1707793551_6755276" 7390 onclick="$('#vhsjs_view_146_1707793551_6755276').hide(); 7391 $('#vhsjs_hide_146_1707793551_6755276').show(); 7392 $('#145_1707793551_6755195').slideDown(function() { 7393 if (typeof Masonry === 'function') { 7394 $('.use_masonry').masonry(); 7395 }; 7396 7397 });" 7398 ><i class="fa fa-caret-right"></i> 7399 <span class="hover_link">Abstract</span></a 7400 ><a 7401 class="clickable no-decoration" 7402 id="vhsjs_hide_146_1707793551_6755276" 7403 onclick="$('#145_1707793551_6755195').hide(function() { 7404 if (typeof Masonry === 'function') { 7405 $('.use_masonry').masonry(); 7406 }; 7407 }); 7408 $('#vhsjs_hide_146_1707793551_6755276').hide(); 7409 $('#vhsjs_view_146_1707793551_6755276').show();" 7410 style="display: none" 7411 ><i class="fa fa-caret-down"></i> 7412 <span class="hover_link">Abstract</span></a 7413 > 7414 <div 7415 data-display-control="146_1707793551_6755276" 7416 id="145_1707793551_6755195" 7417 style="display: none" 7418 > 7419 <div class="arrow-slidedown"> 7420 <blockquote> 7421 In a high-tech production environment, capacity 7422 investment and production planning are often 7423 based on the demand information from 7424 manufacturers within a supply chain. A supplier 7425 solicits forecast information from a 7426 manufacturer, and the manufacturer provides 7427 demand forecasts that are updated on a rolling 7428 horizon basis. Problems arise with this setup if 7429 the manufacturer provides volatile forecast 7430 quantities due to the market's fluctuating 7431 demand or internal bias. As a result, suppliers' 7432 mistrust regarding forecast quantities grows, 7433 leading to adjusted production plans based on 7434 planners' anecdotal experience. The paper 7435 presents a decision model to determine the 7436 reliability of forecasts provided by 7437 manufacturers to facilitate better production 7438 planning. The study also suggests alternate 7439 forecasting techniques in case of low 7440 reliability. To evaluate the effectiveness of 7441 the proposed approach, a simulation study is 7442 conducted for different manufacturers and 7443 scenarios. Our experiments showed an average 7444 cost reduction of 14% across all instances. 7445 </blockquote> 7446 </div> 7447 </div> 7448 </div> 7449 </div> 7450 <div class="slot-urls"></div> 7451 <a href="/wsc23papers/068.pdf" target="_blank">pdf</a 7452 ><br /> 7453 </div> 7454 <div class="slot-entry"> 7455 <a name="con347" tabindex="-1"></a> 7456 <div class="slot-title-line">
7457 <span class="slot-title" 7458 >Digital Twin Based Learning Framework for Adaptive 7459 Fault Diagnosis in Microgrids with Autonomous 7460 Reconfiguration Capabilities</span 7461 > 7462 </div> 7463 <div class="slot-authors"> 7464 Temitope Runsewe, Abdurrahman Yavuz, and Nurcin Celik 7465 (University of Miami) 7466 </div> 7467 <div class="slot-abstract"> 7468 <div> 7469 <a 7470 class="clickable no-decoration" 7471 id="vhsjs_view_148_1707793551_6778123" 7472 onclick="$('#vhsjs_view_148_1707793551_6778123').hide(); 7473 $('#vhsjs_hide_148_1707793551_6778123').show(); 7474 $('#147_1707793551_6778042').slideDown(function() { 7475 if (typeof Masonry === 'function') { 7476 $('.use_masonry').masonry(); 7477 }; 7478 7479 });" 7480 ><i class="fa fa-caret-right"></i> 7481 <span class="hover_link">Abstract</span></a 7482 ><a 7483 class="clickable no-decoration" 7484 id="vhsjs_hide_148_1707793551_6778123" 7485 onclick="$('#147_1707793551_6778042').hide(function() { 7486 if (typeof Masonry === 'function') { 7487 $('.use_masonry').masonry(); 7488 }; 7489 }); 7490 $('#vhsjs_hide_148_1707793551_6778123').hide(); 7491 $('#vhsjs_view_148_1707793551_6778123').show();" 7492 style="display: none" 7493 ><i class="fa fa-caret-down"></i> 7494 <span class="hover_link">Abstract</span></a 7495 > 7496 <div 7497 data-display-control="148_1707793551_6778123" 7498 id="147_1707793551_6778042" 7499 style="display: none" 7500 > 7501 <div class="arrow-slidedown"> 7502 <blockquote> 7503 The world is increasingly reliant on energy 7504 systems, making them a critical infrastructure 7505 for essential services. This also makes them 7506 vulnerable to attacks, which can result in 7507 significant disruptions and damage. Microgrid 7508 (MG) monitoring systems play a crucial role in 7509 ensuring the safety and reliability of energy 7510 systems. However, traditional fault diagnosis 7511 techniques are limited to already established 7512 faults due to the use of only historical data, 7513 making it challenging to keep up with the 7514 increasing demand for safety and reliability. 7515 This paper proposes a digital twin based machine 7516 learning (DTML) framework for fault diagnosis in 7517 MG monitoring systems, with a focus on assessing 7518 the resilience of MG end-to-end systems to 7519 potential disruptions from adversaries. The 7520 proposed framework utilizes digital twin based 7521 random forest (RF) and support vector machine 7522 (SVM) and logistic regression (LR) model and 7523 shows that the RF based model outperforms other 7524 models with an accuracy of 95%. 7525 </blockquote> 7526 </div> 7527 </div> 7528 </div> 7529 </div> 7530 <div class="slot-urls"></div> 7531 <a href="/wsc23papers/069.pdf" target="_blank">pdf</a 7532 ><br /> 7533 </div> 7534 </div> 7535 </div> 7536 <div class="centered"> 7537 <div class="top-link"><a href="#top">Return to Top</a></div> 7538 </div> 7539 <hr /> 7540 </div> 7541 <div class="area-section"> 7542 <div class="centered"> 7543 <a name="ptrack129" tabindex="-1"></a> 7544 <div class="section-title"> 7545 Environment Sustainability and Resilience 7546 </div> 7547 </div> 7548 <div class="section-entry"> 7549 <div class="session-entry">
7550 <span class="session-event-type">Technical Session</span 7551 ><span class="type-track-spacer"> · </span 7552 ><span class="program-track" 7553 >Environment Sustainability and Resilience</span 7554 ><br /> 7555 <div class="session-title">Critical Infrastructures</div> 7556 <div class="session-chair"> 7557 Chair: Raymond Smith (East Carolina University)<br /> 7558 </div> 7559 <div class="slot-entry"> 7560 <a name="con101" tabindex="-1"></a> 7561 <div class="slot-title-line"> 7562 <span class="slot-title" 7563 >A Network Theory to Quantify and Bound Cyber-risk in 7564 IT/OT Systems</span 7565 > 7566 </div> 7567 <div> 7568 <span class="BAP award" 7569 >Best Contributed Applied Paper - Finalist</span 7570 > 7571 </div> 7572 <div class="slot-authors"> 7573 Ranjan Pal (MIT Sloan School of Management), Rohan 7574 Xavier Sequeira (University of Southern California), and 7575 Sander Zeijlemaker and Michael Siegel (MIT Sloan School 7576 of Management) 7577 </div> 7578 <div class="slot-abstract"> 7579 <div> 7580 <a 7581 class="clickable no-decoration" 7582 id="vhsjs_view_150_1707793551_6850877" 7583 onclick="$('#vhsjs_view_150_1707793551_6850877').hide(); 7584 $('#vhsjs_hide_150_1707793551_6850877').show(); 7585 $('#149_1707793551_6850796').slideDown(function() { 7586 if (typeof Masonry === 'function') { 7587 $('.use_masonry').masonry(); 7588 }; 7589 7590 });" 7591 ><i class="fa fa-caret-right"></i> 7592 <span class="hover_link">Abstract</span></a 7593 ><a 7594 class="clickable no-decoration" 7595 id="vhsjs_hide_150_1707793551_6850877" 7596 onclick="$('#149_1707793551_6850796').hide(function() { 7597 if (typeof Masonry === 'function') { 7598 $('.use_masonry').masonry(); 7599 }; 7600 }); 7601 $('#vhsjs_hide_150_1707793551_6850877').hide(); 7602 $('#vhsjs_view_150_1707793551_6850877').show();" 7603 style="display: none" 7604 ><i class="fa fa-caret-down"></i> 7605 <span class="hover_link">Abstract</span></a 7606 > 7607 <div 7608 data-display-control="150_1707793551_6850877" 7609 id="149_1707793551_6850796" 7610 style="display: none" 7611 > 7612 <div class="arrow-slidedown"> 7613 <blockquote>
7614 IT/OT driven industrial control systems (ICSs) 7615 such as water/power/transportation networks are 7616 increasingly meeting the daily functional needs 7617 of civilian society around the globe. This, 7618 alongside making societal businesses more 7619 automated, efficient, productive, and 7620 profitable. However, often poorly configured IoT 7621 security settings increase the chances of 7622 occurrence of (nation-sponsored) stealthy 7623 spread-based APT malware attacks in ICSs that 7624 might go undetected over a considerable period 7625 of time. The ICS enterprise management is often 7626 keen to get apriori statistical estimates of 7627 cyber-loss impact post any cyber-attack event 7628 such that it can plan ahead on its 7629 cyber-resilience budget. In this paper, we 7630 propose the first mathematical theory, based 7631 upon stochastic processes and concentration 7632 inequalities, to (a) statistically quantify 7633 apriori the cyber-loss impact (distribution) on 7634 an ICS infrastructure network post an APT 7635 cyber-attack event, and subsequently (b) bound 7636 the tail of such a cyber-risk distribution, for 7637 arbitrary impact distributions. 7638 </blockquote> 7639 </div> 7640 </div> 7641 </div> 7642 </div> 7643 <div class="slot-urls"></div> 7644 <a href="/wsc23papers/070.pdf" target="_blank">pdf</a 7645 ><br /> 7646 </div> 7647 <div class="slot-entry"> 7648 <a name="inv184" tabindex="-1"></a> 7649 <div class="slot-title-line"> 7650 <span class="slot-title" 7651 >Safeguarding Infrastructure from Cyber Threats with 7652 NLP-based Information Retrieval</span 7653 > 7654 </div> 7655 <div class="slot-authors"> 7656 Christin J. Salley, Neda Mohammadi, and John E. Taylor 7657 (Georgia Institute of Technology) 7658 </div> 7659 <div class="slot-abstract"> 7660 <div> 7661 <a 7662 class="clickable no-decoration" 7663 id="vhsjs_view_152_1707793551_6875057" 7664 onclick="$('#vhsjs_view_152_1707793551_6875057').hide(); 7665 $('#vhsjs_hide_152_1707793551_6875057').show(); 7666 $('#151_1707793551_687498').slideDown(function() { 7667 if (typeof Masonry === 'function') { 7668 $('.use_masonry').masonry(); 7669 }; 7670 7671 });" 7672 ><i class="fa fa-caret-right"></i> 7673 <span class="hover_link">Abstract</span></a 7674 ><a 7675 class="clickable no-decoration" 7676 id="vhsjs_hide_152_1707793551_6875057" 7677 onclick="$('#151_1707793551_687498').hide(function() { 7678 if (typeof Masonry === 'function') { 7679 $('.use_masonry').masonry(); 7680 }; 7681 }); 7682 $('#vhsjs_hide_152_1707793551_6875057').hide(); 7683 $('#vhsjs_view_152_1707793551_6875057').show();" 7684 style="display: none" 7685 ><i class="fa fa-caret-down"></i>
7686 <span class="hover_link">Abstract</span></a 7687 > 7688 <div 7689 data-display-control="152_1707793551_6875057" 7690 id="151_1707793551_687498" 7691 style="display: none" 7692 > 7693 <div class="arrow-slidedown"> 7694 <blockquote> 7695 Natural disasters disrupt systems, leading to 7696 critical infrastructure vulnerabilities prone to 7697 cyber-attacks. The MITRE ATT&CK Enterprise 7698 Matrix is a knowledge base for threat analyses 7699 in the cybersecurity community. Existing 7700 processes to derive possible attack 7701 methodologies from this Matrix are largely 7702 manual and time-consuming. It is essential to 7703 automate the information retrieval process to 7704 reduce human errors, improve efficiency, and 7705 free up resources for identifying unrevealed 7706 cyber-attacks. We propose a framework that 7707 incorporates Natural Language Processing (NLP) 7708 and Text Mining to automatically generate sets 7709 of attack paths from the technique descriptions 7710 in the Matrix. The framework generates 7711 similarity between techniques based on their 7712 descriptions and creates an output showing 7713 potential pathways an adversary can take to 7714 infiltrate a system. The outputs are compared 7715 against an annotated approach and attack report. 7716 The results of this study provide an approach to 7717 more quickly and effectively assess potential 7718 cyber-attacks towards protecting critical 7719 infrastructure. 7720 </blockquote> 7721 </div> 7722 </div> 7723 </div> 7724 </div> 7725 <div class="slot-urls"></div> 7726 <a href="/wsc23papers/071.pdf" target="_blank">pdf</a 7727 ><br /> 7728 </div> 7729 <div class="slot-entry"> 7730 <a name="cea115" tabindex="-1"></a> 7731 <div class="slot-title-line"> 7732 <span class="slot-title" 7733 >Modeling of Circular Economy Strategies for CFRP-made 7734 Aircrafts</span 7735 > 7736 </div> 7737 <div class="slot-authors"> 7738 Arnd Schirrmann and Uwe Beier (Airbus) 7739 </div> 7740 <div class="slot-abstract"> 7741 <div> 7742 <a 7743 class="clickable no-decoration" 7744 id="vhsjs_view_154_1707793551_6895576" 7745 onclick="$('#vhsjs_view_154_1707793551_6895576').hide(); 7746 $('#vhsjs_hide_154_1707793551_6895576').show(); 7747 $('#153_1707793551_6895497').slideDown(function() { 7748 if (typeof Masonry === 'function') { 7749 $('.use_masonry').masonry(); 7750 }; 7751 7752 });" 7753 ><i class="fa fa-caret-right"></i> 7754 <span class="hover_link">Abstract</span></a 7755 ><a 7756 class="clickable no-decoration" 7757 id="vhsjs_hide_154_1707793551_6895576" 7758 onclick="$('#153_1707793551_6895497').hide(function() { 7759 if (typeof Masonry === 'function') { 7760 $('.use_masonry').masonry(); 7761 }; 7762 }); 7763 $('#vhsjs_hide_154_1707793551_6895576').hide(); 7764 $('#vhsjs_view_154_1707793551_6895576').show();" 7765 style="display: none" 7766 ><i class="fa fa-caret-down"></i>
7767 <span class="hover_link">Abstract</span></a 7768 > 7769 <div 7770 data-display-control="154_1707793551_6895576" 7771 id="153_1707793551_6895497" 7772 style="display: none" 7773 > 7774 <div class="arrow-slidedown"> 7775 <blockquote> 7776 In a circular economy, recycling of materials at 7777 the end of a product's life cycle is a key 7778 issue. This paper discusses the sustainability 7779 impacts of different recycling strategies for 7780 CFPR-made aircraft and how they weigh up against 7781 alternative measures such as waste reduction and 7782 lower material consumption in the manufacture of 7783 the product. The analysis includes environmental 7784 and cost impacts for different strategies and 7785 market scenarios. A quantitative system dynamic 7786 simulation of the life cycle of an aircraft 7787 program is used. The subject of the life cycle 7788 simulation model is the CFRP mass flow, CO2 7789 emissions and associated costs. In addition, the 7790 effects of R&T investments in new technologies 7791 for recycling and waste prevention as well as 7792 the reduction of material consumption were 7793 investigated. 7794 </blockquote> 7795 </div> 7796 </div> 7797 </div> 7798 </div> 7799 <div class="slot-urls"></div> 7800 <a href="/wsc23papers/cea115.pdf" target="_blank">pdf</a 7801 ><br /> 7802 </div> 7803 </div> 7804 <div class="session-entry"> 7805 <span class="session-event-type">Technical Session</span 7806 ><span class="type-track-spacer"> · </span 7807 ><span class="program-track" 7808 >Environment Sustainability and Resilience</span 7809 ><br /> 7810 <div class="session-title">Food and Supply Chains</div> 7811 <div class="session-chair"> 7812 Chair: Virginia Fani (University of Florence)<br /> 7813 </div> 7814 <div class="slot-entry"> 7815 <a name="con228" tabindex="-1"></a> 7816 <div class="slot-title-line"> 7817 <span class="slot-title" 7818 >System Dynamics Simulation of External Supply Chain 7819 Disruptions on a Simplified Semiconductor Supply 7820 Chain</span 7821 > 7822 </div> 7823 <div class="slot-authors"> 7824 Anna Christina Hartwick, Abdelgafar Ismail, Beatriz 7825 Kalil Valladão Novais, Mohammed Zeeshan, and Hans 7826 Ehm (Infineon Technologies AG) 7827 </div> 7828 <div class="slot-abstract"> 7829 <div> 7830 <a 7831 class="clickable no-decoration" 7832 id="vhsjs_view_156_1707793551_6954722" 7833 onclick="$('#vhsjs_view_156_1707793551_6954722').hide(); 7834 $('#vhsjs_hide_156_1707793551_6954722').show(); 7835 $('#155_1707793551_6954641').slideDown(function() { 7836 if (typeof Masonry === 'function') { 7837 $('.use_masonry').masonry(); 7838 }; 7839 7840 });" 7841 ><i class="fa fa-caret-right"></i>
7842 <span class="hover_link">Abstract</span></a 7843 ><a 7844 class="clickable no-decoration" 7845 id="vhsjs_hide_156_1707793551_6954722" 7846 onclick="$('#155_1707793551_6954641').hide(function() { 7847 if (typeof Masonry === 'function') { 7848 $('.use_masonry').masonry(); 7849 }; 7850 }); 7851 $('#vhsjs_hide_156_1707793551_6954722').hide(); 7852 $('#vhsjs_view_156_1707793551_6954722').show();" 7853 style="display: none" 7854 ><i class="fa fa-caret-down"></i> 7855 <span class="hover_link">Abstract</span></a 7856 > 7857 <div 7858 data-display-control="156_1707793551_6954722" 7859 id="155_1707793551_6954641" 7860 style="display: none" 7861 > 7862 <div class="arrow-slidedown"> 7863 <blockquote> 7864 Due to the vitality of semiconductor products 7865 for other industries, the production of 7866 semiconductors and impact of external 7867 disruptions on the semiconductor supply chain 7868 should be well understood. As semiconductor 7869 manufacturing is accompanied with intrinsic long 7870 manufacturing cycle times ranging from 50 to 100 7871 days where operations run 24/7, 365 days per 7872 year, correct understanding of potential 7873 disturbances should be considered. Examples of 7874 these disturbances include pandemics, extreme 7875 weather events, geopolitical tensions and war. 7876 These hazards pose various risks for supply 7877 chains, for example, the bullwhip and ripple 7878 effect. To simulate the result of such risks, a 7879 simplified system dynamics model of a typical 7880 semiconductor manufacturing supply chain was 7881 constructed using the Anylogic Software. The 7882 model serves as a what-if scenario foundation to 7883 evaluate certain external circumstances 7884 dependent on current global situations to 7885 enhance supply chain resilience 7886 </blockquote> 7887 </div> 7888 </div> 7889 </div> 7890 </div> 7891 <div class="slot-urls"></div> 7892 <a href="/wsc23papers/072.pdf" target="_blank">pdf</a 7893 ><br /> 7894 </div> 7895 <div class="slot-entry"> 7896 <a name="inv181" tabindex="-1"></a> 7897 <div class="slot-title-line"> 7898 <span class="slot-title" 7899 >An Agent-Based Model of Agricultural Land Use in 7900 Support of Local Food Systems</span 7901 > 7902 </div> 7903 <div class="slot-authors"> 7904 Poojan Patel and Caroline Krejci (University of Texas at 7905 Arlington), Nicholas Schwab (University of Northern 7906 Iowa), and Michael Dorneich (Iowa State University) 7907 </div> 7908 <div class="slot-abstract"> 7909 <div> 7910 <a 7911 class="clickable no-decoration"
7912 id="vhsjs_view_158_1707793551_6978538" 7913 onclick="$('#vhsjs_view_158_1707793551_6978538').hide(); 7914 $('#vhsjs_hide_158_1707793551_6978538').show(); 7915 $('#157_1707793551_697846').slideDown(function() { 7916 if (typeof Masonry === 'function') { 7917 $('.use_masonry').masonry(); 7918 }; 7919 7920 });" 7921 ><i class="fa fa-caret-right"></i> 7922 <span class="hover_link">Abstract</span></a 7923 ><a 7924 class="clickable no-decoration" 7925 id="vhsjs_hide_158_1707793551_6978538" 7926 onclick="$('#157_1707793551_697846').hide(function() { 7927 if (typeof Masonry === 'function') { 7928 $('.use_masonry').masonry(); 7929 }; 7930 }); 7931 $('#vhsjs_hide_158_1707793551_6978538').hide(); 7932 $('#vhsjs_view_158_1707793551_6978538').show();" 7933 style="display: none" 7934 ><i class="fa fa-caret-down"></i> 7935 <span class="hover_link">Abstract</span></a 7936 > 7937 <div 7938 data-display-control="158_1707793551_6978538" 7939 id="157_1707793551_697846" 7940 style="display: none" 7941 > 7942 <div class="arrow-slidedown"> 7943 <blockquote> 7944 Local food systems, in which consumers source 7945 food from nearby farmers, offer a sustainable 7946 alternative to the modern industrial food supply 7947 system. However, scaling up local food 7948 production to meet consumer demand will require 7949 farmers to allocate more land to this purpose. 7950 This paper describes an agent-based model that 7951 represents commodity-producing Iowa farmers and 7952 their decisions about converting some of their 7953 acreage to specialty crop production for local 7954 consumption. Farmer agents’ land-use 7955 decisions are informed by messages passed to 7956 them via their social connections with other 7957 farmers in their communities and messages from 7958 agricultural extension agents. Preliminary 7959 experimentation revealed that leveraging 7960 extension agents to increase the frequency and 7961 strength of messages to farmers in support of 7962 local food production has a modest positive 7963 impact on adoption. By itself, however, this 7964 intervention is unlikely to yield significant 7965 improvements to food system sustainability. 7966 </blockquote> 7967 </div> 7968 </div> 7969 </div> 7970 </div> 7971 <div class="slot-urls"></div> 7972 <a href="/wsc23papers/074.pdf" target="_blank">pdf</a 7973 ><br /> 7974 </div> 7975 </div> 7976 <div class="session-entry"> 7977 <span class="session-event-type">Technical Session</span 7978 ><span class="type-track-spacer"> · </span 7979 ><span class="program-track" 7980 >Environment Sustainability and Resilience</span 7981 ><br /> 7982 <div class="session-title"> 7983 Simulation for Sustainability 7984 </div> 7985 <div class="session-chair"> 7986 Chair: Jonathan M. Gilligan (Vanderbilt University)<br /> 7987 </div> 7988 <div class="slot-entry"> 7989 <a name="con209" tabindex="-1"></a> 7990 <div class="slot-title-line">
7991 <span class="slot-title" 7992 >Sustainability Assessment Through Simulation: The 7993 Case Of Fashion Renting</span 7994 > 7995 </div> 7996 <div class="slot-authors"> 7997 Virginia Fani and Romeo Bandinelli (University of 7998 Florence) 7999 </div> 8000 <div class="slot-abstract"> 8001 <div> 8002 <a 8003 class="clickable no-decoration" 8004 id="vhsjs_view_160_1707793551_703358" 8005 onclick="$('#vhsjs_view_160_1707793551_703358').hide(); 8006 $('#vhsjs_hide_160_1707793551_703358').show(); 8007 $('#159_1707793551_7033498').slideDown(function() { 8008 if (typeof Masonry === 'function') { 8009 $('.use_masonry').masonry(); 8010 }; 8011 8012 });" 8013 ><i class="fa fa-caret-right"></i> 8014 <span class="hover_link">Abstract</span></a 8015 ><a 8016 class="clickable no-decoration" 8017 id="vhsjs_hide_160_1707793551_703358" 8018 onclick="$('#159_1707793551_7033498').hide(function() { 8019 if (typeof Masonry === 'function') { 8020 $('.use_masonry').masonry(); 8021 }; 8022 }); 8023 $('#vhsjs_hide_160_1707793551_703358').hide(); 8024 $('#vhsjs_view_160_1707793551_703358').show();" 8025 style="display: none" 8026 ><i class="fa fa-caret-down"></i> 8027 <span class="hover_link">Abstract</span></a 8028 > 8029 <div 8030 data-display-control="160_1707793551_703358" 8031 id="159_1707793551_7033498" 8032 style="display: none" 8033 > 8034 <div class="arrow-slidedown"> 8035 <blockquote> 8036 The fashion industry is widely known as one of 8037 the most environmentally impacting. To address 8038 the overconsumption issue, the fashion renting 8039 business model allows renting clothes or 8040 accessories instead of buying them, extending 8041 the useful life of products. However, concerns 8042 about the sustainability of fashion renting 8043 supply chains are arisen, especially due to 8044 reverse logistics. In this context, a hybrid 8045 simulation model is developed to support fashion 8046 companies in the design and evaluation of 8047 renting supply chain configurations. Through 8048 Discrete Event Simulation (DES) logistics flows 8049 are represented, while Agent-Based Modeling 8050 (ABM) integrated with Geographic Information 8051 System (GIS) allow to represent supply 8052 chain’s nodes in the real environment. GIS 8053 concurs to estimate the sustainability of the 8054 supply chain importing effective data related to 8055 the covered distances. The proposed parametric 8056 model will enable performing scenario analyses 8057 to assess the best configuration in terms of 8058 environmental impact. 8059 </blockquote> 8060 </div> 8061 </div> 8062 </div> 8063 </div> 8064 <div class="slot-urls"></div> 8065 <a href="/wsc23papers/075.pdf" target="_blank">pdf</a 8066 ><br /> 8067 </div> 8068 <div class="slot-entry"> 8069 <a name="cea108" tabindex="-1"></a> 8070 <div class="slot-title-line">
8071 <span class="slot-title" 8072 >Simulative Analysis of the Sustainability Driven 8073 Transformation of Casting Plants</span 8074 > 8075 </div> 8076 <div class="slot-authors"> 8077 Johannes Dettelbacher, Wolfgang Schlüter, and 8078 Alexander Buchele (Ansbach University of Applied 8079 Sciences) 8080 </div> 8081 <div class="slot-abstract"> 8082 <div> 8083 <a 8084 class="clickable no-decoration" 8085 id="vhsjs_view_162_1707793551_7055008" 8086 onclick="$('#vhsjs_view_162_1707793551_7055008').hide(); 8087 $('#vhsjs_hide_162_1707793551_7055008').show(); 8088 $('#161_1707793551_7054927').slideDown(function() { 8089 if (typeof Masonry === 'function') { 8090 $('.use_masonry').masonry(); 8091 }; 8092 8093 });" 8094 ><i class="fa fa-caret-right"></i> 8095 <span class="hover_link">Abstract</span></a 8096 ><a 8097 class="clickable no-decoration" 8098 id="vhsjs_hide_162_1707793551_7055008" 8099 onclick="$('#161_1707793551_7054927').hide(function() { 8100 if (typeof Masonry === 'function') { 8101 $('.use_masonry').masonry(); 8102 }; 8103 }); 8104 $('#vhsjs_hide_162_1707793551_7055008').hide(); 8105 $('#vhsjs_view_162_1707793551_7055008').show();" 8106 style="display: none" 8107 ><i class="fa fa-caret-down"></i> 8108 <span class="hover_link">Abstract</span></a 8109 > 8110 <div 8111 data-display-control="162_1707793551_7055008" 8112 id="161_1707793551_7054927" 8113 style="display: none" 8114 > 8115 <div class="arrow-slidedown"> 8116 <blockquote> 8117 The current energy crisis and high fossil fuel 8118 costs are challenging energy intensive 8119 industries such as non-ferrous foundries. It is 8120 therefore important to promote the transition to 8121 renewable energy sources with the 8122 electrification of melting units. This pilot 8123 study is the first to simulate the transition of 8124 conventional foundries to sustainable 8125 technologies. For this purpose, a simulation 8126 model based on a selected example company is 8127 developed. It takes into account the energy 8128 consumption and the logistical effects of a 8129 converted operation. The simulation model is 8130 implemented as a hybrid simulation combining a 8131 discrete event simulation at the plant level and 8132 a process simulation within the furnaces. The 8133 study shows how a sustainable energy supply can 8134 be achieved in foundries. The effects of 8135 efficiency as well as energy costs and emissions 8136 are also taken into account. 8137 </blockquote> 8138 </div> 8139 </div> 8140 </div> 8141 </div> 8142 <div class="slot-urls"></div> 8143 <a href="/wsc23papers/cea108.pdf" target="_blank">pdf</a 8144 ><br /> 8145 </div> 8146 <div class="slot-entry"> 8147 <a name="inv134" tabindex="-1"></a> 8148 <div class="slot-title-line">
8149 <span class="slot-title" 8150 >A Customizable Community-Building-Energy-Modeling 8151 Decision Support System (CCBEM-DSS) for Net-Zero 8152 Planning in Developing Countries</span 8153 > 8154 </div> 8155 <div class="slot-authors"> 8156 Omprakash Ramalingam Rethnam and Albert Thomas (Indian 8157 Institute of Technology Bombay) 8158 </div> 8159 <div class="slot-abstract"> 8160 <div> 8161 <a 8162 class="clickable no-decoration" 8163 id="vhsjs_view_164_1707793551_707606" 8164 onclick="$('#vhsjs_view_164_1707793551_707606').hide(); 8165 $('#vhsjs_hide_164_1707793551_707606').show(); 8166 $('#163_1707793551_707598').slideDown(function() { 8167 if (typeof Masonry === 'function') { 8168 $('.use_masonry').masonry(); 8169 }; 8170 8171 });" 8172 ><i class="fa fa-caret-right"></i> 8173 <span class="hover_link">Abstract</span></a 8174 ><a 8175 class="clickable no-decoration" 8176 id="vhsjs_hide_164_1707793551_707606" 8177 onclick="$('#163_1707793551_707598').hide(function() { 8178 if (typeof Masonry === 'function') { 8179 $('.use_masonry').masonry(); 8180 }; 8181 }); 8182 $('#vhsjs_hide_164_1707793551_707606').hide(); 8183 $('#vhsjs_view_164_1707793551_707606').show();" 8184 style="display: none" 8185 ><i class="fa fa-caret-down"></i> 8186 <span class="hover_link">Abstract</span></a 8187 > 8188 <div 8189 data-display-control="164_1707793551_707606" 8190 id="163_1707793551_707598" 8191 style="display: none" 8192 > 8193 <div class="arrow-slidedown"> 8194 <blockquote> 8195 Buildings contribute to about 40% of global 8196 energy-related CO2 emissions, and reducing 8197 energy demand in buildings has become one of the 8198 vital components of the current climate change 8199 mitigation strategies. Optimizing energy for the 8200 urban building stock by energy-efficient 8201 retrofits is becoming increasingly popular in 8202 developed countries where the functional and 8203 construction elements of the stock are uniform, 8204 along with the updated stock database already 8205 built in desirable standard formats for energy 8206 simulation exchange. However, a decision support 8207 system to arrive at energy-efficient retrofits 8208 for developing countries where the building 8209 stock is highly diverse, with varying 8210 construction and operational philosophies, and 8211 has no readily available datasets of existing 8212 stock is highly challenging. To close this gap, 8213 this study suggests an adaptable decentralized 8214 community building energy simulation and 8215 modeling schema using free and open-source tools 8216 for retrofit decision-making. 8217 </blockquote> 8218 </div> 8219 </div> 8220 </div> 8221 </div> 8222 <div class="slot-urls"></div> 8223 <a href="/wsc23papers/266.pdf" target="_blank">pdf</a 8224 ><br /> 8225 </div> 8226 </div> 8227 <div class="session-entry">
8228 <span class="session-event-type">Technical Session</span 8229 ><span class="type-track-spacer"> · </span 8230 ><span class="program-track" 8231 >Environment Sustainability and Resilience</span 8232 ><br /> 8233 <div class="session-title"> 8234 Electric and Autonomous Transportation 8235 </div> 8236 <div class="session-chair"> 8237 Chair: Neda Mohammadi (Georgia Institute of Technology)<br /> 8238 </div> 8239 <div class="slot-entry"> 8240 <a name="con118" tabindex="-1"></a> 8241 <div class="slot-title-line"> 8242 <span class="slot-title" 8243 >Simulation, Optimization and Control of Trajectories 8244 of ASVs Performing HACBS Monitoring Missions in Lentic 8245 Waters</span 8246 > 8247 </div> 8248 <div class="slot-authors"> 8249 Alfredo Gonzalez-Calvin, Lía García-Perez, 8250 José Luis Risco-Martín, and Eva Besada-Portas 8251 (Complutense University of Madrid) 8252 </div> 8253 <div class="slot-abstract"> 8254 <div> 8255 <a 8256 class="clickable no-decoration" 8257 id="vhsjs_view_166_1707793551_713462" 8258 onclick="$('#vhsjs_view_166_1707793551_713462').hide(); 8259 $('#vhsjs_hide_166_1707793551_713462').show(); 8260 $('#165_1707793551_7134538').slideDown(function() { 8261 if (typeof Masonry === 'function') { 8262 $('.use_masonry').masonry(); 8263 }; 8264 8265 });" 8266 ><i class="fa fa-caret-right"></i> 8267 <span class="hover_link">Abstract</span></a 8268 ><a 8269 class="clickable no-decoration" 8270 id="vhsjs_hide_166_1707793551_713462" 8271 onclick="$('#165_1707793551_7134538').hide(function() { 8272 if (typeof Masonry === 'function') { 8273 $('.use_masonry').masonry(); 8274 }; 8275 }); 8276 $('#vhsjs_hide_166_1707793551_713462').hide(); 8277 $('#vhsjs_view_166_1707793551_713462').show();" 8278 style="display: none" 8279 ><i class="fa fa-caret-down"></i> 8280 <span class="hover_link">Abstract</span></a 8281 > 8282 <div 8283 data-display-control="166_1707793551_713462" 8284 id="165_1707793551_7134538" 8285 style="display: none" 8286 > 8287 <div class="arrow-slidedown"> 8288 <blockquote> 8289 Harmful Algae and Cyanobacteria Blooms (HACBs) 8290 are dangerous dynamic processes for the 8291 users/inhabitants of the hydric resources. Their 8292 development and contingency plans can be 8293 anticipated by using Autonomous Surface Vehicles 8294 (ASVs) equipped with a self-driven system 8295 capable of deciding how to displace the ASV and 8296 its multi-parametric probe to take measurements 8297 in the 3D locations of the water body where the 8298 HACB is likely to occur. This paper presents a 8299 new self-driven system for that purpose, 8300 consistent on 1) an offline trajectory planner 8301 for the ASV that exploits the information 8302 provided by a commercial HACBs simulator to 8303 optimize, in turn, the ASV horizontal and probe 8304 vertical displacements;
8304 and 2) a guidance and 8305 control system specially designed for making the 8306 ASV follow the planned trajectories. The paper 8307 also presents a comprehensive set of simulations 8308 to evaluate our proposal's performance and 8309 adjust its parameters. 8310 </blockquote> 8311 </div> 8312 </div> 8313 </div> 8314 </div> 8315 <div class="slot-urls"></div> 8316 <a href="/wsc23papers/076.pdf" target="_blank">pdf</a 8317 ><br /> 8318 </div> 8319 <div class="slot-entry"> 8320 <a name="con284" tabindex="-1"></a> 8321 <div class="slot-title-line"> 8322 <span class="slot-title" 8323 >Lightweight Smart Charging vs. Immediate Charging 8324 with Buffer Storage: Towards a Simulation Study for 8325 Electric Vehicle Grid Integration at Workplaces</span 8326 > 8327 </div> 8328 <div class="slot-authors"> 8329 Paul Benz and Marco Pruckner (Universität 8330 Würzburg) 8331 </div> 8332 <div class="slot-abstract"> 8333 <div> 8334 <a 8335 class="clickable no-decoration" 8336 id="vhsjs_view_168_1707793551_7156966" 8337 onclick="$('#vhsjs_view_168_1707793551_7156966').hide(); 8338 $('#vhsjs_hide_168_1707793551_7156966').show(); 8339 $('#167_1707793551_7156882').slideDown(function() { 8340 if (typeof Masonry === 'function') { 8341 $('.use_masonry').masonry(); 8342 }; 8343 8344 });" 8345 ><i class="fa fa-caret-right"></i> 8346 <span class="hover_link">Abstract</span></a 8347 ><a 8348 class="clickable no-decoration" 8349 id="vhsjs_hide_168_1707793551_7156966" 8350 onclick="$('#167_1707793551_7156882').hide(function() { 8351 if (typeof Masonry === 'function') { 8352 $('.use_masonry').masonry(); 8353 }; 8354 }); 8355 $('#vhsjs_hide_168_1707793551_7156966').hide(); 8356 $('#vhsjs_view_168_1707793551_7156966').show();" 8357 style="display: none" 8358 ><i class="fa fa-caret-down"></i> 8359 <span class="hover_link">Abstract</span></a 8360 > 8361 <div 8362 data-display-control="168_1707793551_7156966" 8363 id="167_1707793551_7156882" 8364 style="display: none" 8365 > 8366 <div class="arrow-slidedown"> 8367 <blockquote> 8368 The present study investigates the extension of 8369 an existing simulation model combining system 8370 dynamics and discrete event simulation by linear 8371 optimization for an electric vehicle charging 8372 system. The existing simulation framework is 8373 extended by a smart charging strategy based on 8374 linear programming in order to exploit the 8375 flexibility of real charging processes at a 8376 workplace parking lot for a better integration 8377 of solar photovoltaic electricity generation. 8378 Therefore, different smart charging strategies 8379 are evaluated. In multiple simulation runs, the 8380 strategies are compared with immediate charging 8381 using a stationary battery energy storage system 8382 for intermediate storage of electricity 8383 generated by solar photovoltaic. Results show 8384 that smart charging strategies can achieve 8385 similarly good results with respect to the
8386 self-sufficiency rate and self-consumption rate. 8387 In the context of a 100kWp PV system the 8388 combination of optimizing charging rates and 8389 stationary battery energy storage resulted in 8390 self-sufficiency rates of more than 90% in the 8391 simulation. 8392 </blockquote> 8393 </div> 8394 </div> 8395 </div> 8396 </div> 8397 <div class="slot-urls"></div> 8398 <a href="/wsc23papers/077.pdf" target="_blank">pdf</a 8399 ><br /> 8400 </div> 8401 <div class="slot-entry"> 8402 <a name="cea161" tabindex="-1"></a> 8403 <div class="slot-title-line"> 8404 <span class="slot-title" 8405 >A Simulation-Based Decision Support Tool for Direct 8406 Current Fast Charger Installations</span 8407 > 8408 </div> 8409 <div class="slot-authors"> 8410 Cathy Rupp (BC Hydro); Deep Jariwala, Suellen Ventura, 8411 and Scott Nason (SAS Institute (Canada) , Inc); Bahar 8412 Biller (SAS Institute, Inc); and Yanan Sun and Parvir 8413 Girn (BC Hydro) 8414 </div> 8415 <div class="slot-abstract"> 8416 <div> 8417 <a 8418 class="clickable no-decoration" 8419 id="vhsjs_view_170_1707793551_7179558" 8420 onclick="$('#vhsjs_view_170_1707793551_7179558').hide(); 8421 $('#vhsjs_hide_170_1707793551_7179558').show(); 8422 $('#169_1707793551_7179477').slideDown(function() { 8423 if (typeof Masonry === 'function') { 8424 $('.use_masonry').masonry(); 8425 }; 8426 8427 });" 8428 ><i class="fa fa-caret-right"></i> 8429 <span class="hover_link">Abstract</span></a 8430 ><a 8431 class="clickable no-decoration" 8432 id="vhsjs_hide_170_1707793551_7179558" 8433 onclick="$('#169_1707793551_7179477').hide(function() { 8434 if (typeof Masonry === 'function') { 8435 $('.use_masonry').masonry(); 8436 }; 8437 }); 8438 $('#vhsjs_hide_170_1707793551_7179558').hide(); 8439 $('#vhsjs_view_170_1707793551_7179558').show();" 8440 style="display: none" 8441 ><i class="fa fa-caret-down"></i> 8442 <span class="hover_link">Abstract</span></a 8443 > 8444 <div 8445 data-display-control="170_1707793551_7179558" 8446 id="169_1707793551_7179477" 8447 style="display: none" 8448 > 8449 <div class="arrow-slidedown"> 8450 <blockquote> 8451 We develop a simulation-based tool for 8452 supporting direct current fast charger (DCFC) 8453 installation decisions. Our simulation captures 8454 details of the DCFC network configuration, 8455 non-stationary arrival patterns of the electric 8456 vehicles to the fast charging DCFC stations, 8457 various DCFC attributes, charging time 8458 distributions, and customer behavior. The 8459 statistical analysis of the simulation generated 8460 output data produces various key performance 8461 indicators (KPIs) including DCFC utilizations, 8462 number of electric vehicles charged and left 8463 uncharged, and queueing experience of the 8464 customers. One of the key challenges of 8465 developing this simulation is its validation: we 8466 have validated the simulation with the 8467 historical DCFC charging session data and past 8468 observations of the DCFC utilizations. The 8469 resulting data-driven simulation is used for 8470 supporting DCFC planning through its capability
8471 to conduct scenario analysis and predict various 8472 KPIs. 8473 </blockquote> 8474 </div> 8475 </div> 8476 </div> 8477 </div> 8478 <div class="slot-urls"></div> 8479 <a href="/wsc23papers/cea161.pdf" target="_blank">pdf</a 8480 ><br /> 8481 </div> 8482 </div> 8483 <div class="session-entry"> 8484 <span class="session-event-type">Technical Session</span 8485 ><span class="type-track-spacer"> · </span 8486 ><span class="program-track" 8487 >Environment Sustainability and Resilience</span 8488 ><br /> 8489 <div class="session-title"> 8490 Water and Environmental Resources 8491 </div> 8492 <div class="session-chair"> 8493 Chair: Christin Salley (Georgia Institute of 8494 Technology)<br /> 8495 </div> 8496 <div class="slot-entry"> 8497 <a name="con110" tabindex="-1"></a> 8498 <div class="slot-title-line"> 8499 <span class="slot-title" 8500 >Equity-Driven Management of Essential Environmental 8501 Resources Under Price-Based Consumption</span 8502 > 8503 </div> 8504 <div class="slot-authors"> 8505 Shai Amouyal and Noa Zychlinski (Technion - Israel 8506 Institute of Technology) 8507 </div> 8508 <div class="slot-abstract"> 8509 <div> 8510 <a 8511 class="clickable no-decoration" 8512 id="vhsjs_view_172_1707793551_7235184" 8513 onclick="$('#vhsjs_view_172_1707793551_7235184').hide(); 8514 $('#vhsjs_hide_172_1707793551_7235184').show(); 8515 $('#171_1707793551_7235103').slideDown(function() { 8516 if (typeof Masonry === 'function') { 8517 $('.use_masonry').masonry(); 8518 }; 8519 8520 });" 8521 ><i class="fa fa-caret-right"></i> 8522 <span class="hover_link">Abstract</span></a 8523 ><a 8524 class="clickable no-decoration" 8525 id="vhsjs_hide_172_1707793551_7235184" 8526 onclick="$('#171_1707793551_7235103').hide(function() { 8527 if (typeof Masonry === 'function') { 8528 $('.use_masonry').masonry(); 8529 }; 8530 }); 8531 $('#vhsjs_hide_172_1707793551_7235184').hide(); 8532 $('#vhsjs_view_172_1707793551_7235184').show();" 8533 style="display: none" 8534 ><i class="fa fa-caret-down"></i> 8535 <span class="hover_link">Abstract</span></a 8536 > 8537 <div 8538 data-display-control="172_1707793551_7235184" 8539 id="171_1707793551_7235103" 8540 style="display: none" 8541 > 8542 <div class="arrow-slidedown"> 8543 <blockquote> 8544 The global climate crisis and population growth 8545 restrict the availability of essential 8546 environmental resources such as water and energy 8547 and this situation continues to deteriorate. If 8548 and when conditions become extreme, only the 8549 well-offs will have access to these valuable 8550 resources. With that in mind, we look for 8551 solutions to achieve equity within societies 8552 while preserving, the degree possible, natural 8553 resources. We suggest a method for setting 8554 differential pricing for each population 8555 stratum, so that each spends a relatively 8556 similar percentage of their income on these 8557 basic commodities, without depleting valuable 8558 resources. Our method optimizes the prices while 8559 simultaneously estimating the unknown 8560 consumption–price relation. We show the 8561 effectiveness of our method based on data from 8562 Israel and through extensive simulation 8563 experiments reflecting different levels of 8564 income inequality within societies, different 8565 consumption–price relations, and resource 8566 availability. Our study shows that equity and 8567 resource preservation can go hand-in-hand. 8568 </blockquote> 8569 </div> 8570 </div> 8571 </div> 8572 </div> 8573 <div class="slot-urls"></div> 8574 <a href="/wsc23papers/078.pdf" target="_blank">pdf</a 8575 ><br /> 8576 </div> 8577 <div class="slot-entry"> 8578 <a name="inv179" tabindex="-1"></a> 8579 <div class="slot-title-line">
8580 <span class="slot-title" 8581 >Modeling the Dynamics of Sediment Transport, Tides, 8582 and Sea-Level Rise: Implications for the Resilience of 8583 Coastal Bengal</span 8584 > 8585 </div> 8586 <div class="slot-authors"> 8587 Christopher M. Tasich, Jonathan M. Gilligan, and George 8588 M. Hornberger (Vanderbilt University) 8589 </div> 8590 <div class="slot-abstract"> 8591 <div> 8592 <a 8593 class="clickable no-decoration" 8594 id="vhsjs_view_174_1707793551_7258265" 8595 onclick="$('#vhsjs_view_174_1707793551_7258265').hide(); 8596 $('#vhsjs_hide_174_1707793551_7258265').show(); 8597 $('#173_1707793551_7258186').slideDown(function() { 8598 if (typeof Masonry === 'function') { 8599 $('.use_masonry').masonry(); 8600 }; 8601 8602 });" 8603 ><i class="fa fa-caret-right"></i> 8604 <span class="hover_link">Abstract</span></a 8605 ><a 8606 class="clickable no-decoration" 8607 id="vhsjs_hide_174_1707793551_7258265" 8608 onclick="$('#173_1707793551_7258186').hide(function() { 8609 if (typeof Masonry === 'function') { 8610 $('.use_masonry').masonry(); 8611 }; 8612 }); 8613 $('#vhsjs_hide_174_1707793551_7258265').hide(); 8614 $('#vhsjs_view_174_1707793551_7258265').show();" 8615 style="display: none" 8616 ><i class="fa fa-caret-down"></i> 8617 <span class="hover_link">Abstract</span></a 8618 > 8619 <div 8620 data-display-control="174_1707793551_7258265" 8621 id="173_1707793551_7258186" 8622 style="display: none" 8623 > 8624 <div class="arrow-slidedown"> 8625 <blockquote> 8626 The coastal zone of the 8627 Ganges-Brahmaputra-Meghna (GBM) Delta is widely 8628 recognized as one of the most vulnerable places 8629 to sea-level rise (SLR), with around 57 million 8630 people living within 5 m of sea level. Sediment 8631 transported by the Ganges, Brahmaputra, and 8632 Meghna rivers has the potential to raise the 8633 land and offset SLR. There is significant 8634 uncertainty in future sediment supply and SLR, 8635 which raises questions about the sustainability 8636 of the delta. We present a simple model, driven 8637 by basic physics, to estimate the evolution of 8638 the landscape under different conditions at low 8639 computational cost. Using a single tuning 8640 parameter, the model can match observed rates of 8641 land aggradation. We find a strong negative 8642 feedback, which robustly brings land elevation 8643 into equilibrium with changing sea level. We 8644 discuss how this model can be used to 8645 investigate the dynamics of sediment transport 8646 and the sustainability of the GBM Delta.
8647 </blockquote> 8648 </div> 8649 </div> 8650 </div> 8651 </div> 8652 <div class="slot-urls"></div> 8653 <a href="/wsc23papers/079.pdf" target="_blank">pdf</a 8654 ><br /> 8655 </div> 8656 <div class="slot-entry"> 8657 <a name="cea153" tabindex="-1"></a> 8658 <div class="slot-title-line"> 8659 <span class="slot-title" 8660 >Infrastructure Planning Using a Dynamic Simulation to 8661 Improve Sustainability and Resilience: Case Study for 8662 a Coastal Watershed</span 8663 > 8664 </div> 8665 <div class="slot-authors"> 8666 Raymond Smith (East Carolina University) 8667 </div> 8668 <div class="slot-abstract"> 8669 <div> 8670 <a 8671 class="clickable no-decoration" 8672 id="vhsjs_view_176_1707793551_7287776" 8673 onclick="$('#vhsjs_view_176_1707793551_7287776').hide(); 8674 $('#vhsjs_hide_176_1707793551_7287776').show(); 8675 $('#175_1707793551_7287698').slideDown(function() { 8676 if (typeof Masonry === 'function') { 8677 $('.use_masonry').masonry(); 8678 }; 8679 8680 });" 8681 ><i class="fa fa-caret-right"></i> 8682 <span class="hover_link">Abstract</span></a 8683 ><a 8684 class="clickable no-decoration" 8685 id="vhsjs_hide_176_1707793551_7287776" 8686 onclick="$('#175_1707793551_7287698').hide(function() { 8687 if (typeof Masonry === 'function') { 8688 $('.use_masonry').masonry(); 8689 }; 8690 }); 8691 $('#vhsjs_hide_176_1707793551_7287776').hide(); 8692 $('#vhsjs_view_176_1707793551_7287776').show();" 8693 style="display: none" 8694 ><i class="fa fa-caret-down"></i> 8695 <span class="hover_link">Abstract</span></a 8696 > 8697 <div 8698 data-display-control="176_1707793551_7287776" 8699 id="175_1707793551_7287698" 8700 style="display: none" 8701 > 8702 <div class="arrow-slidedown"> 8703 <blockquote> 8704 Climate change presents a significant challenge 8705 for many coastal communities as sea level rise 8706 is expected to cause widespread and chronic 8707 flood inundation. This study examines the case 8708 of a coastal watershed of ecological importance, 8709 which is threatened by sea level rise and land 8710 subsidence, as well as seasonal severe storms. 8711 The health of the watershed and flood inundation 8712 protection to the community depends on water 8713 outflow; something which sea level rise will 8714 further restrict. Infrastructure planning for an 8715 active water management solution resilient to 8716 severe storms and electrical grid disruptions is 8717 needed. A dynamic simulation is used to evaluate 8718 microgrid energy system design performance and 8719 effectiveness in powering a critical
8720 infrastructure pumping station during 8721 storm-related electrical grid outage and 8722 restoration scenarios. 8723 </blockquote> 8724 </div> 8725 </div> 8726 </div> 8727 </div> 8728 <div class="slot-urls"></div> 8729 <a href="/wsc23papers/cea153.pdf" target="_blank">pdf</a 8730 ><br /> 8731 </div> 8732 </div> 8733 </div> 8734 <div class="centered"> 8735 <div class="top-link"><a href="#top">Return to Top</a></div> 8736 </div> 8737 <hr /> 8738 </div> 8739 <div class="area-section"> 8740 <div class="centered"> 8741 <a name="ptrack102" tabindex="-1"></a> 8742 <div class="section-title">Introductory Tutorials</div> 8743 </div> 8744 <div class="centered track-chair"> 8745 <span class="track-chair-role" 8746 >Track Coordinator - Introductory Tutorials: </span 8747 ><span class="track-chair-names" 8748 >Sanjay Jain (The George Washington University), Chang-Han 8749 Rhee (Northwestern University)</span 8750 > 8751 </div> 8752 <div class="section-entry"> 8753 <div class="session-entry"> 8754 <span class="session-event-type">Tutorial</span 8755 ><span class="type-track-spacer"> · </span 8756 ><span class="program-track">Introductory Tutorials</span 8757 ><br /> 8758 <div class="session-title"> 8759 Importance Sampling for Minimization of Tail Risks: A 8760 Tutorial 8761 </div> 8762 <div class="session-chair"> 8763 Chair: Chang-Han Rhee (Northwestern University)<br /> 8764 </div> 8765 <div class="slot-entry"> 8766 <a name="inv177" tabindex="-1"></a> 8767 <div class="slot-authors"> 8768 Anand Deo (Indian Institute of Management Bangalore) and 8769 Karthyek Murthy (Singapore University of Technology and 8770 Design) 8771 </div> 8772 <div class="slot-abstract"> 8773 <div> 8774 <a 8775 class="clickable no-decoration" 8776 id="vhsjs_view_178_1707793551_7353423" 8777 onclick="$('#vhsjs_view_178_1707793551_7353423').hide(); 8778 $('#vhsjs_hide_178_1707793551_7353423').show(); 8779 $('#177_1707793551_735334').slideDown(function() { 8780 if (typeof Masonry === 'function') { 8781 $('.use_masonry').masonry(); 8782 }; 8783 8784 });" 8785 ><i class="fa fa-caret-right"></i> 8786 <span class="hover_link">Abstract</span></a 8787 ><a 8788 class="clickable no-decoration" 8789 id="vhsjs_hide_178_1707793551_7353423" 8790 onclick="$('#177_1707793551_735334').hide(function() { 8791 if (typeof Masonry === 'function') { 8792 $('.use_masonry').masonry(); 8793 }; 8794 }); 8795 $('#vhsjs_hide_178_1707793551_7353423').hide(); 8796 $('#vhsjs_view_178_1707793551_7353423').show();" 8797 style="display: none" 8798 ><i class="fa fa-caret-down"></i> 8799 <span class="hover_link">Abstract</span></a 8800 > 8801 <div 8802 data-display-control="178_1707793551_7353423" 8803 id="177_1707793551_735334" 8804 style="display: none" 8805 > 8806 <div class="arrow-slidedown"> 8807 <blockquote> 8808 This paper provides an introductory overview of 8809 how one may employ importance sampling (IS) 8810 effectively as a tool for solving stochastic 8811 optimization formulations incorporating tail 8812 risk measures such as Conditional Value-at-Risk. 8813 Approximating the tail risk measure by its 8814 sample average approximation, while appealing 8815 due to its simplicity and universality in use, 8816 requires a large number of samples to be able to 8817 arrive at risk-minimizing decisions with high 8818 confidence. In simulation, IS is among the most 8819 prominent methods for substantially reducing the 8820 sample requirement while estimating 8821 probabilities of rare tail events. Can IS be 8822 similarly effective for optimization as well? 8823 This tutorial aims to provide an overview of the 8824 two key ingredients in this regard, namely, (i) 8825 how one may arrive at an effective importance 8826 sampling change of measure prescription at every 8827 decision, and (ii) the prominent techniques 8828 available for integrating such a prescription 8829 within a solution paradigm for stochastic 8830 optimization. 8831 </blockquote> 8832 </div> 8833 </div> 8834 </div> 8835 </div> 8836 <div class="slot-urls"></div> 8837 <a href="/wsc23papers/120.pdf" target="_blank">pdf</a 8838 ><br /> 8839 </div> 8840 </div> 8841 <div class="session-entry">
8842 <span class="session-event-type">Tutorial</span 8843 ><span class="type-track-spacer"> · </span 8844 ><span class="program-track">Introductory Tutorials</span 8845 ><br /> 8846 <div class="session-title"> 8847 Event Graphs: Syntax, Semantics, and Implementation 8848 </div> 8849 <div class="session-chair"> 8850 Chair: Md Tariqul Islam (Purdue University)<br /> 8851 </div> 8852 <div class="slot-entry"> 8853 <a name="inv197" tabindex="-1"></a> 8854 <div class="slot-authors"> 8855 Murat M. Gunal (Fenerbahce University); Yahya Ismail 8856 Osais (King Fahd University of Petroleum and Minerals, 8857 Interdisc. Research Center for Intellig. Secure 8858 Systems); and Gerd Wagner (Brandenburg University of 8859 Technology) 8860 </div> 8861 <div class="slot-abstract"> 8862 <div> 8863 <a 8864 class="clickable no-decoration" 8865 id="vhsjs_view_180_1707793551_7414422" 8866 onclick="$('#vhsjs_view_180_1707793551_7414422').hide(); 8867 $('#vhsjs_hide_180_1707793551_7414422').show(); 8868 $('#179_1707793551_7414346').slideDown(function() { 8869 if (typeof Masonry === 'function') { 8870 $('.use_masonry').masonry(); 8871 }; 8872 8873 });" 8874 ><i class="fa fa-caret-right"></i> 8875 <span class="hover_link">Abstract</span></a 8876 ><a 8877 class="clickable no-decoration" 8878 id="vhsjs_hide_180_1707793551_7414422" 8879 onclick="$('#179_1707793551_7414346').hide(function() { 8880 if (typeof Masonry === 'function') { 8881 $('.use_masonry').masonry(); 8882 }; 8883 }); 8884 $('#vhsjs_hide_180_1707793551_7414422').hide(); 8885 $('#vhsjs_view_180_1707793551_7414422').show();" 8886 style="display: none" 8887 ><i class="fa fa-caret-down"></i> 8888 <span class="hover_link">Abstract</span></a 8889 > 8890 <div 8891 data-display-control="180_1707793551_7414422" 8892 id="179_1707793551_7414346" 8893 style="display: none" 8894 > 8895 <div class="arrow-slidedown"> 8896 <blockquote> 8897 This tutorial aims to introduce Event Graphs 8898 (EGs), invented 40 years ago by Lee Schruben to 8899 allow event-based modeling of discrete dynamic 8900 systems. Their simplicity and naturalness in 8901 causality modelling and simulation modelling 8902 made EGs popular in research and practice. In a 8903 simulation, an event causes state changes in a 8904 system as well as other events to happen in the 8905 future. EGs provide a parsimonious diagram 8906 representation for the Event Scheduling paradigm 8907 of Discrete Event Simulation. We first introduce 8908 their visual syntax and informal semantics, and 8909 then present a recent extension by adding 8910 objects to EGs. Our tutorial also includes an 8911 introduction to the formal semantics of EGs and 8912 a Python implementation for executing EGs. 8913 </blockquote> 8914 </div> 8915 </div> 8916 </div> 8917 </div> 8918 <div class="slot-urls"></div> 8919 <a href="/wsc23papers/121.pdf" target="_blank">pdf</a 8920 ><br /> 8921 </div> 8922 </div> 8923 <div class="session-entry">
8924 <span class="session-event-type">Tutorial</span 8925 ><span class="type-track-spacer"> · </span 8926 ><span class="program-track">Introductory Tutorials</span 8927 ><br /> 8928 <div class="session-title"> 8929 Simulation-Driven Digital Twins: The DNA of Resilient 8930 Supply Chains 8931 </div> 8932 <div class="session-chair"> 8933 Chair: David T. Sturrock (Simio LLC)<br /> 8934 </div> 8935 <div class="slot-entry"> 8936 <a name="inv202" tabindex="-1"></a> 8937 <div class="slot-authors"> 8938 Stephan Biller (Purdue University) and Paul Venditti, 8939 Jinxin Yi, Xi Jiang, and Bahar Biller (SAS Institute, 8940 Inc) 8941 </div> 8942 <div class="slot-abstract"> 8943 <div> 8944 <a 8945 class="clickable no-decoration" 8946 id="vhsjs_view_182_1707793551_7472937" 8947 onclick="$('#vhsjs_view_182_1707793551_7472937').hide(); 8948 $('#vhsjs_hide_182_1707793551_7472937').show(); 8949 $('#181_1707793551_7472847').slideDown(function() { 8950 if (typeof Masonry === 'function') { 8951 $('.use_masonry').masonry(); 8952 }; 8953 8954 });" 8955 ><i class="fa fa-caret-right"></i> 8956 <span class="hover_link">Abstract</span></a 8957 ><a 8958 class="clickable no-decoration" 8959 id="vhsjs_hide_182_1707793551_7472937" 8960 onclick="$('#181_1707793551_7472847').hide(function() { 8961 if (typeof Masonry === 'function') { 8962 $('.use_masonry').masonry(); 8963 }; 8964 }); 8965 $('#vhsjs_hide_182_1707793551_7472937').hide(); 8966 $('#vhsjs_view_182_1707793551_7472937').show();" 8967 style="display: none" 8968 ><i class="fa fa-caret-down"></i> 8969 <span class="hover_link">Abstract</span></a 8970 > 8971 <div 8972 data-display-control="182_1707793551_7472937" 8973 id="181_1707793551_7472847" 8974 style="display: none" 8975 > 8976 <div class="arrow-slidedown"> 8977 <blockquote> 8978 This tutorial defines what a digital twin is and 8979 outlines its four required characteristics. 8980 Digital twins are developed to derive insights 8981 to control entities and processes in the digital 8982 world with simulation as one of the key 8983 technologies lying at the heart of this 8984 development. The resulting insights are used to 8985 prescribe actions in the physical world to fix 8986 future problems before they happen. This 8987 tutorial describes the key digital twin 8988 development functions together with the digital 8989 twin enabling technologies with focus on the use 8990 of simulation for process twin development. The 8991 corresponding functions and technologies are 8992 displayed on several different digital twin 8993 development frameworks with the potential to 8994 serve as guides for practitioners interested in 8995 developing digital twin solutions. We conclude 8996 with an example of a supply chain digital twin 8997 use case and the role of simulation and AI in 8998 the twin development. 8999 </blockquote> 9000 </div> 9001 </div> 9002 </div> 9003 </div> 9004 <div class="slot-urls"></div> 9005 <a href="/wsc23papers/122.pdf" target="_blank">pdf</a 9006 ><br /> 9007 </div> 9008 </div> 9009 <div class="session-entry">
9010 <span class="session-event-type">Tutorial</span 9011 ><span class="type-track-spacer"> · </span 9012 ><span class="program-track">Introductory Tutorials</span 9013 ><br /> 9014 <div class="session-title"> 9015 Tested Success Tips for Simulation Project Excellence 9016 </div> 9017 <div class="session-chair"> 9018 Chair: Björn Johansson (Chalmers University of 9019 Technology)<br /> 9020 </div> 9021 <div class="slot-entry"> 9022 <a name="con103" tabindex="-1"></a> 9023 <div class="slot-authors"> 9024 David T. Sturrock (Simio LLC) 9025 </div> 9026 <div class="slot-abstract"> 9027 <div> 9028 <a 9029 class="clickable no-decoration" 9030 id="vhsjs_view_184_1707793551_7532747" 9031 onclick="$('#vhsjs_view_184_1707793551_7532747').hide(); 9032 $('#vhsjs_hide_184_1707793551_7532747').show(); 9033 $('#183_1707793551_7532663').slideDown(function() { 9034 if (typeof Masonry === 'function') { 9035 $('.use_masonry').masonry(); 9036 }; 9037 9038 });" 9039 ><i class="fa fa-caret-right"></i> 9040 <span class="hover_link">Abstract</span></a 9041 ><a 9042 class="clickable no-decoration" 9043 id="vhsjs_hide_184_1707793551_7532747" 9044 onclick="$('#183_1707793551_7532663').hide(function() { 9045 if (typeof Masonry === 'function') { 9046 $('.use_masonry').masonry(); 9047 }; 9048 }); 9049 $('#vhsjs_hide_184_1707793551_7532747').hide(); 9050 $('#vhsjs_view_184_1707793551_7532747').show();" 9051 style="display: none" 9052 ><i class="fa fa-caret-down"></i> 9053 <span class="hover_link">Abstract</span></a 9054 > 9055 <div
9056 data-display-control="184_1707793551_7532747" 9057 id="183_1707793551_7532663" 9058 style="display: none" 9059 > 9060 <div class="arrow-slidedown"> 9061 <blockquote> 9062 How can you make your projects successful? 9063 Modeling can certainly be fun, but it can also 9064 be quite challenging. With the new demands of 9065 Smart Factories, Digital Twins, and Digital 9066 Transformation, the challenges multiply. You 9067 want your first and every project to be 9068 successful, so you can justify continued work. 9069 Unfortunately, a simulation project is much more 9070 than simply building a model -- the skills 9071 required for success go well beyond knowing a 9072 particular simulation tool. 9073 </blockquote> 9074 </div> 9075 </div> 9076 </div> 9077 </div> 9078 <div class="slot-urls"></div> 9079 <a href="/wsc23papers/123.pdf" target="_blank">pdf</a 9080 ><br /> 9081 </div> 9082 </div> 9083 <div class="session-entry"> 9084 <span class="session-event-type">Tutorial</span 9085 ><span class="type-track-spacer"> · </span 9086 ><span class="program-track">Introductory Tutorials</span 9087 ><br /> 9088 <div class="session-title"> 9089 Design and Analysis of Simulation Experiments Using Three 9090 Simple Statistical Formulas 9091 </div> 9092 <div class="session-chair"> 9093 Chair: Sanjay Jain (The George Washington University)<br /> 9094 </div> 9095 <div class="slot-entry"> 9096 <a name="inv119" tabindex="-1"></a> 9097 <div class="slot-authors"> 9098 Averill Law (Averill M. Law & Associates, Inc.) 9099 </div> 9100 <div class="slot-abstract"> 9101 <div> 9102 <a 9103 class="clickable no-decoration" 9104 id="vhsjs_view_186_1707793551_758164" 9105 onclick="$('#vhsjs_view_186_1707793551_758164').hide(); 9106 $('#vhsjs_hide_186_1707793551_758164').show(); 9107 $('#185_1707793551_758146').slideDown(function() { 9108 if (typeof Masonry === 'function') { 9109 $('.use_masonry').masonry(); 9110 }; 9111 9112 });" 9113 ><i class="fa fa-caret-right"></i> 9114 <span class="hover_link">Abstract</span></a 9115 ><a 9116 class="clickable no-decoration" 9117 id="vhsjs_hide_186_1707793551_758164" 9118 onclick="$('#185_1707793551_758146').hide(function() { 9119 if (typeof Masonry === 'function') { 9120 $('.use_masonry').masonry(); 9121 }; 9122 }); 9123 $('#vhsjs_hide_186_1707793551_758164').hide(); 9124 $('#vhsjs_view_186_1707793551_758164').show();" 9125 style="display: none" 9126 ><i class="fa fa-caret-down"></i> 9127 <span class="hover_link">Abstract</span></a 9128 > 9129 <div 9130 data-display-control="186_1707793551_758164" 9131 id="185_1707793551_758146" 9132 style="display: none" 9133 > 9134 <div class="arrow-slidedown"> 9135 <blockquote> 9136 Output-data analysis is arguably the 9137 most-researched topic in the field of simulation 9138 modeling, with more than 1000 technical papers 9139 having been written. However, many of the 9140 published papers are highly mathematical in 9141 nature, making them difficult to understand for 9142 many simulation practitioners. In this tutorial, 9143 we discuss the replication and 9144 replication/deletion approaches which can 9145 address most analysis problems using three 9146 simple formulas (or expressions) from a first 9147 undergraduate statistics course. Although the 9148 replication approaches discussed above are 9149 widely used for estimating the mean of a single 9150 simulated system, we show that the same three 9151 formulas can also be used to compare any number 9152 of simulated systems, to handle multiple system 9153 performance measures simultaneously, and also to 9154 estimate performance measures such as 9155 probabilities and percentiles rather than just 9156 means. We also discuss a relatively simple 9157 graphical methodology for determining a warmup 9158 period if steady-state characteristics are of 9159 interest. 9160 </blockquote> 9161 </div> 9162 </div> 9163 </div> 9164 </div> 9165 <div class="slot-urls"></div> 9166 <a href="/wsc23papers/124.pdf" target="_blank">pdf</a 9167 ><br /> 9168 </div> 9169 </div> 9170 <div class="session-entry">
9171 <span class="session-event-type">Tutorial</span 9172 ><span class="type-track-spacer"> · </span 9173 ><span class="program-track">Introductory Tutorials</span 9174 ><br /> 9175 <div class="session-title"> 9176 Statistical Uncertainty Quantification for Expensive 9177 Black-Box Models: Methodologies and Input Uncertainty 9178 Applications 9179 </div> 9180 <div class="session-chair"> 9181 Chair: Chang-Han Rhee (Northwestern University)<br /> 9182 </div> 9183 <div class="slot-entry"> 9184 <a name="inv143" tabindex="-1"></a> 9185 <div class="slot-authors"> 9186 Henry Lam (Columbia University) 9187 </div> 9188 <div class="slot-abstract"> 9189 <div> 9190 <a 9191 class="clickable no-decoration" 9192 id="vhsjs_view_188_1707793551_7631223" 9193 onclick="$('#vhsjs_view_188_1707793551_7631223').hide(); 9194 $('#vhsjs_hide_188_1707793551_7631223').show(); 9195 $('#187_1707793551_7631142').slideDown(function() { 9196 if (typeof Masonry === 'function') { 9197 $('.use_masonry').masonry(); 9198 }; 9199 9200 });" 9201 ><i class="fa fa-caret-right"></i> 9202 <span class="hover_link">Abstract</span></a 9203 ><a 9204 class="clickable no-decoration" 9205 id="vhsjs_hide_188_1707793551_7631223" 9206 onclick="$('#187_1707793551_7631142').hide(function() { 9207 if (typeof Masonry === 'function') { 9208 $('.use_masonry').masonry(); 9209 }; 9210 }); 9211 $('#vhsjs_hide_188_1707793551_7631223').hide(); 9212 $('#vhsjs_view_188_1707793551_7631223').show();" 9213 style="display: none" 9214 ><i class="fa fa-caret-down"></i> 9215 <span class="hover_link">Abstract</span></a 9216 > 9217 <div 9218 data-display-control="188_1707793551_7631223" 9219 id="187_1707793551_7631142" 9220 style="display: none" 9221 > 9222 <div class="arrow-slidedown"> 9223 <blockquote> 9224 This tutorial reviews methodologies for 9225 quantifying statistical uncertainty in 9226 computationally expensive black-box models, 9227 which arise frequently in data-driven simulation 9228 analyses under input uncertainty. When facing 9229 these models, it can be difficult to run 9230 repeated evaluations due to computation cost, 9231 and also to obtain auxiliary information such as 9232 gradients due to analytical intractability, thus 9233 rendering many traditional statistical 9234 approaches challenging to apply. We describe 9235 several lines of approaches to resolve these 9236 challenges, including data-splitting methods 9237 based on batching variants, a recent so-called 9238 cheap bootstrap approach, and subsampling 9239 schemes. We discuss the applications of these 9240 approaches to simulation, including problems 9241 suffering from both aleatory error exhibited via 9242 Monte Carlo noises and epistemic error stemming 9243 from the input uncertainty. 9244 </blockquote> 9245 </div> 9246 </div> 9247 </div> 9248 </div> 9249 <div class="slot-urls"></div> 9250 <a href="/wsc23papers/125.pdf" target="_blank">pdf</a 9251 ><br /> 9252 </div> 9253 </div> 9254 <div class="session-entry">
9255 <span class="session-event-type">Tutorial</span 9256 ><span class="type-track-spacer"> · </span 9257 ><span class="program-track">Introductory Tutorials</span 9258 ><br /> 9259 <div class="session-title"> 9260 Tutorial: Basics of Metamodeling 9261 </div> 9262 <div class="session-chair"> 9263 Chair: Paulo Victor Freitas Lopes (Chalmers University of 9264 Technology, Aeronautics Institute of Technology)<br /> 9265 </div> 9266 <div class="slot-entry"> 9267 <a name="inv204" tabindex="-1"></a> 9268 <div class="slot-authors"> 9269 Russell Barton (The Pennsylvania State University) 9270 </div> 9271 <div class="slot-abstract"> 9272 <div> 9273 <a 9274 class="clickable no-decoration" 9275 id="vhsjs_view_190_1707793551_7673018" 9276 onclick="$('#vhsjs_view_190_1707793551_7673018').hide(); 9277 $('#vhsjs_hide_190_1707793551_7673018').show(); 9278 $('#189_1707793551_767294').slideDown(function() { 9279 if (typeof Masonry === 'function') { 9280 $('.use_masonry').masonry(); 9281 }; 9282 9283 });" 9284 ><i class="fa fa-caret-right"></i> 9285 <span class="hover_link">Abstract</span></a 9286 ><a 9287 class="clickable no-decoration" 9288 id="vhsjs_hide_190_1707793551_7673018" 9289 onclick="$('#189_1707793551_767294').hide(function() { 9290 if (typeof Masonry === 'function') { 9291 $('.use_masonry').masonry(); 9292 }; 9293 }); 9294 $('#vhsjs_hide_190_1707793551_7673018').hide(); 9295 $('#vhsjs_view_190_1707793551_7673018').show();" 9296 style="display: none" 9297 ><i class="fa fa-caret-down"></i> 9298 <span class="hover_link">Abstract</span></a 9299 > 9300 <div 9301 data-display-control="190_1707793551_7673018" 9302 id="189_1707793551_767294" 9303 style="display: none" 9304 > 9305 <div class="arrow-slidedown"> 9306 <blockquote> 9307 Metamodels are fast-to-compute mathematical 9308 models that are designed to mimic the 9309 input-output behavior of discrete-event or other 9310 complex simulation models. Linear regression 9311 metamodels have the longest history, but other 9312 model forms include Gaussian process regression 9313 and neural networks. This introductory tutorial 9314 highlights basic issues in choosing a metamodel 9315 type and specific form, and making simulation 9316 runs to fit the metamodel. The tutorial ends 9317 with a warning on potential pitfalls, and 9318 suggestions on further reading to expand your 9319 knowledge of metamodeling. 9320 </blockquote> 9321 </div> 9322 </div> 9323 </div> 9324 </div> 9325 <div class="slot-urls"></div> 9326 <a href="/wsc23papers/126.pdf" target="_blank">pdf</a 9327 ><br /> 9328 </div> 9329 </div> 9330 <div class="session-entry">
9331 <span class="session-event-type">Tutorial</span 9332 ><span class="type-track-spacer"> · </span 9333 ><span class="program-track">Introductory Tutorials</span 9334 ><br /> 9335 <div class="session-title"> 9336 An Introduction to Discrete-event Modeling and Simulation 9337 with DEVS 9338 </div> 9339 <div class="session-chair"> 9340 Chair: Russell R. Barton (Pennsylvania State 9341 University)<br /> 9342 </div> 9343 <div class="slot-entry"> 9344 <a name="inv123" tabindex="-1"></a> 9345 <div class="slot-title-line"> 9346 <span class="slot-title" 9347 >An Introduction to Discrete-Event Modeling and 9348 Simulation with DEVS</span 9349 > 9350 </div> 9351 <div class="slot-authors"> 9352 Yentl Van Tendeloo and Randy Paredis (University of 9353 Antwerp) and Hans Vangheluwe (University of Antwerp, 9354 Flanders Make) 9355 </div> 9356 <div class="slot-abstract"> 9357 <div> 9358 <a 9359 class="clickable no-decoration" 9360 id="vhsjs_view_192_1707793551_7727807" 9361 onclick="$('#vhsjs_view_192_1707793551_7727807').hide(); 9362 $('#vhsjs_hide_192_1707793551_7727807').show(); 9363 $('#191_1707793551_7727723').slideDown(function() { 9364 if (typeof Masonry === 'function') { 9365 $('.use_masonry').masonry(); 9366 }; 9367 9368 });" 9369 ><i class="fa fa-caret-right"></i> 9370 <span class="hover_link">Abstract</span></a 9371 ><a 9372 class="clickable no-decoration" 9373 id="vhsjs_hide_192_1707793551_7727807" 9374 onclick="$('#191_1707793551_7727723').hide(function() { 9375 if (typeof Masonry === 'function') { 9376 $('.use_masonry').masonry(); 9377 }; 9378 }); 9379 $('#vhsjs_hide_192_1707793551_7727807').hide(); 9380 $('#vhsjs_view_192_1707793551_7727807').show();" 9381 style="display: none" 9382 ><i class="fa fa-caret-down"></i> 9383 <span class="hover_link">Abstract</span></a 9384 > 9385 <div 9386 data-display-control="192_1707793551_7727807" 9387 id="191_1707793551_7727723" 9388 style="display: none" 9389 > 9390 <div class="arrow-slidedown"> 9391 <blockquote> 9392 The Discrete-Event System Specification (DEVS) 9393 is a formalism devised by Bernard Zeigler in the 9394 late 1970s for modeling complex dynamical 9395 systems using a discrete-event abstraction. At 9396 this abstraction level, a timed sequence of 9397 pertinent "events'' input to a system causes 9398 instantaneous changes to the state of the 9399 system. The main advantages of DEVS are its 9400 precise, implementation independent 9401 specification, and its support for modular, 9402 hierarchical composition. This tutorial 9403 introduces the Classic DEVS formalism in a 9404 bottom-up fashion, using a simple traffic light 9405 example. The syntax and operational semantics of 9406 Atomic (i.e., non-hierarchical) and of Coupled 9407 (i.e., hierarchical, connecting interacting 9408 components) models are introduced. Finally, a 9409 simplified DEVS model for performance analysis 9410 of vessel movements in the Port of Antwerp is 9411 presented. All examples in the paper use 9412 PythonPDEVS, though other DEVS tools could 9413 equally well be used. We conclude with 9414 suggestions for further reading on DEVS theory, 9415 variants, and tools. 9416 </blockquote> 9417 </div> 9418 </div> 9419 </div> 9420 </div> 9421 <div class="slot-urls"></div> 9422 <a href="/wsc23papers/127.pdf" target="_blank">pdf</a 9423 ><br /> 9424 </div> 9425 </div> 9426 </div> 9427 <div class="centered"> 9428 <div class="top-link"><a href="#top">Return to Top</a></div> 9429 </div> 9430 <hr /> 9431 </div> 9432 <div class="area-section"> 9433 <div class="centered"> 9434 <a name="ptrack116" tabindex="-1"></a> 9435 <div class="section-title">Healthcare and Life Sciences</div> 9436 </div> 9437 <div class="centered track-chair">
9438 <span class="track-chair-role" 9439 >Track Coordinator - Healthcare and Life Sciences: </span 9440 ><span class="track-chair-names" 9441 >Bjorn Berg (University of Minnesota), Masoud Fakhimi 9442 (University of Surrey), Tugce Martagan (Eindhoven University 9443 of Technology)</span 9444 > 9445 </div> 9446 <div class="section-entry"> 9447 <div class="session-entry"> 9448 <span class="session-event-type">Technical Session</span 9449 ><span class="type-track-spacer"> · </span 9450 ><span class="program-track" 9451 >Healthcare and Life Sciences</span 9452 ><br /> 9453 <div class="session-title"> 9454 Simulation Modeling for COVID I 9455 </div> 9456 <div class="session-chair"> 9457 Chair: Christine Currie (University of Southampton)<br /> 9458 </div> 9459 <div class="slot-entry"> 9460 <a name="con153" tabindex="-1"></a> 9461 <div class="slot-title-line"> 9462 <span class="slot-title" 9463 >Using Simulation to Study the Impact of Covid-19 9464 Policies on the Availability of Childcare</span 9465 > 9466 </div> 9467 <div class="slot-authors"> 9468 Adam Cahall, Jasmine Eng, Jane Gao, Ben Hilbert, and 9469 Jamol Pender (Cornell University) 9470 </div> 9471 <div class="slot-abstract"> 9472 <div> 9473 <a 9474 class="clickable no-decoration" 9475 id="vhsjs_view_194_1707793551_7793958" 9476 onclick="$('#vhsjs_view_194_1707793551_7793958').hide(); 9477 $('#vhsjs_hide_194_1707793551_7793958').show(); 9478 $('#193_1707793551_7793877').slideDown(function() { 9479 if (typeof Masonry === 'function') { 9480 $('.use_masonry').masonry(); 9481 }; 9482 9483 });" 9484 ><i class="fa fa-caret-right"></i> 9485 <span class="hover_link">Abstract</span></a 9486 ><a 9487 class="clickable no-decoration" 9488 id="vhsjs_hide_194_1707793551_7793958" 9489 onclick="$('#193_1707793551_7793877').hide(function() { 9490 if (typeof Masonry === 'function') { 9491 $('.use_masonry').masonry(); 9492 }; 9493 }); 9494 $('#vhsjs_hide_194_1707793551_7793958').hide(); 9495 $('#vhsjs_view_194_1707793551_7793958').show();" 9496 style="display: none" 9497 ><i class="fa fa-caret-down"></i> 9498 <span class="hover_link">Abstract</span></a 9499 > 9500 <div 9501 data-display-control="194_1707793551_7793958" 9502 id="193_1707793551_7793877" 9503 style="display: none" 9504 > 9505 <div class="arrow-slidedown"> 9506 <blockquote> 9507 The COVID-19 pandemic has had a profound impact 9508 on the lives of working parents, who are 9509 struggling to balance their responsibilities at 9510 work and at home, as well as childcare providers 9511 who are working hard to keep their doors open. 9512 In this paper, we examine the effect of 9513 childcare policies on the availability of 9514 childcare. Specifically, we investigate how 9515 classroom size, the likelihood of COVID-19 9516 infection, and the number of days a classroom 9517 may need to close affect the amount of time 9518 parents will need to stay at home with their 9519 children. Our results show that even low 9520 probabilities of infection combined with
9521 stringent policies can have a large impact on 9522 the duration of a child's exclusion from 9523 childcare services. 9524 </blockquote> 9525 </div> 9526 </div> 9527 </div> 9528 </div> 9529 <div class="slot-urls"></div> 9530 <a href="/wsc23papers/080.pdf" target="_blank">pdf</a 9531 ><br /> 9532 </div> 9533 <div class="slot-entry"> 9534 <a name="inv194" tabindex="-1"></a> 9535 <div class="slot-title-line"> 9536 <span class="slot-title" 9537 >Enhancing Pandemic Preparedness Using Mean Field and 9538 Simulation Modeling</span 9539 > 9540 </div> 9541 <div class="slot-authors"> 9542 Mohammad Dehghanimohammadabadi (Northeastern University) 9543 and Gökçe Dayanıklı (University of 9544 Illinois at Urbana-Champaign) 9545 </div> 9546 <div class="slot-abstract"> 9547 <div> 9548 <a 9549 class="clickable no-decoration" 9550 id="vhsjs_view_196_1707793551_781545" 9551 onclick="$('#vhsjs_view_196_1707793551_781545').hide(); 9552 $('#vhsjs_hide_196_1707793551_781545').show(); 9553 $('#195_1707793551_7815368').slideDown(function() { 9554 if (typeof Masonry === 'function') { 9555 $('.use_masonry').masonry(); 9556 }; 9557 9558 });" 9559 ><i class="fa fa-caret-right"></i> 9560 <span class="hover_link">Abstract</span></a 9561 ><a 9562 class="clickable no-decoration" 9563 id="vhsjs_hide_196_1707793551_781545" 9564 onclick="$('#195_1707793551_7815368').hide(function() { 9565 if (typeof Masonry === 'function') { 9566 $('.use_masonry').masonry(); 9567 }; 9568 }); 9569 $('#vhsjs_hide_196_1707793551_781545').hide(); 9570 $('#vhsjs_view_196_1707793551_781545').show();" 9571 style="display: none" 9572 ><i class="fa fa-caret-down"></i> 9573 <span class="hover_link">Abstract</span></a 9574 > 9575 <div 9576 data-display-control="196_1707793551_781545" 9577 id="195_1707793551_7815368" 9578 style="display: none" 9579 > 9580 <div class="arrow-slidedown"> 9581 <blockquote> 9582 The COVID-19 pandemic has emphasized the 9583 importance of preparedness and response plans 9584 for healthcare providers and rational responses 9585 from society to effectively manage infectious 9586 disease outbreaks. Strategic guidelines should 9587 be created to ensure the availability of 9588 required resources while considering the 9589 rational response of individuals under different 9590 policy scenarios. This study uses a 9591 simulation-optimization-game theory approach to 9592 first determine the daily number of infected 9593 people in response to social distancing policies 9594 in a game theoretical setup. Second, this daily 9595 number of infected people is used in a 9596 simulation to determine an optimal replenishment 9597 policy for restocking personal protective 9598 equipment (PPE) items. The model incorporates a 9599 combination of mean field games modeling and a 9600 simulation model in Simio to perform 9601 optimization tasks. This approach aims to 9602 guarantee the availability of required resources 9603 by taking into account the rational response of 9604 individuals under different policy sce
9604narios. 9605 </blockquote> 9606 </div> 9607 </div> 9608 </div> 9609 </div> 9610 <div class="slot-urls"></div> 9611 <a href="/wsc23papers/081.pdf" target="_blank">pdf</a 9612 ><br /> 9613 </div> 9614 <div class="slot-entry"> 9615 <a name="con259" tabindex="-1"></a> 9616 <div class="slot-title-line"> 9617 <span class="slot-title" 9618 >Equitable Allocation of Scarce Resources during the 9619 COVID-19 Pandemic: A Case Study for Convalescent 9620 Plasma Distribution</span 9621 > 9622 </div> 9623 <div class="slot-authors"> 9624 Jasdeep Singh Dhahan and Alexander Rutherford (Simon 9625 Fraser University), Andrew Shih (University of British 9626 Columbia), Na Li (University of Calgary), and Douglas 9627 Down (McMaster University) 9628 </div> 9629 <div class="slot-abstract"> 9630 <div> 9631 <a 9632 class="clickable no-decoration" 9633 id="vhsjs_view_198_1707793551_784038" 9634 onclick="$('#vhsjs_view_198_1707793551_784038').hide(); 9635 $('#vhsjs_hide_198_1707793551_784038').show(); 9636 $('#197_1707793551_7840302').slideDown(function() { 9637 if (typeof Masonry === 'function') { 9638 $('.use_masonry').masonry(); 9639 }; 9640 9641 });" 9642 ><i class="fa fa-caret-right"></i> 9643 <span class="hover_link">Abstract</span></a 9644 ><a 9645 class="clickable no-decoration" 9646 id="vhsjs_hide_198_1707793551_784038" 9647 onclick="$('#197_1707793551_7840302').hide(function() { 9648 if (typeof Masonry === 'function') { 9649 $('.use_masonry').masonry(); 9650 }; 9651 }); 9652 $('#vhsjs_hide_198_1707793551_784038').hide(); 9653 $('#vhsjs_view_198_1707793551_784038').show();" 9654 style="display: none" 9655 ><i class="fa fa-caret-down"></i> 9656 <span class="hover_link">Abstract</span></a 9657 > 9658 <div 9659 data-display-control="198_1707793551_784038" 9660 id="197_1707793551_7840302" 9661 style="display: none" 9662 > 9663 <div class="arrow-slidedown"> 9664 <blockquote> 9665 Resource planning during pandemics presents many 9666 challenges and equitable decisions about 9667 resource allocation must be made. There is no 9668 standard definition of equity. Robust 9669 mathematical formulations can require a lot of 9670 data. In a novel pandemic there is limited 9671 historical information available to inform 9672 decisions. Decision makers can look to define 9673 equity through population proportions 9674 (pro-rata). This notion of equity is readily 9675 implementable. We present a practical framework 9676 for an equitable allocation of scarce resources 9677 using population proportions, disease 9678 demographics, and resource utilization. We 9679 assess our framework using a stochastic 9680 simulation model, calibrated to COVID-19 case
9681 data, in a case study for convalescent plasma 9682 distribution in the context of the clinical 9683 trial CONCOR-1. We show that pro-rata resource 9684 allocation can be inequitable and that decision 9685 makers can consider readily available 9686 information, such as resource utilization and 9687 case data, to inform equity and proactively 9688 manage scarce resources during a pandemic. 9689 </blockquote> 9690 </div> 9691 </div> 9692 </div> 9693 </div> 9694 <div class="slot-urls"></div> 9695 <a href="/wsc23papers/082.pdf" target="_blank">pdf</a 9696 ><br /> 9697 </div> 9698 </div> 9699 <div class="session-entry"> 9700 <span class="session-event-type">Technical Session</span 9701 ><span class="type-track-spacer"> · </span 9702 ><span class="program-track" 9703 >Healthcare and Life Sciences</span 9704 ><br /> 9705 <div class="session-title"> 9706 Simulation Modeling for COVID II 9707 </div> 9708 <div class="session-chair"> 9709 Chair: Yuming Sun (Georgia Institute of Technology)<br /> 9710 </div> 9711 <div class="slot-entry"> 9712 <a name="con253" tabindex="-1"></a> 9713 <div class="slot-title-line"> 9714 <span class="slot-title" 9715 >A Multi-Team Multi-Model Collaborative COVID-19 9716 Forecasting Hub for India</span 9717 > 9718 </div> 9719 <div class="slot-authors"> 9720 Aniruddha Adiga (University of Virginia); Siva Athreya 9721 (International Centre for Theoretical Sciences-TIFR, 9722 Indian Statistical Institute); Kantha Rao Bhimala (CSIR 9723 Fourth Paradigm Institute); Ambedkar Dukkipati and Tony 9724 Gracious (Indian Institute of Science); Shubham Gupta 9725 (IBM Research Europe); Benjamin Hurt, Gursharn Kaur, 9726 Bryan Lewis, and Madhav Marathe (University of 9727 Virginia); Vidyadhar Mudkavi and Gopal Krishna Patra 9728 (CSIR Fourth Paradigm Institute); Przemyslaw Porebski 9729 (University of Virginia); Nihesh Rathod and Rajesh 9730 Sundaresan (Indian Institute of Science); Srinivasan 9731 Venkataramanan (University of Virginia); and Sarath 9732 Yasodharan (Indian Institute of Science) 9733 </div> 9734 <div class="slot-abstract"> 9735 <div> 9736 <a 9737 class="clickable no-decoration" 9738 id="vhsjs_view_200_1707793551_7895677" 9739 onclick="$('#vhsjs_view_200_1707793551_7895677').hide(); 9740 $('#vhsjs_hide_200_1707793551_7895677').show(); 9741 $('#199_1707793551_78956').slideDown(function() { 9742 if (typeof Masonry === 'function') { 9743 $('.use_masonry').masonry(); 9744 }; 9745 9746 });" 9747 ><i class="fa fa-caret-right"></i> 9748 <span class="hover_link">Abstract</span></a 9749 ><a 9750 class="clickable no-decoration" 9751 id="vhsjs_hide_200_1707793551_7895677" 9752 onclick="$('#199_1707793551_78956').hide(function() { 9753 if (typeof Masonry === 'function') { 9754 $('.use_masonry').masonry(); 9755 }; 9756 }); 9757 $('#vhsjs_hide_200_1707793551_7895677').hide(); 9758 $('#vhsjs_view_200_1707793551_7895677').show();" 9759 style="display: none" 9760 ><i class="fa fa-caret-down"></i>
9761 <span class="hover_link">Abstract</span></a 9762 > 9763 <div 9764 data-display-control="200_1707793551_7895677" 9765 id="199_1707793551_78956" 9766 style="display: none" 9767 > 9768 <div class="arrow-slidedown"> 9769 <blockquote> 9770 During the COVID-19 pandemic, India has seen 9771 some of the highest number of cases and deaths. 9772 Quality of data, continuously changing policy, 9773 and public health response made forecasting 9774 extremely difficult. Given the challenges in 9775 real-time forecasting, several countries had 9776 started a multi-team collaborative effort. 9777 Inspired by these works, academic partners from 9778 India and the United States setup a repository 9779 for aggregating India-specific forecasts from 9780 multiple teams. In this paper, we describe the 9781 effort and the challenges in setting up the 9782 repository. We discuss the development of 9783 simulations of compartmental models to model 9784 specific waves of the pandemic and show that the 9785 simulation model designed specifically for the 9786 Omicron wave was able to predict the onset and 9787 peak sizes accurately. We employed a 9788 median-based ensemble model to aggregate the 9789 individual forecasts. We observed that 9790 median-based ensemble was relatively stable 9791 compared to the constituent models and was one 9792 of better performing models. 9793 </blockquote> 9794 </div> 9795 </div> 9796 </div> 9797 </div> 9798 <div class="slot-urls"></div> 9799 <a href="/wsc23papers/083.pdf" target="_blank">pdf</a 9800 ><br /> 9801 </div> 9802 <div class="slot-entry"> 9803 <a name="inv166" tabindex="-1"></a> 9804 <div class="slot-title-line"> 9805 <span class="slot-title" 9806 >Multi-criteria Simulation Optimization for COVID-19 9807 Testing in Schools</span 9808 > 9809 </div> 9810 <div class="slot-authors"> 9811 Yiwei Zhang, Maria Mayorga, Julie Ivy, and Julie Swann 9812 (North Carolina State University) 9813 </div> 9814 <div class="slot-abstract"> 9815 <div> 9816 <a 9817 class="clickable no-decoration" 9818 id="vhsjs_view_202_1707793551_7919667" 9819 onclick="$('#vhsjs_view_202_1707793551_7919667').hide(); 9820 $('#vhsjs_hide_202_1707793551_7919667').show(); 9821 $('#201_1707793551_7919586').slideDown(function() { 9822 if (typeof Masonry === 'function') { 9823 $('.use_masonry').masonry(); 9824 }; 9825 9826 });" 9827 ><i class="fa fa-caret-right"></i> 9828 <span class="hover_link">Abstract</span></a 9829 ><a 9830 class="clickable no-decoration" 9831 id="vhsjs_hide_202_1707793551_7919667" 9832 onclick="$('#201_1707793551_7919586').hide(function() { 9833 if (typeof Masonry === 'function') { 9834 $('.use_masonry').masonry(); 9835 }; 9836 }); 9837 $('#vhsjs_hide_202_1707793551_7919667').hide(); 9838 $('#vhsjs_view_202_1707793551_7919667').show();" 9839 style="display: none" 9840 ><i class="fa fa-caret-down"></i>
9841 <span class="hover_link">Abstract</span></a 9842 > 9843 <div 9844 data-display-control="202_1707793551_7919667" 9845 id="201_1707793551_7919586" 9846 style="display: none" 9847 > 9848 <div class="arrow-slidedown"> 9849 <blockquote> 9850 Evidence has shown that random screening tests 9851 are effective in reducing COVID-19 infections in 9852 schools. However, test administration may be 9853 hindered due to a limited budget or low 9854 participation caused by pandemic fatigue. Thus, 9855 we seek to balance the number of tests 9856 administered with end-of-semester infections. To 9857 do this we use an SEIR model to simulate 9858 SARS-CoV-2 transmissions within K-12 schools, 9859 design a multi-objective simulation optimization 9860 problem, and tune an NSGA-II algorithm to find 9861 the best testing schedules. We find the Pareto 9862 front of optimal schedules of screening tests, 9863 which can be used by stakeholders to inform test 9864 administration strategies. We discuss insights 9865 about the characteristics of optimal strategies, 9866 for example, when there are limited number of 9867 tests available or a desire to use few tests, 9868 the optimal plan is to perform the tests earlier 9869 in the semester and at higher intensity. 9870 </blockquote> 9871 </div> 9872 </div> 9873 </div> 9874 </div> 9875 <div class="slot-urls"></div> 9876 <a href="/wsc23papers/084.pdf" target="_blank">pdf</a 9877 ><br /> 9878 </div> 9879 <div class="slot-entry"> 9880 <a name="cea152" tabindex="-1"></a> 9881 <div class="slot-title-line"> 9882 <span class="slot-title" 9883 >Endogenous Human Behavior in Models of COVID-19 9884 Transmission: A Systematic Scoping Review</span 9885 > 9886 </div> 9887 <div class="slot-authors"> 9888 Alisa Hamilton (Johns Hopkins University, Center for 9889 Systems Science and Engineering; One Health Trust); 9890 Fardad Haghpanah, Sasha Tulchinsky, Nodar Kipshidze, and 9891 Suprena Poleon (One Health Trust); Gary Lin (Johns 9892 Hopkins Applied Physics Laboratory, One Health Trust); 9893 Hongru Du and Lauren Gardner (Johns Hopkins University, 9894 Center for Systems Science and Engineering); and Eili 9895 Klein (One Health Trust; Johns Hopkins University, 9896 Department of Emergency Medicine) 9897 </div> 9898 <div class="slot-abstract"> 9899 <div> 9900 <a 9901 class="clickable no-decoration" 9902 id="vhsjs_view_204_1707793551_7943113" 9903 onclick="$('#vhsjs_view_204_1707793551_7943113').hide(); 9904 $('#vhsjs_hide_204_1707793551_7943113').show(); 9905 $('#203_1707793551_794303').slideDown(function() { 9906 if (typeof Masonry === 'function') { 9907 $('.use_masonry').masonry(); 9908 }; 9909 9910 });" 9911 ><i class="fa fa-caret-right"></i>
9912 <span class="hover_link">Abstract</span></a 9913 ><a 9914 class="clickable no-decoration" 9915 id="vhsjs_hide_204_1707793551_7943113" 9916 onclick="$('#203_1707793551_794303').hide(function() { 9917 if (typeof Masonry === 'function') { 9918 $('.use_masonry').masonry(); 9919 }; 9920 }); 9921 $('#vhsjs_hide_204_1707793551_7943113').hide(); 9922 $('#vhsjs_view_204_1707793551_7943113').show();" 9923 style="display: none" 9924 ><i class="fa fa-caret-down"></i> 9925 <span class="hover_link">Abstract</span></a 9926 > 9927 <div 9928 data-display-control="204_1707793551_7943113" 9929 id="203_1707793551_794303" 9930 style="display: none" 9931 > 9932 <div class="arrow-slidedown"> 9933 <blockquote> 9934 While mathematical models of disease have been 9935 important drivers of public policy since the 9936 eighteenth century, the incorporation of 9937 endogenous behavior driven by risk perception is 9938 a relatively recent phenomenon (Klein et al., 9939 2007). Models incorporating behavior as 9940 endogenous variables may enhance their 9941 usefulness by providing an explicit mechanism 9942 for how behavior varies in response to public 9943 health measures and epidemic dynamics, resulting 9944 in a more nuanced understanding of disease 9945 transmission. We conducted a systematic scoping 9946 review to understand the extent to which 9947 endogenous behavior was incorporated into models 9948 of COVID-19 transmission. 9949 </blockquote> 9950 </div> 9951 </div> 9952 </div> 9953 </div> 9954 <div class="slot-urls"></div> 9955 <a href="/wsc23papers/cea152.pdf" target="_blank">pdf</a 9956 ><br /> 9957 </div> 9958 </div> 9959 <div class="session-entry"> 9960 <span class="session-event-type">Technical Session</span 9961 ><span class="type-track-spacer"> · </span 9962 ><span class="program-track" 9963 >Healthcare and Life Sciences</span 9964 ><br /> 9965 <div class="session-title"> 9966 Improving Emergency Department Efficiency Using Simulation 9967 </div> 9968 <div class="session-chair"> 9969 Chair: Vishnunarayan Girishan Prabhu (University of North 9970 Carolina at Charlotte)<br /> 9971 </div> 9972 <div class="slot-entry"> 9973 <a name="con205" tabindex="-1"></a> 9974 <div class="slot-title-line"> 9975 <span class="slot-title" 9976 >Measuring Emergency Department Resilience to Demand 9977 Surge: A Discrete-Event Simulation Framework</span 9978 > 9979 </div> 9980 <div class="slot-authors"> 9981 Eman Ouda, Andrei Sleptchenko, and Mecit Can Emre 9982 Simsekler (Khalifa University) and Ghada R. El-Eid 9983 (Sheikh Shakhbout Medical City) 9984 </div> 9985 <div class="slot-abstract"> 9986 <div> 9987 <a 9988 class="clickable no-decoration" 9989 id="vhsjs_view_206_1707793551_7989888" 9990 onclick="$('#vhsjs_view_206_1707793551_7989888').hide(); 9991 $('#vhsjs_hide_206_1707793551_7989888').show(); 9992 $('#205_1707793551_7989802').slideDown(function() { 9993 if (typeof Masonry === 'function') { 9994 $('.use_masonry').masonry(); 9995 }; 9996 9997 });" 9998 ><i class="fa fa-caret-right"></i>
9999 <span class="hover_link">Abstract</span></a 10000 ><a 10001 class="clickable no-decoration" 10002 id="vhsjs_hide_206_1707793551_7989888" 10003 onclick="$('#205_1707793551_7989802').hide(function() { 10004 if (typeof Masonry === 'function') { 10005 $('.use_masonry').masonry(); 10006 }; 10007 }); 10008 $('#vhsjs_hide_206_1707793551_7989888').hide(); 10009 $('#vhsjs_view_206_1707793551_7989888').show();" 10010 style="display: none" 10011 ><i class="fa fa-caret-down"></i> 10012 <span class="hover_link">Abstract</span></a 10013 > 10014 <div 10015 data-display-control="206_1707793551_7989888" 10016 id="205_1707793551_7989802" 10017 style="display: none" 10018 > 10019 <div class="arrow-slidedown"> 10020 <blockquote> 10021 This research explores the resilience components 10022 in emergency departments (EDs) during surges 10023 through discrete-event simulation (DES). By 10024 focusing on the resistance and recoverability 10025 components, the resilience of the ED is 10026 analyzed, as well as the flow of the patient and 10027 the resources required at each step. A 10028 simulation is developed to model an ED in the 10029 UAE and validated through collected timestamps. 10030 The results demonstrate the ordinary conditions 10031 of the ED and its calculated resilience, 10032 recoverability, and resistance, as well as its 10033 strength under conditions of surge demand. To 10034 investigate the impact of resources on the 10035 ED’s resilience, the resilience triangle 10036 is analyzed, and different interventions are 10037 applied by adding physicians, nurses, and beds 10038 and their effects. The methodology and 10039 simulation model provides significant insights 10040 to ED managers to evaluate and improve their 10041 department’s resilience during surges and 10042 emergencies. 10043 </blockquote> 10044 </div> 10045 </div> 10046 </div> 10047 </div> 10048 <div class="slot-urls"></div> 10049 <a href="/wsc23papers/085.pdf" target="_blank">pdf</a 10050 ><br /> 10051 </div> 10052 <div class="slot-entry"> 10053 <a name="con320" tabindex="-1"></a> 10054 <div class="slot-title-line"> 10055 <span class="slot-title" 10056 >Analysis of the Resilience of an Emergency 10057 Department: the Case of Accident with Multiple 10058 Victims</span 10059 > 10060 </div> 10061 <div class="slot-authors"> 10062 Mariela Ester Rodriguez (National University of Jujuy); 10063 Francesc Boixader (Computer Science School, Autonomous 10064 University of Barcelona); Francisco Epelde (Consultant 10065 Internal Medicine, Autonomous University of Barcelona); 10066 Alvaro Wong (Autonomous University of Barcelona); Eva 10067 Bruballa (Computer Science School, Autonomous University 10068 of Barcelona); Armando De Giusti (National University of 10069 La Plata); and Dolores Rexach and Emilio Luque 10070 (Autonomous University of Barcelona) 10071 </div> 10072 <div class="slot-abstract"> 10073 <div> 10074 <a 10075 class="clickable no-decoration" 10076 id="vhsjs_view_208_1707793551_8016827" 10077 onclick="$('#vhsjs_view_208_1707793551_8016827').hide(); 10078 $('#vhsjs_hide_208_1707793551_8016827').show(); 10079 $('#207_1707793551_8016746').slideDown(function() { 10080 if (typeof Masonry === 'function') { 10081 $('.use_masonry').masonry(); 10082 }; 10083 10084 });" 10085 ><i class="fa fa-caret-right"></i>
10086 <span class="hover_link">Abstract</span></a 10087 ><a 10088 class="clickable no-decoration" 10089 id="vhsjs_hide_208_1707793551_8016827" 10090 onclick="$('#207_1707793551_8016746').hide(function() { 10091 if (typeof Masonry === 'function') { 10092 $('.use_masonry').masonry(); 10093 }; 10094 }); 10095 $('#vhsjs_hide_208_1707793551_8016827').hide(); 10096 $('#vhsjs_view_208_1707793551_8016827').show();" 10097 style="display: none" 10098 ><i class="fa fa-caret-down"></i> 10099 <span class="hover_link">Abstract</span></a 10100 > 10101 <div 10102 data-display-control="208_1707793551_8016827" 10103 id="207_1707793551_8016746" 10104 style="display: none" 10105 > 10106 <div class="arrow-slidedown"> 10107 <blockquote> 10108 The care of multiple victims such as natural 10109 disasters in an Emergency Department is 10110 critical. This differs from ordinary care by the 10111 number of patients that arrive, their severity 10112 and the insufficient staff for these events. 10113 Designing and simulating this real life scenario 10114 will be useful for disaster management decision 10115 makers. The objective of this simulation is to 10116 model a system with resilience to critical 10117 situations. To model the input of this research, 10118 we worked with the percentage of patients 10119 received by the Cauquenes Hospital during the 10120 Chilean Earthquake February 27, 2010. A 10121 comparison of two situations is made: the 10122 admission of patients before an earthquake with 10123 a normal daily attention versus the admission of 10124 patients before an earthquake and the activation 10125 of the relief chain. The latter situation allows 10126 the system to be resilient and adapt quickly to 10127 its new reality. 10128 </blockquote> 10129 </div> 10130 </div> 10131 </div> 10132 </div> 10133 <div class="slot-urls"></div> 10134 <a href="/wsc23papers/086.pdf" target="_blank">pdf</a 10135 ><br /> 10136 </div> 10137 <div class="slot-entry"> 10138 <a name="inv176" tabindex="-1"></a> 10139 <div class="slot-title-line"> 10140 <span class="slot-title" 10141 >A Generalized Symbiotic Simulation Model of an 10142 Emergency Department for Real-Time Operational 10143 Decision-Making</span 10144 > 10145 </div> 10146 <div class="slot-authors"> 10147 Alexander R. Heib, Christine S. M. Currie, Bhakti 10148 Stephan Onggo, and Honora K. Smith (University of 10149 Southampton) and James Kerr (Hampshire Hospitals NHS 10150 Foundation Trust) 10151 </div> 10152 <div class="slot-abstract"> 10153 <div> 10154 <a 10155 class="clickable no-decoration" 10156 id="vhsjs_view_210_1707793551_8040168" 10157 onclick="$('#vhsjs_view_210_1707793551_8040168').hide(); 10158 $('#vhsjs_hide_210_1707793551_8040168').show(); 10159 $('#209_1707793551_804009').slideDown(function() { 10160 if (typeof Masonry === 'function') { 10161 $('.use_masonry').masonry(); 10162 }; 10163 10164 });" 10165 ><i class="fa fa-caret-right"></i>
10166 <span class="hover_link">Abstract</span></a 10167 ><a 10168 class="clickable no-decoration" 10169 id="vhsjs_hide_210_1707793551_8040168" 10170 onclick="$('#209_1707793551_804009').hide(function() { 10171 if (typeof Masonry === 'function') { 10172 $('.use_masonry').masonry(); 10173 }; 10174 }); 10175 $('#vhsjs_hide_210_1707793551_8040168').hide(); 10176 $('#vhsjs_view_210_1707793551_8040168').show();" 10177 style="display: none" 10178 ><i class="fa fa-caret-down"></i> 10179 <span class="hover_link">Abstract</span></a 10180 > 10181 <div 10182 data-display-control="210_1707793551_8040168" 10183 id="209_1707793551_804009" 10184 style="display: none" 10185 > 10186 <div class="arrow-slidedown"> 10187 <blockquote> 10188 We describe the design of a generalizable 10189 simulation model of an emergency department (ED) 10190 that forms part of a symbiotic simulation tool 10191 designed to improve short-term decision-making. 10192 While the paper will give an overview of the 10193 planned symbiotic simulation tool, our focus 10194 here is on the generalizability of the 10195 simulation model. The model is coded such that 10196 the routing logic of patient pathways are not 10197 explicitly defined but are instead included as 10198 an input parameter. By structuring the model 10199 this way, the pathways can instead be discovered 10200 through process mining methods on standard 10201 healthcare transactions data. This enables the 10202 simulation model to be applied to other EDs 10203 without redesigning all of the logical flows 10204 within the model. As symbiotic simulation tools 10205 are designed for ongoing use within the system 10206 they model, utilizing process mining also allows 10207 for automating recalibration of the patient 10208 pathways if changes occur in the physical 10209 system. 10210 </blockquote> 10211 </div> 10212 </div> 10213 </div> 10214 </div> 10215 <div class="slot-urls"></div> 10216 <a href="/wsc23papers/087.pdf" target="_blank">pdf</a 10217 ><br /> 10218 </div> 10219 </div> 10220 <div class="session-entry"> 10221 <span class="session-event-type">Technical Session</span 10222 ><span class="type-track-spacer"> · </span 10223 ><span class="program-track" 10224 >Healthcare and Life Sciences</span 10225 ><br /> 10226 <div class="session-title"> 10227 Discrete-event Simulation Models to Inform Healthcare 10228 Decisions 10229 </div> 10230 <div class="session-chair"> 10231 Chair: Marta Staff (University of Exeter)<br /> 10232 </div> 10233 <div class="slot-entry"> 10234 <a name="con113" tabindex="-1"></a> 10235 <div class="slot-title-line"> 10236 <span class="slot-title" 10237 >Estimating Quantile Fields for a Simulated Model of a 10238 Homeless Care System</span 10239 > 10240 </div> 10241 <div class="slot-authors"> 10242 Dashi I. Singham (Naval Postgradaute School) 10243 </div> 10244 <div class="slot-abstract"> 10245 <div> 10246 <a 10247 class="clickable no-decoration" 10248 id="vhsjs_view_212_1707793551_8085656" 10249 onclick="$('#vhsjs_view_212_1707793551_8085656').hide(); 10250 $('#vhsjs_hide_212_1707793551_8085656').show(); 10251 $('#211_1707793551_8085573').slideDown(function() { 10252 if (typeof Masonry === 'function') { 10253 $('.use_masonry').masonry(); 10254 }; 10255 10256 });" 10257 ><i class="fa fa-caret-right"></i>
10258 <span class="hover_link">Abstract</span></a 10259 ><a 10260 class="clickable no-decoration" 10261 id="vhsjs_hide_212_1707793551_8085656" 10262 onclick="$('#211_1707793551_8085573').hide(function() { 10263 if (typeof Masonry === 'function') { 10264 $('.use_masonry').masonry(); 10265 }; 10266 }); 10267 $('#vhsjs_hide_212_1707793551_8085656').hide(); 10268 $('#vhsjs_view_212_1707793551_8085656').show();" 10269 style="display: none" 10270 ><i class="fa fa-caret-down"></i> 10271 <span class="hover_link">Abstract</span></a 10272 > 10273 <div 10274 data-display-control="212_1707793551_8085656" 10275 id="211_1707793551_8085573" 10276 style="display: none" 10277 > 10278 <div class="arrow-slidedown"> 10279 <blockquote> 10280 We construct a simulation model of a homeless 10281 care system to determine the amount of new 10282 housing and emergency shelter needed to support 10283 the growing unsheltered population in Alameda 10284 County, California. To quantify the performance 10285 of the system, we assess the number of people 10286 having unmet need via an estimate of the 10287 quantile field using a recently developed 10288 batching method. This approach helps right-size 10289 the amount of housing and shelter resources 10290 needed to quickly provide services to the 10291 unsheltered population. We find that with a 10292 large investment in housing to help the system 10293 reach steady state, current levels of emergency 10294 shelter may be sufficient to serve those with 10295 unmet need. 10296 </blockquote> 10297 </div> 10298 </div> 10299 </div> 10300 </div> 10301 <div class="slot-urls"></div> 10302 <a href="/wsc23papers/088.pdf" target="_blank">pdf</a 10303 ><br /> 10304 </div> 10305 <div class="slot-entry"> 10306 <a name="con333" tabindex="-1"></a> 10307 <div class="slot-title-line"> 10308 <span class="slot-title" 10309 >Measuring the Operational Impacts of Right-Sizing 10310 Prenatal Care Using Simulation</span 10311 > 10312 </div> 10313 <div class="slot-authors"> 10314 Leena Ghrayeb, Timothy Bryan, Meghana Kandiraju, Tejas 10315 Maire, Yuanbo Zhang, Amy Cohn, and Alex Peahl 10316 (University of Michigan) 10317 </div> 10318 <div class="slot-abstract"> 10319 <div> 10320 <a 10321 class="clickable no-decoration" 10322 id="vhsjs_view_214_1707793551_811081" 10323 onclick="$('#vhsjs_view_214_1707793551_811081').hide(); 10324 $('#vhsjs_hide_214_1707793551_811081').show(); 10325 $('#213_1707793551_811073').slideDown(function() { 10326 if (typeof Masonry === 'function') { 10327 $('.use_masonry').masonry(); 10328 }; 10329 10330 });" 10331 ><i class="fa fa-caret-right"></i>
10332 <span class="hover_link">Abstract</span></a 10333 ><a 10334 class="clickable no-decoration" 10335 id="vhsjs_hide_214_1707793551_811081" 10336 onclick="$('#213_1707793551_811073').hide(function() { 10337 if (typeof Masonry === 'function') { 10338 $('.use_masonry').masonry(); 10339 }; 10340 }); 10341 $('#vhsjs_hide_214_1707793551_811081').hide(); 10342 $('#vhsjs_view_214_1707793551_811081').show();" 10343 style="display: none" 10344 ><i class="fa fa-caret-down"></i> 10345 <span class="hover_link">Abstract</span></a 10346 > 10347 <div 10348 data-display-control="214_1707793551_811081" 10349 id="213_1707793551_811073" 10350 style="display: none" 10351 > 10352 <div class="arrow-slidedown"> 10353 <blockquote> 10354 Despite high levels of spending on prenatal 10355 care, the U.S. has the worst maternal mortality 10356 outcomes amongst peer high-income nations. In 10357 response to a growing need for modernized 10358 prenatal care policies, national prenatal care 10359 stakeholders have developed a new model of 10360 prenatal care, which moves away from a 10361 “one-size-fits-all” model of 10362 prenatal care delivery, and instead tailors care 10363 to patients’ specific needs. In this 10364 article, we develop a data-driven discrete event 10365 simulation model to quantify the operational 10366 impacts of adopting this new care paradigm. We 10367 consider a case study of a large academic health 10368 center, and derive input parameters for the 10369 model from historical data. Our results suggest 10370 that when compared with the 10371 “one-size-fits-all” model of care, 10372 the new tailored care policy leads to reduced 10373 patient delays, as well as a reduction in 10374 overbooking, implying increased flexibility in 10375 the system. 10376 </blockquote> 10377 </div> 10378 </div> 10379 </div> 10380 </div> 10381 <div class="slot-urls"></div> 10382 <a href="/wsc23papers/089.pdf" target="_blank">pdf</a 10383 ><br /> 10384 </div> 10385 <div class="slot-entry"> 10386 <a name="inv154" tabindex="-1"></a> 10387 <div class="slot-title-line"> 10388 <span class="slot-title" 10389 >Open-Source Modeling for Orthopedic Elective Capacity 10390 Planning Using Discrete-Event Simulation</span 10391 > 10392 </div> 10393 <div class="slot-authors"> 10394 Alison Harper, Martin Pitt, and Thomas Monks (University 10395 of Exeter) 10396 </div> 10397 <div class="slot-abstract"> 10398 <div> 10399 <a 10400 class="clickable no-decoration" 10401 id="vhsjs_view_216_1707793551_8143234" 10402 onclick="$('#vhsjs_view_216_1707793551_8143234').hide(); 10403 $('#vhsjs_hide_216_1707793551_8143234').show(); 10404 $('#215_1707793551_814315').slideDown(function() { 10405 if (typeof Masonry === 'function') { 10406 $('.use_masonry').masonry(); 10407 }; 10408 10409 });" 10410 ><i class="fa fa-caret-right"></i>
10411 <span class="hover_link">Abstract</span></a 10412 ><a 10413 class="clickable no-decoration" 10414 id="vhsjs_hide_216_1707793551_8143234" 10415 onclick="$('#215_1707793551_814315').hide(function() { 10416 if (typeof Masonry === 'function') { 10417 $('.use_masonry').masonry(); 10418 }; 10419 }); 10420 $('#vhsjs_hide_216_1707793551_8143234').hide(); 10421 $('#vhsjs_view_216_1707793551_8143234').show();" 10422 style="display: none" 10423 ><i class="fa fa-caret-down"></i> 10424 <span class="hover_link">Abstract</span></a 10425 > 10426 <div 10427 data-display-control="216_1707793551_8143234" 10428 id="215_1707793551_814315" 10429 style="display: none" 10430 > 10431 <div class="arrow-slidedown"> 10432 <blockquote> 10433 The increase in elective surgical waiting lists 10434 as a result of the COVID-19 pandemic is creating 10435 significant consequences for health services 10436 worldwide. In the UK, the allocation of capital 10437 funds to increase capacity for managing elective 10438 waits has created planning and operational 10439 challenges for health services. This paper 10440 reports on the development and deployment of an 10441 interactive web-based discrete-event simulation 10442 model for supporting capacity planning of 10443 surgical activity and ward stay in a proposed 10444 new ring-fenced orthopedic facility in a UK 10445 health service. The model is free and 10446 open-source and developed to be generic and 10447 applicable for new capacity planning of elective 10448 recovery in orthopedics in other regions. With 10449 minor adaptations it can also be readily 10450 modified for application to other specialties. 10451 Given the current relevance of managing record 10452 elective waiting lists, there is potential 10453 widespread applicability of the simulation model 10454 which is supported by our open approach to 10455 modeling. 10456 </blockquote> 10457 </div> 10458 </div> 10459 </div> 10460 </div> 10461 <div class="slot-urls"></div> 10462 <a href="/wsc23papers/090.pdf" target="_blank">pdf</a 10463 ><br /> 10464 </div> 10465 </div> 10466 <div class="session-entry"> 10467 <span class="session-event-type">Technical Session</span 10468 ><span class="type-track-spacer"> · </span 10469 ><span class="program-track" 10470 >Healthcare and Life Sciences</span 10471 ><br /> 10472 <div class="session-title"> 10473 Simulation Modeling for Covid-19 III 10474 </div> 10475 <div class="session-chair"> 10476 Chair: Arindam Fadikar (Argonne National Laboratory)<br /> 10477 </div> 10478 <div class="slot-entry"> 10479 <a name="con270" tabindex="-1"></a> 10480 <div class="slot-title-line"> 10481 <span class="slot-title" 10482 >Evaluating Parallelization Strategies for Large-Scale 10483 Individual-Based Infectious Disease Simulations</span 10484 > 10485 </div> 10486 <div class="slot-authors"> 10487 Johannes Ponge (University of Münster), Lukas Bayer 10488 (RPTU Kaiserslautern-Landau), Dennis Horstkemper 10489 (University of Münster), Wolfgang Bock (RPTU 10490 Kaiserslautern-Landau), and Bernd Hellingrath and 10491 André Karch (University of Münster) 10492 </div> 10493 <div class="slot-abstract"> 10494 <div> 10495 <a 10496 class="clickable no-decoration" 10497 id="vhsjs_view_218_1707793551_8191028" 10498 onclick="$('#vhsjs_view_218_1707793551_8191028').hide(); 10499 $('#vhsjs_hide_218_1707793551_8191028').show(); 10500 $('#217_1707793551_8190944').slideDown(function() { 10501 if (typeof Masonry === 'function') { 10502 $('.use_masonry').masonry(); 10503 }; 10504 10505 });" 10506 ><i class="fa fa-caret-right"></i>
10507 <span class="hover_link">Abstract</span></a 10508 ><a 10509 class="clickable no-decoration" 10510 id="vhsjs_hide_218_1707793551_8191028" 10511 onclick="$('#217_1707793551_8190944').hide(function() { 10512 if (typeof Masonry === 'function') { 10513 $('.use_masonry').masonry(); 10514 }; 10515 }); 10516 $('#vhsjs_hide_218_1707793551_8191028').hide(); 10517 $('#vhsjs_view_218_1707793551_8191028').show();" 10518 style="display: none" 10519 ><i class="fa fa-caret-down"></i> 10520 <span class="hover_link">Abstract</span></a 10521 > 10522 <div 10523 data-display-control="218_1707793551_8191028" 10524 id="217_1707793551_8190944" 10525 style="display: none" 10526 > 10527 <div class="arrow-slidedown"> 10528 <blockquote> 10529 Individual-based models (IBMs) of infectious 10530 disease dynamics with full-country populations 10531 often suffer from high runtimes. While there are 10532 approaches to parallelize simulations, many 10533 prominent epidemic models exhibit single-core 10534 implementations, suggesting a lack of consensus 10535 among the research community on whether 10536 parallelization is desirable or achievable. 10537 Rising demands in model scope and complexity, 10538 however, imply that performance will continue to 10539 be a bottleneck. In this paper, we discuss the 10540 requirements and challenges of parallel IBMs in 10541 general and the German Epidemic Micro-Simulation 10542 System (GEMS) in particular. While the 10543 exploitation of unique model characteristics can 10544 yield significant performance improvement 10545 potential, parallelization strategies generally 10546 necessitate trade-offs in either hardware 10547 requirements, model fidelity, or implementation 10548 complexity. Therefore, the selection of 10549 parallelization strategies requires a 10550 comprehensive assessment. We present a 10551 point-based evaluation scheme to assess the 10552 potential of parallelization strategies as our 10553 main contribution and exemplify its application 10554 in the context of GEMS. 10555 </blockquote> 10556 </div> 10557 </div> 10558 </div> 10559 </div> 10560 <div class="slot-urls"></div> 10561 <a href="/wsc23papers/091.pdf" target="_blank">pdf</a 10562 ><br /> 10563 </div> 10564 <div class="slot-entry"> 10565 <a name="con344" tabindex="-1"></a> 10566 <div class="slot-title-line"> 10567 <span class="slot-title" 10568 >Determining the Impact of Facility Layout Methods on 10569 Walk-in Covid-19 Vaccine Clinics: A Theoretical 10570 Exploration</span 10571 > 10572 </div> 10573 <div class="slot-authors"> 10574 S. Yasaman Ahmadi and Jennifer Lather (University of 10575 Nebraska Lincoln) 10576 </div> 10577 <div class="slot-abstract"> 10578 <div> 10579 <a 10580 class="clickable no-decoration" 10581 id="vhsjs_view_220_1707793551_8214436" 10582 onclick="$('#vhsjs_view_220_1707793551_8214436').hide(); 10583 $('#vhsjs_hide_220_1707793551_8214436').show(); 10584 $('#219_1707793551_8214355').slideDown(function() { 10585 if (typeof Masonry === 'function') { 10586 $('.use_masonry').masonry(); 10587 }; 10588 10589 });" 10590 ><i class="fa fa-caret-right"></i>
10591 <span class="hover_link">Abstract</span></a 10592 ><a 10593 class="clickable no-decoration" 10594 id="vhsjs_hide_220_1707793551_8214436" 10595 onclick="$('#219_1707793551_8214355').hide(function() { 10596 if (typeof Masonry === 'function') { 10597 $('.use_masonry').masonry(); 10598 }; 10599 }); 10600 $('#vhsjs_hide_220_1707793551_8214436').hide(); 10601 $('#vhsjs_view_220_1707793551_8214436').show();" 10602 style="display: none" 10603 ><i class="fa fa-caret-down"></i> 10604 <span class="hover_link">Abstract</span></a 10605 > 10606 <div 10607 data-display-control="220_1707793551_8214436" 10608 id="219_1707793551_8214355" 10609 style="display: none" 10610 > 10611 <div class="arrow-slidedown"> 10612 <blockquote> 10613 Ensuring safety and public health is a paramount 10614 concern in mass vaccination against contagious 10615 respiratory infections. This study examines the 10616 effects of layout methods and path routing 10617 decisions on average patient travel distance 10618 (TD) and time-in-system (TIS) within the context 10619 of a theoretical mass vaccination clinic. Two 10620 distinct layout methods, Perimeter and 10621 Serpentine, are evaluated in conjunction with 10622 two path routing conditions, Cyclical and 10623 Unidirectional. Employing discrete-event 10624 simulation, the study investigates multiple 10625 patient turnouts and clinic operational hours. 10626 The results reveal the significant impact of 10627 layout on average TD, underscoring the 10628 heightened efficiency of the Perimeter layout 10629 and Unidirectional path. Furthermore, the 10630 findings highlight the significant effect of 10631 layout method on TIS when considering optimal 10632 staffing configurations. Conversely, the 10633 analysis indicates that path directionality does 10634 not exert a statistically significant effect. 10635 This study emphasizes the critical role of 10636 layout design in optimizing vaccination clinics 10637 for efficiency and effectiveness. 10638 </blockquote> 10639 </div> 10640 </div> 10641 </div> 10642 </div> 10643 <div class="slot-urls"></div> 10644 <a href="/wsc23papers/092.pdf" target="_blank">pdf</a 10645 ><br /> 10646 </div> 10647 <div class="slot-entry"> 10648 <a name="con323" tabindex="-1"></a> 10649 <div class="slot-title-line"> 10650 <span class="slot-title" 10651 >A Network-based Analytics Framework For 10652 High-resolution Agent-Based Epidemic Simulation 10653 Ensembles</span 10654 > 10655 </div> 10656 <div class="slot-authors"> 10657 Amro Alabsi Aljundi, Galen Harrison, Jiangzhuo Chen, 10658 Madhav Marathe, Henning S. Mortveit, Anil Vullikanti, 10659 and Abhijin Adiga (University of Virginia) 10660 </div> 10661 <div class="slot-abstract"> 10662 <div> 10663 <a 10664 class="clickable no-decoration" 10665 id="vhsjs_view_222_1707793551_8238652" 10666 onclick="$('#vhsjs_view_222_1707793551_8238652').hide(); 10667 $('#vhsjs_hide_222_1707793551_8238652').show(); 10668 $('#221_1707793551_8238568').slideDown(function() { 10669 if (typeof Masonry === 'function') { 10670 $('.use_masonry').masonry(); 10671 }; 10672 10673 });" 10674 ><i class="fa fa-caret-right"></i>
10675 <span class="hover_link">Abstract</span></a 10676 ><a 10677 class="clickable no-decoration" 10678 id="vhsjs_hide_222_1707793551_8238652" 10679 onclick="$('#221_1707793551_8238568').hide(function() { 10680 if (typeof Masonry === 'function') { 10681 $('.use_masonry').masonry(); 10682 }; 10683 }); 10684 $('#vhsjs_hide_222_1707793551_8238652').hide(); 10685 $('#vhsjs_view_222_1707793551_8238652').show();" 10686 style="display: none" 10687 ><i class="fa fa-caret-down"></i> 10688 <span class="hover_link">Abstract</span></a 10689 > 10690 <div 10691 data-display-control="222_1707793551_8238652" 10692 id="221_1707793551_8238568" 10693 style="display: none" 10694 > 10695 <div class="arrow-slidedown"> 10696 <blockquote> 10697 High-resolution network-based contagion models 10698 are being increasingly used to study complex 10699 disease scenarios. Due to network-induced 10700 heterogeneity and sophisticated disease and 10701 intervention models, even simple simulation 10702 exercises can lead to large volumes of complex 10703 simulation outcomes. New approaches are required 10704 to analyze them. Simulations of such network 10705 spread processes can be viewed as attributed 10706 temporal graphs. We describe a network-based 10707 analytics framework that enables a user to 10708 leverage this graphical viewpoint and apply 10709 graph mining methods to perform fine-grained 10710 analysis of the simulation outcomes and the 10711 underlying network. The framework is based on a 10712 microservices-oriented architecture, and is 10713 designed to be general, adaptable, and scalable. 10714 We demonstrate its utility through a case study 10715 motivated by the COVID-19 pandemic involving the 10716 spread of two variants on a large realistic 10717 population network with multiple interventions. 10718 We study the transmissions within and between 10719 age-groups, importance of non-essential 10720 interactions, and efficacy of interventions. 10721 </blockquote> 10722 </div> 10723 </div> 10724 </div> 10725 </div> 10726 <div class="slot-urls"></div> 10727 <a href="/wsc23papers/315.pdf" target="_blank">pdf</a 10728 ><br /> 10729 </div> 10730 </div> 10731 <div class="session-entry"> 10732 <span class="session-event-type">Technical Session</span 10733 ><span class="type-track-spacer"> · </span 10734 ><span class="program-track" 10735 >Healthcare and Life Sciences</span 10736 ><br /> 10737 <div class="session-title">Medical Decision Analysis</div> 10738 <div class="session-chair"> 10739 Chair: Navonil Mustafee (University of Exeter, The 10740 Business School)<br /> 10741 </div> 10742 <div class="slot-entry"> 10743 <a name="cea101" tabindex="-1"></a> 10744 <div class="slot-title-line">
10745 <span class="slot-title" 10746 >Continuous-Time Survival Model Study Designs for 10747 Heart Recovery Applications</span 10748 > 10749 </div> 10750 <div class="slot-authors"> 10751 Jason Bodnar (ABIOMED, Inc.) 10752 </div> 10753 <div class="slot-abstract"> 10754 <div> 10755 <a 10756 class="clickable no-decoration" 10757 id="vhsjs_view_224_1707793551_8283389" 10758 onclick="$('#vhsjs_view_224_1707793551_8283389').hide(); 10759 $('#vhsjs_hide_224_1707793551_8283389').show(); 10760 $('#223_1707793551_8283308').slideDown(function() { 10761 if (typeof Masonry === 'function') { 10762 $('.use_masonry').masonry(); 10763 }; 10764 10765 });" 10766 ><i class="fa fa-caret-right"></i> 10767 <span class="hover_link">Abstract</span></a 10768 ><a 10769 class="clickable no-decoration" 10770 id="vhsjs_hide_224_1707793551_8283389" 10771 onclick="$('#223_1707793551_8283308').hide(function() { 10772 if (typeof Masonry === 'function') { 10773 $('.use_masonry').masonry(); 10774 }; 10775 }); 10776 $('#vhsjs_hide_224_1707793551_8283389').hide(); 10777 $('#vhsjs_view_224_1707793551_8283389').show();" 10778 style="display: none" 10779 ><i class="fa fa-caret-down"></i> 10780 <span class="hover_link">Abstract</span></a 10781 > 10782 <div 10783 data-display-control="224_1707793551_8283389" 10784 id="223_1707793551_8283308" 10785 style="display: none" 10786 > 10787 <div class="arrow-slidedown"> 10788 <blockquote> 10789 Due to the aging global population, the science 10790 of heart recovery is an essential area for 10791 research to improve patient health, reduce 10792 time-to-discharge, and delay overall mortality. 10793 New medical device technology is needed to 10794 advance these goals. For the medical community 10795 to gain trust in and use these technologies in 10796 their hospital environments, optimal study 10797 design and proper execution of randomized 10798 controlled trials is necessary. Such RCTs will 10799 result in the collection of valid scientific 10800 evidence for establishing the new device’s 10801 risk and benefit profile in targeted patient 10802 populations. Continuous time-to-event survival 10803 models are commonly used to determine the amount 10804 of data needed to demonstrate an improvement in 10805 these profiles over current standard-of-care 10806 therapies. This paper will compare simulated 10807 power functions and sample size requirements for 10808 a variety of survival methods in a two-sample 10809 RCT setting. Simulation scenarios will encompass 10810 various effect sizes, survival distribution 10811 forms, and time-to-event density functions. 10812 </blockquote> 10813 </div> 10814 </div> 10815 </div> 10816 </div> 10817 <div class="slot-urls"></div> 10818 <a href="/wsc23papers/cea101.pdf" target="_blank">pdf</a 10819 ><br /> 10820 </div> 10821 <div class="slot-entry"> 10822 <a name="cea124" tabindex="-1"></a> 10823 <div class="slot-title-line">
10824 <span class="slot-title" 10825 >KSIM 2.0: A Simulation of Kidney Allocation Using 10826 OPTN Records</span 10827 > 10828 </div> 10829 <div class="slot-authors"> 10830 Masoud Barah (Northwestern University), Vikram Kilambi 10831 (RAND Corporation), and Sanjay Mehrotra (Northwestern 10832 University) 10833 </div> 10834 <div class="slot-abstract"> 10835 <div> 10836 <a 10837 class="clickable no-decoration" 10838 id="vhsjs_view_226_1707793551_8304791" 10839 onclick="$('#vhsjs_view_226_1707793551_8304791').hide(); 10840 $('#vhsjs_hide_226_1707793551_8304791').show(); 10841 $('#225_1707793551_8304708').slideDown(function() { 10842 if (typeof Masonry === 'function') { 10843 $('.use_masonry').masonry(); 10844 }; 10845 10846 });" 10847 ><i class="fa fa-caret-right"></i> 10848 <span class="hover_link">Abstract</span></a 10849 ><a 10850 class="clickable no-decoration" 10851 id="vhsjs_hide_226_1707793551_8304791" 10852 onclick="$('#225_1707793551_8304708').hide(function() { 10853 if (typeof Masonry === 'function') { 10854 $('.use_masonry').masonry(); 10855 }; 10856 }); 10857 $('#vhsjs_hide_226_1707793551_8304791').hide(); 10858 $('#vhsjs_view_226_1707793551_8304791').show();" 10859 style="display: none" 10860 ><i class="fa fa-caret-down"></i> 10861 <span class="hover_link">Abstract</span></a 10862 > 10863 <div 10864 data-display-control="226_1707793551_8304791" 10865 id="225_1707793551_8304708" 10866 style="display: none" 10867 > 10868 <div class="arrow-slidedown"> 10869 <blockquote> 10870 The Organ Procurement and Transplantation 10871 Network (OPTN) in the US allocates kidneys for 10872 transplantation, but nearly one fifth of kidneys 10873 from deceased donors are not utilized due to the 10874 avoidance of transplantation for kidneys that 10875 have been removed from a donor for too long. To 10876 be able to provide clinically relevant 10877 recommendations to the OPTN contractor, we 10878 updated the KSIM discrete event simulation of 10879 kidney allocation in the academic literature 10880 using actual OPTN individual-level records for 10881 patients and donors. As a case study, we 10882 simulated offering kidneys at high risk of 10883 discard to the first accepting transplant center 10884 after 10 hours of accumulated cold time and 10885 found increased utilization. The updated model 10886 allows for greater clinical fidelity and can be 10887 embedded in medical decision support systems. 10888 </blockquote> 10889 </div> 10890 </div> 10891 </div> 10892 </div> 10893 <div class="slot-urls"></div> 10894 <a href="/wsc23papers/cea124.pdf" target="_blank">pdf</a 10895 ><br /> 10896 </div> 10897 <div class="slot-entry"> 10898 <a name="con206" tabindex="-1"></a> 10899 <div class="slot-title-line">
10900 <span class="slot-title" 10901 >Modeling and Simulation of the SARS-CoV-2 Lung 10902 Infection and Immune Response with Cell-DEVS</span 10903 > 10904 </div> 10905 <div class="slot-authors"> 10906 Ali Ayadi (University of Strasbourg, ICube laboratory); 10907 Claudia Frydman (Aix Marseille Université); and Quy 10908 Thanh Le (Da Nang University of Science and Technology) 10909 </div> 10910 <div class="slot-abstract"> 10911 <div> 10912 <a 10913 class="clickable no-decoration" 10914 id="vhsjs_view_228_1707793551_8328772" 10915 onclick="$('#vhsjs_view_228_1707793551_8328772').hide(); 10916 $('#vhsjs_hide_228_1707793551_8328772').show(); 10917 $('#227_1707793551_832869').slideDown(function() { 10918 if (typeof Masonry === 'function') { 10919 $('.use_masonry').masonry(); 10920 }; 10921 10922 });" 10923 ><i class="fa fa-caret-right"></i> 10924 <span class="hover_link">Abstract</span></a 10925 ><a 10926 class="clickable no-decoration" 10927 id="vhsjs_hide_228_1707793551_8328772" 10928 onclick="$('#227_1707793551_832869').hide(function() { 10929 if (typeof Masonry === 'function') { 10930 $('.use_masonry').masonry(); 10931 }; 10932 }); 10933 $('#vhsjs_hide_228_1707793551_8328772').hide(); 10934 $('#vhsjs_view_228_1707793551_8328772').show();" 10935 style="display: none" 10936 ><i class="fa fa-caret-down"></i> 10937 <span class="hover_link">Abstract</span></a 10938 > 10939 <div 10940 data-display-control="228_1707793551_8328772" 10941 id="227_1707793551_832869" 10942 style="display: none" 10943 > 10944 <div class="arrow-slidedown"> 10945 <blockquote> 10946 Understanding why patients' viral loads vary 10947 dramatically across individuals is a critical 10948 challenge in addressing respiratory infections, 10949 especially the severe acute respiratory syndrome 10950 coronavirus 2 (SARS-CoV-2). The spatial-temporal 10951 dynamics of viral infection in the respiratory 10952 system and the immune system's response remain 10953 difficult to study. Using modelling and 10954 simulation (M&S) techniques may address this 10955 problem. In this paper, we present a novel 10956 modelling approach using the Cell-DEVS formalism 10957 (a combination of Cellular Automata and DEVS), 10958 to simulate the spatial-temporal dynamics of 10959 viral spread in the lungs. Using a 10960 two-dimensional cellular space that mimics a 10961 lung, the proposed approach focuses also on the 10962 immune system response, viral infection spread, 10963 state of lung epithelial tissue damage, and 10964 immune cells' state. We demonstrate the 10965 pertinence of our proposal on three different 10966 scenarios representing three types of patients. 10967 Qualitative evaluation by expert biologists 10968 confirms that the produced simulations match the 10969 observations made on patients. 10970 </blockquote> 10971 </div> 10972 </div> 10973 </div> 10974 </div> 10975 <div class="slot-urls"></div> 10976 <a href="/wsc23papers/100.pdf" target="_blank">pdf</a 10977 ><br /> 10978 </div> 10979 </div> 10980 <div class="session-entry">
10981 <span class="session-event-type">Technical Session</span 10982 ><span class="type-track-spacer"> · </span 10983 ><span class="program-track" 10984 >Healthcare and Life Sciences</span 10985 ><br /> 10986 <div class="session-title"> 10987 Patient Flow Through Healthcare Processes 10988 </div> 10989 <div class="session-chair"> 10990 Chair: Alison Harper (University of Exeter, The Business 10991 School)<br /> 10992 </div> 10993 <div class="slot-entry"> 10994 <a name="con296" tabindex="-1"></a> 10995 <div class="slot-title-line"> 10996 <span class="slot-title" 10997 >Integrating Home Health Care and Patient 10998 Transportation: A Sample Average Approximation 10999 Approach to Optimize Scheduling and Routing</span 11000 > 11001 </div> 11002 <div class="slot-authors"> 11003 Lorena Silvana Reyes Rubiano (Universidad de La Sabana, 11004 RWTH Aachen University); Marcel Müller (Otto von 11005 Guericke University Magdeburg); Jana Voegl (University 11006 of Natural Resources and Life Sciences Vienna); Angelica 11007 Sarmiento (Colombian School of Engineering Julio 11008 Garavito); William Javier Guerrero (Universidad de La 11009 Sabana); and Patrick Hirsch (University of Natural 11010 Resources and Life Sciences Vienna) 11011 </div> 11012 <div class="slot-abstract"> 11013 <div> 11014 <a 11015 class="clickable no-decoration" 11016 id="vhsjs_view_230_1707793551_8391328" 11017 onclick="$('#vhsjs_view_230_1707793551_8391328').hide(); 11018 $('#vhsjs_hide_230_1707793551_8391328').show(); 11019 $('#229_1707793551_839125').slideDown(function() { 11020 if (typeof Masonry === 'function') { 11021 $('.use_masonry').masonry(); 11022 }; 11023 11024 });" 11025 ><i class="fa fa-caret-right"></i> 11026 <span class="hover_link">Abstract</span></a 11027 ><a 11028 class="clickable no-decoration" 11029 id="vhsjs_hide_230_1707793551_8391328" 11030 onclick="$('#229_1707793551_839125').hide(function() { 11031 if (typeof Masonry === 'function') { 11032 $('.use_masonry').masonry(); 11033 }; 11034 }); 11035 $('#vhsjs_hide_230_1707793551_8391328').hide(); 11036 $('#vhsjs_view_230_1707793551_8391328').show();" 11037 style="display: none" 11038 ><i class="fa fa-caret-down"></i> 11039 <span class="hover_link">Abstract</span></a 11040 > 11041 <div 11042 data-display-control="230_1707793551_8391328" 11043 id="229_1707793551_839125" 11044 style="display: none" 11045 > 11046 <div class="arrow-slidedown"> 11047 <blockquote> 11048 This study introduces an innovative strategy for 11049 addressing the Home Healthcare and Dial-a-Ride 11050 Problem (HHCDAP) concerning the transportation 11051 of medical staff and patients, taking into 11052 account the stochastic nature of service and 11053 travel times. The problem involves assigning 11054 suitable medical staff to patients and clients, 11055 determining the order of visits, and identifying 11056 opportunities for medical staff and patients to 11057 share trips. We propose two objective functions 11058 to minimize travel time for drivers and medical 11059 staff. This problem adheres to numerous 11060 constraints, including maximum work duration, 11061 maximum waiting time, professional 11062 qualifications, and vehicle capacity 11063 limitations. We test our approach on a 11064 small-scale instance to understand the 11065 trade-offs between minimizing drivers' travel 11066 time and minimizing the travel and waiting times 11067 of medical staff and patients. Our results 11068 indicate that the proposed strategy enhances the 11069 efficiency of transporting medical staff and 11070 patients. 11071 </blockquote> 11072 </div> 11073 </div> 11074 </div> 11075 </div> 11076 <div class="slot-urls"></div> 11077 <a href="/wsc23papers/093.pdf" target="_blank">pdf</a 11078 ><br /> 11079 </div> 11080 <div class="slot-entry"> 11081 <a name="con309" tabindex="-1"></a> 11082 <div class="slot-title-line">
11083 <span class="slot-title" 11084 >A Preliminary Predictive Simulation Model for Hip and 11085 Knee Replacement Profile-Dependent Pathway 11086 Stages</span 11087 > 11088 </div> 11089 <div class="slot-authors"> 11090 Ahmed Bakali El Kassimi (Ecole des Mines de 11091 Saint-Etienne, Univ Clermont Auvergne, INP Clermont 11092 Auvergne, CNRS, UMR 6158 LIMOS); Marianne Sarazin 11093 (Clinique Médico-Chirurgicale Mutualiste, Groupe 11094 Aésio Santé); Xiaolan Xie (Ecole des Mines de 11095 Saint-Etienne, Univ Clermont Auvergne, INP Clermont 11096 Auvergne, CNRS, UMR 6158 LIMOS); and Pierre-Luc Fresard 11097 and Bertand Semay (Clinique Médico-Chirurgicale 11098 Mutualiste, Groupe Aésio Santé) 11099 </div> 11100 <div class="slot-abstract"> 11101 <div> 11102 <a 11103 class="clickable no-decoration" 11104 id="vhsjs_view_232_1707793551_8414938" 11105 onclick="$('#vhsjs_view_232_1707793551_8414938').hide(); 11106 $('#vhsjs_hide_232_1707793551_8414938').show(); 11107 $('#231_1707793551_8414857').slideDown(function() { 11108 if (typeof Masonry === 'function') { 11109 $('.use_masonry').masonry(); 11110 }; 11111 11112 });" 11113 ><i class="fa fa-caret-right"></i> 11114 <span class="hover_link">Abstract</span></a 11115 ><a 11116 class="clickable no-decoration" 11117 id="vhsjs_hide_232_1707793551_8414938" 11118 onclick="$('#231_1707793551_8414857').hide(function() { 11119 if (typeof Masonry === 'function') { 11120 $('.use_masonry').masonry(); 11121 }; 11122 }); 11123 $('#vhsjs_hide_232_1707793551_8414938').hide(); 11124 $('#vhsjs_view_232_1707793551_8414938').show();" 11125 style="display: none" 11126 ><i class="fa fa-caret-down"></i> 11127 <span class="hover_link">Abstract</span></a 11128 > 11129 <div 11130 data-display-control="232_1707793551_8414938" 11131 id="231_1707793551_8414857" 11132 style="display: none" 11133 > 11134 <div class="arrow-slidedown"> 11135 <blockquote> 11136 Total hip and knee arthroplasty (THA/TKA) 11137 surgeries are frequently performed on elderly 11138 individuals and consist of preoperative, 11139 operative, and rehabilitation stages. Despite 11140 efforts to improve patient satisfaction,there is 11141 a lack of personalized studies that optimize the 11142 THA/TKA pathway. Our aim is to address this gap 11143 by proposing a predictive simulation model that 11144 considers patient-specific factors to enhance 11145 patient satisfaction and organizational 11146 efficiency. To achieve this, we propose using 11147 process mining techniques to analyze the French 11148 national healthcare database and distinguish 11149 between standard care phases and 11150 patient-dependent phases. We then apply machine 11151 learning algorithms to predict specific stages 11152 of care. The insights gained from these analyses 11153 are used to compare and test predicted patient 11154 pathways and their performances using our 11155 simulation model. 11156 </blockquote> 11157 </div> 11158 </div> 11159 </div> 11160 </div> 11161 <div class="slot-urls"></div> 11162 <a href="/wsc23papers/094.pdf" target="_blank">pdf</a 11163 ><br /> 11164 </div> 11165 <div class="slot-entry"> 11166 <a name="inv182" tabindex="-1"></a> 11167 <div class="slot-title-line">
11168 <span class="slot-title" 11169 >Forecasting Patient Arrivals and Optimizing Physician 11170 Shift Scheduling in Emergency Departments</span 11171 > 11172 </div> 11173 <div class="slot-authors"> 11174 Vishnunarayan Girishan Prabhu (University of North 11175 Carolina); Kevin Taaffe (Clemson University); and Ronald 11176 Pirrallo, William Jackson, Michael Ramsay, and Jessica 11177 Hobbs (Prisma Health-Upstate) 11178 </div> 11179 <div class="slot-abstract"> 11180 <div> 11181 <a 11182 class="clickable no-decoration" 11183 id="vhsjs_view_234_1707793551_8440456" 11184 onclick="$('#vhsjs_view_234_1707793551_8440456').hide(); 11185 $('#vhsjs_hide_234_1707793551_8440456').show(); 11186 $('#233_1707793551_8440375').slideDown(function() { 11187 if (typeof Masonry === 'function') { 11188 $('.use_masonry').masonry(); 11189 }; 11190 11191 });" 11192 ><i class="fa fa-caret-right"></i> 11193 <span class="hover_link">Abstract</span></a 11194 ><a 11195 class="clickable no-decoration" 11196 id="vhsjs_hide_234_1707793551_8440456" 11197 onclick="$('#233_1707793551_8440375').hide(function() { 11198 if (typeof Masonry === 'function') { 11199 $('.use_masonry').masonry(); 11200 }; 11201 }); 11202 $('#vhsjs_hide_234_1707793551_8440456').hide(); 11203 $('#vhsjs_view_234_1707793551_8440456').show();" 11204 style="display: none" 11205 ><i class="fa fa-caret-down"></i> 11206 <span class="hover_link">Abstract</span></a 11207 > 11208 <div 11209 data-display-control="234_1707793551_8440456" 11210 id="233_1707793551_8440375" 11211 style="display: none" 11212 > 11213 <div class="arrow-slidedown"> 11214 <blockquote> 11215 Emergency Departments (EDs) are the primary 11216 access points for millions of patients seeking 11217 medical care. The increasing patient demand and 11218 lack of long-term dynamic planning strain the 11219 EDs in providing timely patient care, leading to 11220 crowding. While a well-recognized problem, ED 11221 crowding is still prevalent, where suboptimal 11222 resource allocation is one significant 11223 contributing factor. In this research, we 11224 developed an end-to-end solution that first 11225 forecasted the patient arrivals to the partner 11226 ED and then used an optimization model to 11227 develop an optimal physician staffing schedule 11228 to minimize the combined cost of patient wait 11229 times, handoffs, and physician shifts. Finally, 11230 the new schedule was tested using the validated 11231 simulation model to evaluate the ED performance. 11232 By generating shift schedules based on forecasts 11233 and testing them in the validated simulation 11234 model, we observed that patient time in the ED 11235 and handoffs could be reduced by 5.6% and 9.2% 11236 compared to current practices. 11237 </blockquote> 11238 </div> 11239 </div> 11240 </div> 11241 </div> 11242 <div class="slot-urls"></div> 11243 <a href="/wsc23papers/095.pdf" target="_blank">pdf</a 11244 ><br /> 11245 </div> 11246 </div> 11247 <div class="session-entry">
11248 <span class="session-event-type">Technical Session</span 11249 ><span class="type-track-spacer"> · </span 11250 ><span class="program-track" 11251 >Healthcare and Life Sciences</span 11252 ><br /> 11253 <div class="session-title"> 11254 Hybrid Simulation in Healthcare 11255 </div> 11256 <div class="session-chair"> 11257 Chair: Bjorn Berg (University of Minnesota)<br /> 11258 </div> 11259 <div class="slot-entry"> 11260 <a name="inv203" tabindex="-1"></a> 11261 <div class="slot-title-line"> 11262 <span class="slot-title" 11263 >Hybrid Models with Real-Time Data in Healthcare: A 11264 Focus on Data Synchronization and 11265 Experimentation</span 11266 > 11267 </div> 11268 <div class="slot-authors"> 11269 Navonil Mustafee and Alison Harper (University of 11270 Exeter, The Business School) and Joe Viana (BI Norwegian 11271 Business School) 11272 </div> 11273 <div class="slot-abstract"> 11274 <div> 11275 <a 11276 class="clickable no-decoration" 11277 id="vhsjs_view_236_1707793551_8486993" 11278 onclick="$('#vhsjs_view_236_1707793551_8486993').hide(); 11279 $('#vhsjs_hide_236_1707793551_8486993').show(); 11280 $('#235_1707793551_8486915').slideDown(function() { 11281 if (typeof Masonry === 'function') { 11282 $('.use_masonry').masonry(); 11283 }; 11284 11285 });" 11286 ><i class="fa fa-caret-right"></i> 11287 <span class="hover_link">Abstract</span></a 11288 ><a 11289 class="clickable no-decoration" 11290 id="vhsjs_hide_236_1707793551_8486993" 11291 onclick="$('#235_1707793551_8486915').hide(function() { 11292 if (typeof Masonry === 'function') { 11293 $('.use_masonry').masonry(); 11294 }; 11295 }); 11296 $('#vhsjs_hide_236_1707793551_8486993').hide(); 11297 $('#vhsjs_view_236_1707793551_8486993').show();" 11298 style="display: none" 11299 ><i class="fa fa-caret-down"></i> 11300 <span class="hover_link">Abstract</span></a 11301 > 11302 <div 11303 data-display-control="236_1707793551_8486993" 11304 id="235_1707793551_8486915" 11305 style="display: none" 11306 > 11307 <div class="arrow-slidedown"> 11308 <blockquote> 11309 Conventional simulation models used in 11310 Operations Research and Management Science 11311 (OR/MS) use historical data. With the increasing 11312 availability of real-time data, technologies 11313 commonly associated with applied computing, such 11314 as Data Acquisition Systems (DAS), may need to 11315 be integrated with conventional OR/MS models to 11316 develop Hybrid Models (HMs). We distinguish 11317 between HMs that use only real-time data – 11318 we refer to them as Digital Twins (DTs) – 11319 and those using a combination of historical and 11320 real-time data – called Real-time 11321 Simulation (RtS). Our previous contribution 11322 focused on the challenges of such integration, a 11323 concept referred to as information fusion, and 11324 presented a conceptualization of DT/RtS. This 11325 paper focuses on DT/RtS data synchronization and 11326 methods that could be employed from Parallel and 11327 Distributed Simulation (PADS). The 11328 conceptualizations and discussions reflect on 11329 the authors' experience implementing an RtS of a 11330 network of Emergency Departments and Urgent Care 11331 Centers in the UK. 11332 </blockquote> 11333 </div> 11334 </div> 11335 </div> 11336 </div> 11337 <div class="slot-urls"></div> 11338 <a href="/wsc23papers/096.pdf" target="_blank">pdf</a 11339 ><br /> 11340 </div> 11341 <div class="slot-entry"> 11342 <a name="con193" tabindex="-1"></a> 11343 <div class="slot-title-line">
11344 <span class="slot-title" 11345 >Modeling and Simulation of Genomic Sequencing 11346 Platform Operations</span 11347 > 11348 </div> 11349 <div class="slot-authors"> 11350 Jules Le Lay (Centre Léon Bérard), Vincent 11351 Augusto and Xavier Boucher (Mines Saint-Etienne), Lionel 11352 Perrier (Centre Léon Bérard), and Xiaolan Xie 11353 (Mines Saint-Etienne) 11354 </div> 11355 <div class="slot-abstract"> 11356 <div> 11357 <a 11358 class="clickable no-decoration" 11359 id="vhsjs_view_238_1707793551_8512633" 11360 onclick="$('#vhsjs_view_238_1707793551_8512633').hide(); 11361 $('#vhsjs_hide_238_1707793551_8512633').show(); 11362 $('#237_1707793551_851255').slideDown(function() { 11363 if (typeof Masonry === 'function') { 11364 $('.use_masonry').masonry(); 11365 }; 11366 11367 });" 11368 ><i class="fa fa-caret-right"></i> 11369 <span class="hover_link">Abstract</span></a 11370 ><a 11371 class="clickable no-decoration" 11372 id="vhsjs_hide_238_1707793551_8512633" 11373 onclick="$('#237_1707793551_851255').hide(function() { 11374 if (typeof Masonry === 'function') { 11375 $('.use_masonry').masonry(); 11376 }; 11377 }); 11378 $('#vhsjs_hide_238_1707793551_8512633').hide(); 11379 $('#vhsjs_view_238_1707793551_8512633').show();" 11380 style="display: none" 11381 ><i class="fa fa-caret-down"></i> 11382 <span class="hover_link">Abstract</span></a 11383 > 11384 <div 11385 data-display-control="238_1707793551_8512633" 11386 id="237_1707793551_851255" 11387 style="display: none" 11388 > 11389 <div class="arrow-slidedown"> 11390 <blockquote> 11391 This paper focuses on the healthcare application 11392 field of Genomic Sequencing and addresses the 11393 challenge of efficient organization and ramp-up 11394 of sequencing platforms. High-throughput 11395 sequencing platforms are currently in an 11396 industrial prototyping phase in France for large 11397 national deployment afterwards. In the current 11398 state of our knowledge, there is no 11399 scientifically established generic model nor 11400 decision-making support at the operational level 11401 which could guide the medical authorities in 11402 designing organizational rules, then managing 11403 the deployment of such platforms at the national 11404 level. After analyzing the state of the art, a 11405 simulation model of a genome sequencing platform 11406 is presented, then used as a decision-making 11407 support to manage a ramp-up situation for an 11408 application case of a French sequencing 11409 platform. These first results are discussed, 11410 together with the perspective to develop a 11411 generic model and decision-aid approach. 11412 </blockquote> 11413 </div> 11414 </div> 11415 </div> 11416 </div> 11417 <div class="slot-urls"></div> 11418 <a href="/wsc23papers/097.pdf" target="_blank">pdf</a 11419 ><br /> 11420 </div> 11421 </div> 11422 <div class="session-entry">
11423 <span class="session-event-type">Technical Session</span 11424 ><span class="type-track-spacer"> · </span 11425 ><span class="program-track" 11426 >Healthcare and Life Sciences</span 11427 ><br /> 11428 <div class="session-title"> 11429 Simulation Modeling for Infectious Diseases 11430 </div> 11431 <div class="session-chair"> 11432 Chair: Maria Mayorga (North Carolina State University)<br /> 11433 </div> 11434 <div class="slot-entry"> 11435 <a name="con144" tabindex="-1"></a> 11436 <div class="slot-title-line"> 11437 <span class="slot-title" 11438 >SEAIRD Model to Simulate the Impact of Human 11439 Behaviors</span 11440 > 11441 </div> 11442 <div class="slot-authors"> 11443 Aidan Fahlman and Gabriel Wainer (Carleton University) 11444 </div> 11445 <div class="slot-abstract"> 11446 <div> 11447 <a 11448 class="clickable no-decoration" 11449 id="vhsjs_view_240_1707793551_8569336" 11450 onclick="$('#vhsjs_view_240_1707793551_8569336').hide(); 11451 $('#vhsjs_hide_240_1707793551_8569336').show(); 11452 $('#239_1707793551_8569257').slideDown(function() { 11453 if (typeof Masonry === 'function') { 11454 $('.use_masonry').masonry(); 11455 }; 11456 11457 });" 11458 ><i class="fa fa-caret-right"></i> 11459 <span class="hover_link">Abstract</span></a 11460 ><a 11461 class="clickable no-decoration" 11462 id="vhsjs_hide_240_1707793551_8569336" 11463 onclick="$('#239_1707793551_8569257').hide(function() { 11464 if (typeof Masonry === 'function') { 11465 $('.use_masonry').masonry(); 11466 }; 11467 }); 11468 $('#vhsjs_hide_240_1707793551_8569336').hide(); 11469 $('#vhsjs_view_240_1707793551_8569336').show();" 11470 style="display: none" 11471 ><i class="fa fa-caret-down"></i> 11472 <span class="hover_link">Abstract</span></a 11473 > 11474 <div 11475 data-display-control="240_1707793551_8569336" 11476 id="239_1707793551_8569257" 11477 style="display: none" 11478 > 11479 <div class="arrow-slidedown"> 11480 <blockquote> 11481 Compartmental models have been utilized in the 11482 study and understanding of the COVID-19 11483 pandemic. Traditional models have been expanded 11484 to include geographical level transmission 11485 dynamics and new states. Here, we present a 11486 model based on Cell-DEVS specifications that can 11487 be used to define and study the effects of basic 11488 human behavior. We include mask wearing and 11489 lockdown fatigue, and an adaptable framework 11490 allowing for the rapid prototyping of different 11491 diseases and behaviors. We exemplify how to 11492 build the model and adapt the attributes using 11493 the provinces of Canada as a case study. The 11494 results show the effect mask mandates, mask 11495 wearing, and lockdown fatigue have on case 11496 counts over time. 11497 </blockquote> 11498 </div> 11499 </div> 11500 </div> 11501 </div> 11502 <div class="slot-urls"></div> 11503 <a href="/wsc23papers/098.pdf" target="_blank">pdf</a 11504 ><br /> 11505 </div> 11506 <div class="slot-entry"> 11507 <a name="inv207" tabindex="-1"></a> 11508 <div class="slot-title-line">
11509 <span class="slot-title" 11510 >A Compartmental Simulation Model to Improve 11511 Interventions for Controlling Poliovirus 11512 Outbreaks</span 11513 > 11514 </div> 11515 <div class="slot-authors"> 11516 Yuming Sun, Pinar Keskinocak, and Lauren Steimle 11517 (Georgia Institute of Technology) and Stephanie Kovacs 11518 and Steven Wassilak (Centers for Disease Control and 11519 Prevention) 11520 </div> 11521 <div class="slot-abstract"> 11522 <div> 11523 <a 11524 class="clickable no-decoration" 11525 id="vhsjs_view_242_1707793551_8593323" 11526 onclick="$('#vhsjs_view_242_1707793551_8593323').hide(); 11527 $('#vhsjs_hide_242_1707793551_8593323').show(); 11528 $('#241_1707793551_8593242').slideDown(function() { 11529 if (typeof Masonry === 'function') { 11530 $('.use_masonry').masonry(); 11531 }; 11532 11533 });" 11534 ><i class="fa fa-caret-right"></i> 11535 <span class="hover_link">Abstract</span></a 11536 ><a 11537 class="clickable no-decoration" 11538 id="vhsjs_hide_242_1707793551_8593323" 11539 onclick="$('#241_1707793551_8593242').hide(function() { 11540 if (typeof Masonry === 'function') { 11541 $('.use_masonry').masonry(); 11542 }; 11543 }); 11544 $('#vhsjs_hide_242_1707793551_8593323').hide(); 11545 $('#vhsjs_view_242_1707793551_8593323').show();" 11546 style="display: none" 11547 ><i class="fa fa-caret-down"></i> 11548 <span class="hover_link">Abstract</span></a 11549 > 11550 <div 11551 data-display-control="242_1707793551_8593323" 11552 id="241_1707793551_8593242" 11553 style="display: none" 11554 > 11555 <div class="arrow-slidedown"> 11556 <blockquote> 11557 Poliomyelitis (polio) is an infectious disease 11558 that paralyzed millions of people worldwide 11559 before polio vaccines were available. Despite 11560 the successes of the Global Polio Eradication 11561 Initiative, there are circulating 11562 vaccine-derived poliovirus outbreaks that 11563 require improved interventions. We built a 11564 compartmental model to simulate the spread of 11565 polio that considers mutation of the 11566 live-attenuated virus (in the oral polio 11567 vaccine) to evaluate the effectiveness of 11568 interventions. We validated the model in a case 11569 study of northern Nigeria and tested the impact 11570 of interventions that varied in the number of 11571 vaccination rounds and the target regions. 11572 Results indicated that the model captures polio 11573 dynamics by matching the case counts and their 11574 spatiotemporal and age distributions in the 11575 data. To stop the outbreaks, stakeholders should 11576 conduct aggressive interventions with more 11577 rounds and broader coverage, especially in the
11578 under-vaccinated regions, compared to the 11579 current practice. 11580 </blockquote> 11581 </div> 11582 </div> 11583 </div> 11584 </div> 11585 <div class="slot-urls"></div> 11586 <a href="/wsc23papers/099.pdf" target="_blank">pdf</a 11587 ><br /> 11588 </div> 11589 </div> 11590 <div class="session-entry"> 11591 <span class="session-event-type">Technical Session</span 11592 ><span class="type-track-spacer"> · </span 11593 ><span class="program-track" 11594 >Healthcare and Life Sciences</span 11595 ><br /> 11596 <div class="session-title">Healthcare Operations</div> 11597 <div class="session-chair"> 11598 Chair: Lambros Viennas (University of Surrey, Bridgnorth 11599 Aluminium Ltd.)<br /> 11600 </div> 11601 <div class="slot-entry"> 11602 <a name="con368" tabindex="-1"></a> 11603 <div class="slot-title-line"> 11604 <span class="slot-title" 11605 >Conceptual Modeling for Perishable Inventory: A Case 11606 Study in Human Milk Banking</span 11607 > 11608 </div> 11609 <div class="slot-authors"> 11610 Marta Staff and Navonil Mustafee (University of Exeter) 11611 and Natalie Shenker (Imperial College London) 11612 </div> 11613 <div class="slot-abstract"> 11614 <div> 11615 <a 11616 class="clickable no-decoration" 11617 id="vhsjs_view_244_1707793551_8652558" 11618 onclick="$('#vhsjs_view_244_1707793551_8652558').hide(); 11619 $('#vhsjs_hide_244_1707793551_8652558').show(); 11620 $('#243_1707793551_8652472').slideDown(function() { 11621 if (typeof Masonry === 'function') { 11622 $('.use_masonry').masonry(); 11623 }; 11624 11625 });" 11626 ><i class="fa fa-caret-right"></i> 11627 <span class="hover_link">Abstract</span></a 11628 ><a 11629 class="clickable no-decoration" 11630 id="vhsjs_hide_244_1707793551_8652558" 11631 onclick="$('#243_1707793551_8652472').hide(function() { 11632 if (typeof Masonry === 'function') { 11633 $('.use_masonry').masonry(); 11634 }; 11635 }); 11636 $('#vhsjs_hide_244_1707793551_8652558').hide(); 11637 $('#vhsjs_view_244_1707793551_8652558').show();" 11638 style="display: none" 11639 ><i class="fa fa-caret-down"></i> 11640 <span class="hover_link">Abstract</span></a 11641 > 11642 <div 11643 data-display-control="244_1707793551_8652558" 11644 id="243_1707793551_8652472" 11645 style="display: none" 11646 > 11647 <div class="arrow-slidedown"> 11648 <blockquote> 11649 The Conceptual Modeling (CM) stage of an M&S 11650 study focuses on developing an abstraction of 11651 the real world for subsequent implementation as 11652 a computer model. Several studies have 11653 acknowledged the importance of CM in the success 11654 of simulation projects. Yet, there is a lack of 11655 literature on applying CM frameworks to 11656 real-world case studies, which arguably impedes 11657 the translation of CM research into practice. In 11658 this paper, we present the development of a 11659 conceptual model, using Robinson’s CM 11660 framework, for our case study investigating the
11661 perishable product of human milk within the milk 11662 banking supply chain. We present the application 11663 of the various stages of the framework, 11664 reporting on stakeholder engagement, which has 11665 allowed us to develop a shared view of the CM. 11666 The paper adds to the literature on CM in 11667 practice, providing a detailed narrative on 11668 developing a conceptual model for perishable 11669 inventory management. 11670 </blockquote> 11671 </div> 11672 </div> 11673 </div> 11674 </div> 11675 <div class="slot-urls"></div> 11676 <a href="/wsc23papers/101.pdf" target="_blank">pdf</a 11677 ><br /> 11678 </div> 11679 <div class="slot-entry"> 11680 <a name="con225" tabindex="-1"></a> 11681 <div class="slot-title-line"> 11682 <span class="slot-title" 11683 >Clinical Pathway Clustering Using Surrogate 11684 Likelihoods and Replayability Validation</span 11685 > 11686 </div> 11687 <div class="slot-authors"> 11688 William Thomas Plumb, Alex Bottle, Giuliano Casale, and 11689 Alex Liddle (Imperial College London) 11690 </div> 11691 <div class="slot-abstract"> 11692 <div> 11693 <a 11694 class="clickable no-decoration" 11695 id="vhsjs_view_246_1707793551_8675966" 11696 onclick="$('#vhsjs_view_246_1707793551_8675966').hide(); 11697 $('#vhsjs_hide_246_1707793551_8675966').show(); 11698 $('#245_1707793551_8675883').slideDown(function() { 11699 if (typeof Masonry === 'function') { 11700 $('.use_masonry').masonry(); 11701 }; 11702 11703 });" 11704 ><i class="fa fa-caret-right"></i> 11705 <span class="hover_link">Abstract</span></a 11706 ><a 11707 class="clickable no-decoration" 11708 id="vhsjs_hide_246_1707793551_8675966" 11709 onclick="$('#245_1707793551_8675883').hide(function() { 11710 if (typeof Masonry === 'function') { 11711 $('.use_masonry').masonry(); 11712 }; 11713 }); 11714 $('#vhsjs_hide_246_1707793551_8675966').hide(); 11715 $('#vhsjs_view_246_1707793551_8675966').show();" 11716 style="display: none" 11717 ><i class="fa fa-caret-down"></i> 11718 <span class="hover_link">Abstract</span></a 11719 > 11720 <div 11721 data-display-control="246_1707793551_8675966" 11722 id="245_1707793551_8675883" 11723 style="display: none" 11724 > 11725 <div class="arrow-slidedown"> 11726 <blockquote> 11727 Modelling clinical pathways from Electronic 11728 Health Records (EHRs) can optimize resources and 11729 improve patient care, but current methods for 11730 generating pathway models using clustering have 11731 limitations including scalability and fidelity 11732 of the clusters. We propose a novel pathway 11733 modelling approach using Maximum Likelihood (ML) 11734 data clustering on Markov chain representations 11735 of clinical pathways. Our method is calibrated
11736 to produce clusters with low inter-cluster 11737 variability across the pathways. We use machine 11738 learning with Stochastic Radial Basis Functions 11739 (SRBF) kernels for surrogate optimization to 11740 handle non-convexity and propose an incremental 11741 optimization method to improve scalability. We 11742 also define a methodology based on novel 11743 replayability scores to help analysts compare 11744 the fidelity of alternative clustering results. 11745 Results show that our ML method produces 11746 clusters that have higher fidelity in terms of 11747 replayability scores than k-means based 11748 clustering and in capturing queueing contention, 11749 which is important for bottleneck identification 11750 in healthcare. 11751 </blockquote> 11752 </div> 11753 </div> 11754 </div> 11755 </div> 11756 <div class="slot-urls"></div> 11757 <a href="/wsc23papers/102.pdf" target="_blank">pdf</a 11758 ><br /> 11759 </div> 11760 </div> 11761 <div class="session-entry"> 11762 <span class="session-event-type">Technical Session</span 11763 ><span class="type-track-spacer"> · </span 11764 ><span class="program-track" 11765 >Healthcare and Life Sciences</span 11766 ><br /> 11767 <div class="session-title"> 11768 Applications of Simulation in Healthcare 11769 </div> 11770 <div class="session-chair"> 11771 Chair: Bjorn Berg (University of Minnesota)<br /> 11772 </div> 11773 <div class="slot-entry"> 11774 <a name="con195" tabindex="-1"></a> 11775 <div class="slot-title-line"> 11776 <span class="slot-title" 11777 >A Simulation Model and Dashboard for Predicting 11778 Covid-19 Bed Requirements</span 11779 > 11780 </div> 11781 <div> 11782 <span class="BAP award" 11783 >Best Contributed Applied Paper - Finalist</span 11784 > 11785 </div> 11786 <div class="slot-authors"> 11787 Yin-Chi Chan, Kaya Dreesbeimdiek, Ajith Kumar Parlikad, 11788 and Tom Ridgman (University of Cambridge); Nicholas J. 11789 Matheson and Ben Warne (University of Cambridge, 11790 Cambridge University Hospitals NHS Foundation Trust); 11791 and Denise Franks (Cambridge University Hospitals NHS 11792 Foundation Trust) 11793 </div> 11794 <div class="slot-abstract"> 11795 <div> 11796 <a 11797 class="clickable no-decoration" 11798 id="vhsjs_view_248_1707793551_8726454" 11799 onclick="$('#vhsjs_view_248_1707793551_8726454').hide(); 11800 $('#vhsjs_hide_248_1707793551_8726454').show(); 11801 $('#247_1707793551_8726373').slideDown(function() { 11802 if (typeof Masonry === 'function') { 11803 $('.use_masonry').masonry(); 11804 }; 11805 11806 });" 11807 ><i class="fa fa-caret-right"></i> 11808 <span class="hover_link">Abstract</span></a 11809 ><a 11810 class="clickable no-decoration" 11811 id="vhsjs_hide_248_1707793551_8726454" 11812 onclick="$('#247_1707793551_8726373').hide(function() { 11813 if (typeof Masonry === 'function') { 11814 $('.use_masonry').masonry(); 11815 }; 11816 }); 11817 $('#vhsjs_hide_248_1707793551_8726454').hide(); 11818 $('#vhsjs_view_248_1707793551_8726454').show();" 11819 style="display: none" 11820 ><i class="fa fa-caret-down"></i>
11821 <span class="hover_link">Abstract</span></a 11822 > 11823 <div 11824 data-display-control="248_1707793551_8726454" 11825 id="247_1707793551_8726373" 11826 style="display: none" 11827 > 11828 <div class="arrow-slidedown"> 11829 <blockquote> 11830 The Covid-19 pandemic has placed extraordinary 11831 amounts of stress upon public hospitals 11832 globally. This paper describes a simulation 11833 model for estimating hospital bed demand based 11834 on generated scenarios. Statistical tools were 11835 also developed for generating these scenarios, 11836 in particular, for fitting distributions to 11837 patients' lengths-of-stay and for predicting the 11838 number of daily arrivals of Covid-19 patients. A 11839 web dashboard has been created for ease of use. 11840 The simulation model and statistical tools have 11841 been used to estimate Covid-related bed demand 11842 at an NHS hospital in the East of England. 11843 </blockquote> 11844 </div> 11845 </div> 11846 </div> 11847 </div> 11848 <div class="slot-urls"></div> 11849 <a href="/wsc23papers/103.pdf" target="_blank">pdf</a 11850 ><br /> 11851 </div> 11852 <div class="slot-entry"> 11853 <a name="inv185" tabindex="-1"></a> 11854 <div class="slot-title-line"> 11855 <span class="slot-title" 11856 >Trajectory-Oriented Optimization of Stochastic 11857 Epidemiological Models</span 11858 > 11859 </div> 11860 <div class="slot-authors"> 11861 Arindam Fadikar (Argonne National Laboratory), Mickael 11862 Binois (Inria Centre at Université Côte 11863 d'Azur), Nicholson Collier and Abby Stevens (Argonne 11864 National Laboratory), Kok Ben Toh (Northwestern 11865 University), and Jonathan Ozik (Argonne National 11866 Laboratory) 11867 </div> 11868 <div class="slot-abstract"> 11869 <div> 11870 <a 11871 class="clickable no-decoration" 11872 id="vhsjs_view_250_1707793551_8751736" 11873 onclick="$('#vhsjs_view_250_1707793551_8751736').hide(); 11874 $('#vhsjs_hide_250_1707793551_8751736').show(); 11875 $('#249_1707793551_875165').slideDown(function() { 11876 if (typeof Masonry === 'function') { 11877 $('.use_masonry').masonry(); 11878 }; 11879 11880 });" 11881 ><i class="fa fa-caret-right"></i> 11882 <span class="hover_link">Abstract</span></a 11883 ><a 11884 class="clickable no-decoration" 11885 id="vhsjs_hide_250_1707793551_8751736" 11886 onclick="$('#249_1707793551_875165').hide(function() { 11887 if (typeof Masonry === 'function') { 11888 $('.use_masonry').masonry(); 11889 }; 11890 }); 11891 $('#vhsjs_hide_250_1707793551_8751736').hide(); 11892 $('#vhsjs_view_250_1707793551_8751736').show();" 11893 style="display: none" 11894 ><i class="fa fa-caret-down"></i>
11895 <span class="hover_link">Abstract</span></a 11896 > 11897 <div 11898 data-display-control="250_1707793551_8751736" 11899 id="249_1707793551_875165" 11900 style="display: none" 11901 > 11902 <div class="arrow-slidedown"> 11903 <blockquote> 11904 Epidemiological models must be calibrated to 11905 ground truth for downstream tasks such as 11906 producing forward projections or running what-if 11907 scenarios. The meaning of calibration changes in 11908 case of a stochastic model since output from 11909 such a model is generally described via an 11910 ensemble or a distribution. Each member of the 11911 ensemble is usually mapped to a random number 11912 seed (explicitly or implicitly). With the goal 11913 of finding not only the input parameter settings 11914 but also the random seeds that are consistent 11915 with the ground truth, we propose a class of 11916 Gaussian process (GP) surrogates along with an 11917 optimization strategy based on Thompson 11918 sampling. This Trajectory Oriented Optimization 11919 (TOO) approach produces actual trajectories 11920 close to the empirical observations instead of a 11921 set of parameter settings where only the mean 11922 simulation behavior matches with the ground 11923 truth. 11924 </blockquote> 11925 </div> 11926 </div> 11927 </div> 11928 </div> 11929 <div class="slot-urls"></div> 11930 <a href="/wsc23papers/104.pdf" target="_blank">pdf</a 11931 ><br /> 11932 </div> 11933 <div class="slot-entry"> 11934 <a name="inv171" tabindex="-1"></a> 11935 <div class="slot-title-line"> 11936 <span class="slot-title" 11937 >Modeling the Potential Impact of Community Health 11938 Volunteers in the Diagnosis and Treatment of Buruli 11939 Ulcer</span 11940 > 11941 </div> 11942 <div class="slot-authors"> 11943 Fatumah Atuhaire, Christine S. M. Currie, and Rebecca B. 11944 Hoyle (University of Southampton) 11945 </div> 11946 <div class="slot-abstract"> 11947 <div> 11948 <a 11949 class="clickable no-decoration" 11950 id="vhsjs_view_252_1707793551_8775272" 11951 onclick="$('#vhsjs_view_252_1707793551_8775272').hide(); 11952 $('#vhsjs_hide_252_1707793551_8775272').show(); 11953 $('#251_1707793551_8775194').slideDown(function() { 11954 if (typeof Masonry === 'function') { 11955 $('.use_masonry').masonry(); 11956 }; 11957 11958 });" 11959 ><i class="fa fa-caret-right"></i> 11960 <span class="hover_link">Abstract</span></a 11961 ><a 11962 class="clickable no-decoration" 11963 id="vhsjs_hide_252_1707793551_8775272" 11964 onclick="$('#251_1707793551_8775194').hide(function() { 11965 if (typeof Masonry === 'function') { 11966 $('.use_masonry').masonry(); 11967 }; 11968 }); 11969 $('#vhsjs_hide_252_1707793551_8775272').hide(); 11970 $('#vhsjs_view_252_1707793551_8775272').show();" 11971 style="display: none" 11972 ><i class="fa fa-caret-down"></i>
11973 <span class="hover_link">Abstract</span></a 11974 > 11975 <div 11976 data-display-control="252_1707793551_8775272" 11977 id="251_1707793551_8775194" 11978 style="display: none" 11979 > 11980 <div class="arrow-slidedown"> 11981 <blockquote> 11982 Buruli ulcer (BU) is a debilitating disease 11983 affecting the skin, soft tissue, and bone. It is 11984 the third most common mycobacterial disease in 11985 humans. The mode of transmission is not fully 11986 understood, posing challenges in prevention, and 11987 delayed diagnosis. One effective approach to 11988 promote early diagnosis and treatment is the 11989 utilization of community health volunteers 11990 (CHVs) for active case-finding. In this study, 11991 we developed an agent-based model to investigate 11992 the impact of CHVs in referring BU patients for 11993 treatment. We compared the effects of two 11994 strategies: offering self-referral alone versus 11995 self-referral combined with CHVs, on the early 11996 diagnosis and treatment of BU. Our findings 11997 confirm previous knowledge that integrating CHVs 11998 in active case-finding leads to earlier 11999 detection of BU cases, decreasing the number of 12000 individuals recovering with major disabilities. 12001 </blockquote> 12002 </div> 12003 </div> 12004 </div> 12005 </div> 12006 <div class="slot-urls"></div> 12007 <a href="/wsc23papers/105.pdf" target="_blank">pdf</a 12008 ><br /> 12009 </div> 12010 </div> 12011 </div> 12012 <div class="centered"> 12013 <div class="top-link"><a href="#top">Return to Top</a></div> 12014 </div> 12015 <hr /> 12016 </div> 12017 <div class="area-section"> 12018 <div class="centered"> 12019 <a name="ptrack105" tabindex="-1"></a> 12020 <div class="section-title">Hybrid Simulation</div> 12021 </div> 12022 <div class="centered track-chair"> 12023 <span class="track-chair-role" 12024 >Track Coordinator - Hybrid Simulation: </span 12025 ><span class="track-chair-names" 12026 >Anastasia Anagnostou (Brunel University London), Antuela 12027 Tako (Loughborough University)</span 12028 > 12029 </div> 12030 <div class="section-entry"> 12031 <div class="session-entry"> 12032 <span class="session-event-type">Technical Session</span 12033 ><span class="type-track-spacer"> · </span 12034 ><span class="program-track">Hybrid Simulation</span><br /> 12035 <div class="session-title"> 12036 Hybrid Simulation for Supply Chain Management 12037 </div> 12038 <div class="session-chair"> 12039 Chair: Anastasia Anagnostou (Brunel University London)<br /> 12040 </div> 12041 <div class="slot-entry"> 12042 <a name="con169" tabindex="-1"></a> 12043 <div class="slot-title-line"> 12044 <span class="slot-title" 12045 >Hybrid Discrete-Event Simulation with Repeated 12046 Machine Learning Prediction-Based Quality Inspection 12047 of Inbound Distribution Center Deliveries</span 12048 > 12049 </div> 12050 <div class="slot-authors"> 12051 Joost R. Remmelts and Alexander Hübl (University of 12052 Groningen) 12053 </div> 12054 <div class="slot-abstract"> 12055 <div> 12056 <a 12057 class="clickable no-decoration" 12058 id="vhsjs_view_254_1707793551_8852706" 12059 onclick="$('#vhsjs_view_254_1707793551_8852706').hide(); 12060 $('#vhsjs_hide_254_1707793551_8852706').show(); 12061 $('#253_1707793551_8852618').slideDown(function() { 12062 if (typeof Masonry === 'function') { 12063 $('.use_masonry').masonry(); 12064 }; 12065 12066 });" 12067 ><i class="fa fa-caret-right"></i>
12068 <span class="hover_link">Abstract</span></a 12069 ><a 12070 class="clickable no-decoration" 12071 id="vhsjs_hide_254_1707793551_8852706" 12072 onclick="$('#253_1707793551_8852618').hide(function() { 12073 if (typeof Masonry === 'function') { 12074 $('.use_masonry').masonry(); 12075 }; 12076 }); 12077 $('#vhsjs_hide_254_1707793551_8852706').hide(); 12078 $('#vhsjs_view_254_1707793551_8852706').show();" 12079 style="display: none" 12080 ><i class="fa fa-caret-down"></i> 12081 <span class="hover_link">Abstract</span></a 12082 > 12083 <div 12084 data-display-control="254_1707793551_8852706" 12085 id="253_1707793551_8852618" 12086 style="display: none" 12087 > 12088 <div class="arrow-slidedown"> 12089 <blockquote> 12090 Business-to-business distributors deem it 12091 necessary to inspect the quality of inbound 12092 deliveries to their distribution centers. This 12093 paper observes a company that experiences an 12094 inefficient quality inspection and wishes to 12095 improve the process. The broader 12096 product-receiving process is under-researched in 12097 warehousing literature but possesses 12098 similarities with manufacturing quality control. 12099 The paper aims to extend prediction-based 12100 quality inspection to the warehousing field. It 12101 applies a hybrid model, combining discrete-event 12102 simulation and machine learning multi-label 12103 classification to decrease the required 12104 inspection volume and evaluate its effects on 12105 the ability of the inspection and the workload 12106 and costs of distribution center operations. The 12107 results show that the inspection volume can 12108 drastically be decreased, reducing the workload 12109 and costs at the expense of the inspection 12110 capability of infrequently occurring delivery 12111 quality flaws in training data. The 12112 configuration of the classification model 12113 determines the degree of inspection volume 12114 reduction and wrongly predicted delivery quality 12115 flaws. 12116 </blockquote> 12117 </div> 12118 </div> 12119 </div> 12120 </div> 12121 <div class="slot-urls"></div> 12122 <a href="/wsc23papers/106.pdf" target="_blank">pdf</a 12123 ><br /> 12124 </div> 12125 <div class="slot-entry"> 12126 <a name="con271" tabindex="-1"></a> 12127 <div class="slot-title-line"> 12128 <span class="slot-title" 12129 >A Hybrid System Dynamics/Input-Output Model for 12130 Studying the Impact of Transportation Delays on the 12131 Resilience of National Supply Chains</span 12132 > 12133 </div> 12134 <div class="slot-authors"> 12135 William Steven Bland, Lissette Escobar, Andrew Hong, 12136 Grace Kenneally, A.J. Liberatore, and Scott Rosen (MITRE 12137 Corporation) 12138 </div> 12139 <div class="slot-abstract"> 12140 <div> 12141 <a 12142 class="clickable no-decoration"
12143 id="vhsjs_view_256_1707793551_8879318" 12144 onclick="$('#vhsjs_view_256_1707793551_8879318').hide(); 12145 $('#vhsjs_hide_256_1707793551_8879318').show(); 12146 $('#255_1707793551_887924').slideDown(function() { 12147 if (typeof Masonry === 'function') { 12148 $('.use_masonry').masonry(); 12149 }; 12150 12151 });" 12152 ><i class="fa fa-caret-right"></i> 12153 <span class="hover_link">Abstract</span></a 12154 ><a 12155 class="clickable no-decoration" 12156 id="vhsjs_hide_256_1707793551_8879318" 12157 onclick="$('#255_1707793551_887924').hide(function() { 12158 if (typeof Masonry === 'function') { 12159 $('.use_masonry').masonry(); 12160 }; 12161 }); 12162 $('#vhsjs_hide_256_1707793551_88793
1216218').hide(); 12163 $('#vhsjs_view_256_1707793551_8879318').show();" 12164 style="display: none" 12165 ><i class="fa fa-caret-down"></i> 12166 <span class="hover_link">Abstract</span></a 12167 > 12168 <div 12169 data-display-control="256_1707793551_8879318" 12170 id="255_1707793551_887924" 12171 style="display: none" 12172 > 12173 <div class="arrow-slidedown"> 12174 <blockquote> 12175 In today’s globally interconnected 12176 economy, transportation delays that impact a 12177 specific industry’s supply chain can 12178 quickly propagate to other industries, 12179 dramatically impacting inventory levels and 12180 economic production on the local, state, 12181 national, and global levels. This research 12182 proposes a hybrid System Dynamics and 12183 Input-Output simulation model that represents 12184 the impact of transportation delays on the flow 12185 of goods across industries and between 12186 geographic regions. The model is applied to a 12187 case study involving the port of Los Angeles to 12188 quantify the direct and indirect effects of a 12189 30- and 60-day delay in container movement on 12190 gross output across the 55 major industries in 12191 the United States. The capability to predict the 12192 scope and scale of the economic impact resulting 12193 from various transportation delays provides 12194 decision makers the opportunity to conduct 12195 preliminary what-if analyses which can support 12196 the development of potential mitigation 12197 strategies before the actual shock occurs. 12198 </blockquote> 12199 </div> 12200 </div> 12201 </div> 12202 </div> 12203 <div class="slot-urls"></div> 12204 <a href="/wsc23papers/107.pdf" target="_blank">pdf</a 12205 ><br /> 12206 </div> 12207 <div class="slot-entry"> 12208 <a name="con339" tabindex="-1"></a> 12209 <div class="slot-title-line"> 12210 <span class="slot-title" 12211 >Evaluating the Effectiveness of Countermeasures in 12212 ICT Supply Chains through Elicitation-Informed 12213 Simulation</span 12214 > 12215 </div> 12216 <div class="slot-authors"> 12217 Rong Lei, Samar Saleh, Weihong Grace Guo, Elsayed 12218 Elsayed, and Fred Roberts (Rutgers, The State University 12219 of New Jersey) and Paul Kantor (Paul B Kantor, 12220 Consultant) 12221 </div> 12222 <div class="slot-abstract"> 12223 <div> 12224 <a 12225 class="clickable no-decoration" 12226 id="vhsjs_view_258_1707793551_8902698" 12227 onclick="$('#vhsjs_view_258_1707793551_8902698').hide(); 12228 $('#vhsjs_hide_258_1707793551_8902698').show(); 12229 $('#257_1707793551_8902612').slideDown(function() { 12230 if (typeof Masonry === 'function') { 12231 $('.use_masonry').masonry(); 12232 }; 12233 12234 });" 12235 ><i class="fa fa-caret-right"></i>
12236 <span class="hover_link">Abstract</span></a 12237 ><a 12238 class="clickable no-decoration" 12239 id="vhsjs_hide_258_1707793551_8902698" 12240 onclick="$('#257_1707793551_8902612').hide(function() { 12241 if (typeof Masonry === 'function') { 12242 $('.use_masonry').masonry(); 12243 }; 12244 }); 12245 $('#vhsjs_hide_258_1707793551_8902698').hide(); 12246 $('#vhsjs_view_258_1707793551_8902698').show();" 12247 style="display: none" 12248 ><i class="fa fa-caret-down"></i> 12249 <span class="hover_link">Abstract</span></a 12250 > 12251 <div 12252 data-display-control="258_1707793551_8902698" 12253 id="257_1707793551_8902612" 12254 style="display: none" 12255 > 12256 <div class="arrow-slidedown"> 12257 <blockquote> 12258 Counterfeiting, the production of imitation 12259 goods, is a critical threat in the Information 12260 and Communication Technology (ICT) manufacturing 12261 supply chain (SC). Countermeasures (CMs) are 12262 strategies to mitigate disruptions and enhance a 12263 SC. We present a novel hybrid approach for 12264 assessing and selecting CMs in ICT SCs. Our 12265 model incorporates insights from subject matter 12266 experts (SME), via Delphi elicitation, into the 12267 simulation. This technique is used to study SC 12268 resilience against disruptions caused by 12269 counterfeiting. ICT is an integral part of our 12270 daily lives and life-supporting systems, making 12271 resilience against such threats vital. Using 12272 performance criteria including system service 12273 levels, delivery time, and product quality, our 12274 findings show the importance of integrating 12275 expert knowledge in simulation and the 12276 effectiveness of certain CMs. 12277 </blockquote> 12278 </div> 12279 </div> 12280 </div> 12281 </div> 12282 <div class="slot-urls"></div> 12283 <a href="/wsc23papers/108.pdf" target="_blank">pdf</a 12284 ><br /> 12285 </div> 12286 </div> 12287 <div class="session-entry"> 12288 <span class="session-event-type">Technical Session</span 12289 ><span class="type-track-spacer"> · </span 12290 ><span class="program-track">Hybrid Simulation</span><br /> 12291 <div class="session-title"> 12292 Hybrid Simulation in Manufacturing 12293 </div> 12294 <div class="session-chair"> 12295 Chair: Fernando Barros (University of Coimbra)<br /> 12296 </div> 12297 <div class="slot-entry"> 12298 <a name="con172" tabindex="-1"></a> 12299 <div class="slot-title-line"> 12300 <span class="slot-title" 12301 >Design of a Serious Game for Safety in Manufacturing 12302 Industry Using Hybrid Simulation Modeling: Towards 12303 Eliciting Risk Preferences</span 12304 > 12305 </div> 12306 <div class="slot-authors"> 12307 Hanane El Raoui and John Quigley (University of 12308 Strathclyde), Ayse Aslan and Gokula Vasantha (Edinburgh 12309 Napier University), Jack Hanson and Jonathan Corney 12310 (Edinburgh University), and Andrew Sherlock (National 12311 Manufacturing Institute Scotland/ University of 12312 Strathclyde) 12313 </div> 12314 <div class="slot-abstract"> 12315 <div> 12316 <a 12317 class="clickable no-decoration" 12318 id="vhsjs_view_260_1707793551_8974016" 12319 onclick="$('#vhsjs_view_260_1707793551_8974016').hide(); 12320 $('#vhsjs_hide_260_1707793551_8974016').show(); 12321 $('#259_1707793551_8973935').slideDown(function() { 12322 if (typeof Masonry === 'function') { 12323 $('.use_masonry').masonry(); 12324 }; 12325 12326 });" 12327 ><i class="fa fa-caret-right"></i>
12328 <span class="hover_link">Abstract</span></a 12329 ><a 12330 class="clickable no-decoration" 12331 id="vhsjs_hide_260_1707793551_8974016" 12332 onclick="$('#259_1707793551_8973935').hide(function() { 12333 if (typeof Masonry === 'function') { 12334 $('.use_masonry').masonry(); 12335 }; 12336 }); 12337 $('#vhsjs_hide_260_1707793551_8974016').hide(); 12338 $('#vhsjs_view_260_1707793551_8974016').show();" 12339 style="display: none" 12340 ><i class="fa fa-caret-down"></i> 12341 <span class="hover_link">Abstract</span></a 12342 > 12343 <div 12344 data-display-control="260_1707793551_8974016" 12345 id="259_1707793551_8973935" 12346 style="display: none" 12347 > 12348 <div class="arrow-slidedown"> 12349 <blockquote> 12350 Conventional methods used to elicit risk-taking 12351 preferences have demonstrated significant 12352 disparities with real-world behaviours, 12353 compromising the validity of the data collected. 12354 Serious gaming (SG) provides a high potential to 12355 bridge this gap. This paper presents a serious 12356 game as a novel approach to elicit 12357 risk-preference in an industrial manufacturing 12358 context, focusing on the game-design and 12359 implementation using hybrid simulation 12360 modelling. The developed SG serves as a tool for 12361 conducting incentivized experiments aimed at 12362 assessing human behaviour towards risk, to 12363 inform policy recommendations. The game 12364 incorporates two influential factors in shaping 12365 risk-taking behaviour in a manufacturing 12366 environment, namely the social learning and 12367 production pressure, and use a variety of game 12368 mechanics to promote the players’ 12369 motivation and engagement. A usability study was 12370 conducted with 10 participants using the 12371 Usability Scale System (SUS), to identify 12372 problems in the usability of the game. Results 12373 have shown that our game has a good usability. 12374 </blockquote> 12375 </div> 12376 </div> 12377 </div> 12378 </div> 12379 <div class="slot-urls"></div> 12380 <a href="/wsc23papers/109.pdf" target="_blank">pdf</a 12381 ><br /> 12382 </div> 12383 <div class="slot-entry"> 12384 <a name="con315" tabindex="-1"></a> 12385 <div class="slot-title-line"> 12386 <span class="slot-title" 12387 >Hybrid Simulation of Product Reconditioning: A Case 12388 Study</span 12389 > 12390 </div> 12391 <div class="slot-authors"> 12392 Sean McConville (Air Force Institute of Technology, 12393 University of North Texas); Suman Niranjan and 12394 Arunachalam Narayanan (University of North Texas); and 12395 Joseph Murray (Dayblink Consulting) 12396 </div> 12397 <div class="slot-abstract"> 12398 <div> 12399 <a 12400 class="clickable no-decoration" 12401 id="vhsjs_view_262_1707793551_8998568" 12402 onclick="$('#vhsjs_view_262_1707793551_8998568').hide(); 12403 $('#vhsjs_hide_262_1707793551_8998568').show(); 12404 $('#261_1707793551_8998487').slideDown(function() { 12405 if (typeof Masonry === 'function') { 12406 $('.use_masonry').masonry(); 12407 }; 12408 12409 });" 12410 ><i class="fa fa-caret-right"></i>
12411 <span class="hover_link">Abstract</span></a 12412 ><a 12413 class="clickable no-decoration" 12414 id="vhsjs_hide_262_1707793551_8998568" 12415 onclick="$('#261_1707793551_8998487').hide(function() { 12416 if (typeof Masonry === 'function') { 12417 $('.use_masonry').masonry(); 12418 }; 12419 }); 12420 $('#vhsjs_hide_262_1707793551_8998568').hide(); 12421 $('#vhsjs_view_262_1707793551_8998568').show();" 12422 style="display: none" 12423 ><i class="fa fa-caret-down"></i> 12424 <span class="hover_link">Abstract</span></a 12425 > 12426 <div 12427 data-display-control="262_1707793551_8998568" 12428 id="261_1707793551_8998487" 12429 style="display: none" 12430 > 12431 <div class="arrow-slidedown"> 12432 <blockquote> 12433 To gain economic competitive advantage from the 12434 closed loop supply chain (CLSC), firms must 12435 ensure that the cost of reconditioning products 12436 does not exceed the cost of purchasing new 12437 products. The uncertainties associated with 12438 product returns (i.e., product condition, 12439 quantity etc.) make it difficult for managers to 12440 efficiently allocate resources. This study 12441 develops and employs a hybrid simulation (HS) 12442 model as a decision support tool in a case study 12443 from industry. We demonstrate via our HS that 12444 the company could save significant money each 12445 quarter by converting their existing schedules 12446 from two shifts to single shifts and 12447 redistributing resources. Furthermore, we found 12448 maximizing the subprocess output doesn't 12449 necessarily reduce costs. The company's focus on 12450 output-oriented subprocess evaluation could 12451 impede cost-saving efforts. Future research will 12452 explore how the mix of new and returned items 12453 affects process yield, different resource 12454 configurations, prioritization of product types, 12455 and processing time disparities. 12456 </blockquote> 12457 </div> 12458 </div> 12459 </div> 12460 </div> 12461 <div class="slot-urls"></div> 12462 <a href="/wsc23papers/110.pdf" target="_blank">pdf</a 12463 ><br /> 12464 </div> 12465 <div class="slot-entry"> 12466 <a name="con378" tabindex="-1"></a> 12467 <div class="slot-title-line"> 12468 <span class="slot-title" 12469 >Virtual Planning of a Metal Additive Manufacturing 12470 Factory Using Techno-Economic Hybrid Simulation 12471 Models</span 12472 > 12473 </div> 12474 <div class="slot-authors"> 12475 Eldar Shakirov, Haden Quinlan, and A. John Hart 12476 (Massachusetts Institute of Technology) 12477 </div> 12478 <div class="slot-abstract"> 12479 <div> 12480 <a 12481 class="clickable no-decoration" 12482 id="vhsjs_view_264_1707793551_9020782" 12483 onclick="$('#vhsjs_view_264_1707793551_9020782').hide(); 12484 $('#vhsjs_hide_264_1707793551_9020782').show(); 12485 $('#263_1707793551_9020698').slideDown(function() { 12486 if (typeof Masonry === 'function') { 12487 $('.use_masonry').masonry(); 12488 }; 12489 12490 });" 12491 ><i class="fa fa-caret-right"></i>
12492 <span class="hover_link">Abstract</span></a 12493 ><a 12494 class="clickable no-decoration" 12495 id="vhsjs_hide_264_1707793551_9020782" 12496 onclick="$('#263_1707793551_9020698').hide(function() { 12497 if (typeof Masonry === 'function') { 12498 $('.use_masonry').masonry(); 12499 }; 12500 }); 12501 $('#vhsjs_hide_264_1707793551_9020782').hide(); 12502 $('#vhsjs_view_264_1707793551_9020782').show();" 12503 style="display: none" 12504 ><i class="fa fa-caret-down"></i> 12505 <span class="hover_link">Abstract</span></a 12506 > 12507 <div 12508 data-display-control="264_1707793551_9020782" 12509 id="263_1707793551_9020698" 12510 style="display: none" 12511 > 12512 <div class="arrow-slidedown"> 12513 <blockquote> 12514 Factory simulation can guide leaner production 12515 operations and resilient supply chains by 12516 informing capital allocation and real-time 12517 decision-making. This is especially true for 12518 emerging production methods, like additive 12519 manufacturing (AM), where a lack of expertise 12520 and relative technological novelty make it 12521 difficult to quantitatively assess technology 12522 economics across applications. While reported 12523 cost models provide detailed analysis on the AM 12524 printing process, accurate modeling requires 12525 specific evaluation of process-level and 12526 production-level considerations that 12527 significantly impact factory dynamics and cost. 12528 Advances in factory simulation modeling 12529 therefore promise the development of 12530 comprehensive and actionable cost models. This 12531 paper reviews progress in simulation-based 12532 costing, hybrid simulation, and automated model 12533 generation, and proposes an integrated approach 12534 for cost modeling using an AM-based factory. We 12535 demonstrate the feasibility of this approach by 12536 simulating the production of two common AM part 12537 geometries, and evaluate the associated cost and 12538 time performances of different factory 12539 configurations. 12540 </blockquote> 12541 </div> 12542 </div> 12543 </div> 12544 </div> 12545 <div class="slot-urls"></div> 12546 <a href="/wsc23papers/111.pdf" target="_blank">pdf</a 12547 ><br /> 12548 </div> 12549 </div> 12550 <div class="session-entry"> 12551 <span class="session-event-type">Technical Session</span 12552 ><span class="type-track-spacer"> · </span 12553 ><span class="program-track">Hybrid Simulation</span><br /> 12554 <div class="session-title"> 12555 Hybrid Simulation Methodology 12556 </div> 12557 <div class="session-chair"> 12558 Chair: Steffen Strassburger (Technische Universität 12559 Ilmenau)<br /> 12560 </div> 12561 <div class="slot-entry"> 12562 <a name="inv107" tabindex="-1"></a> 12563 <div class="slot-title-line">
12564 <span class="slot-title" 12565 >Choosing the Right Entity Size to Minimize 12566 Discretization Error in Discrete Event Simulation 12567 Models</span 12568 > 12569 </div> 12570 <div class="slot-authors"> 12571 Leonardo Chwif (IMT), Wilson Pereira (Simulate), and 12572 José Arnaldo Barra Montevechi (Federal University 12573 of Itajubá) 12574 </div> 12575 <div class="slot-abstract"> 12576 <div> 12577 <a 12578 class="clickable no-decoration" 12579 id="vhsjs_view_266_1707793551_9081979" 12580 onclick="$('#vhsjs_view_266_1707793551_9081979').hide(); 12581 $('#vhsjs_hide_266_1707793551_9081979').show(); 12582 $('#265_1707793551_9081893').slideDown(function() { 12583 if (typeof Masonry === 'function') { 12584 $('.use_masonry').masonry(); 12585 }; 12586 12587 });" 12588 ><i class="fa fa-caret-right"></i> 12589 <span class="hover_link">Abstract</span></a 12590 ><a 12591 class="clickable no-decoration" 12592 id="vhsjs_hide_266_1707793551_9081979" 12593 onclick="$('#265_1707793551_9081893').hide(function() { 12594 if (typeof Masonry === 'function') { 12595 $('.use_masonry').masonry(); 12596 }; 12597 }); 12598 $('#vhsjs_hide_266_1707793551_9081979').hide(); 12599 $('#vhsjs_view_266_1707793551_9081979').show();" 12600 style="display: none" 12601 ><i class="fa fa-caret-down"></i> 12602 <span class="hover_link">Abstract</span></a 12603 > 12604 <div 12605 data-display-control="266_1707793551_9081979" 12606 id="265_1707793551_9081893" 12607 style="display: none" 12608 > 12609 <div class="arrow-slidedown"> 12610 <blockquote> 12611 In discrete-event simulation models, the way we 12612 establish the relationship between a real-world 12613 object and the model entity (a single 12614 indivisible object flowing through the model) is 12615 crucial to some classes of problems due to 12616 possible computational unfeasibility. In 12617 addition, the entity size also relates to 12618 results accuracy and simulation running time - a 12619 subject barely explored in the literature. In 12620 this paper, these questions were investigated 12621 through case studies which supported our initial 12622 hypothesis about the general relationships 12623 involved. Then, a simple algorithm was developed 12624 for correctly choosing the best entity size to 12625 provide the desired accuracy, measured as a 12626 discretization error, with promising results. 12627 The limitations of the algorithm are addressed 12628 and some directions for future research are 12629 pointed. 12630 </blockquote> 12631 </div> 12632 </div> 12633 </div> 12634 </div> 12635 <div class="slot-urls"></div> 12636 <a href="/wsc23papers/112.pdf" target="_blank">pdf</a 12637 ><br /> 12638 </div> 12639 <div class="slot-entry"> 12640 <a name="inv108" tabindex="-1"></a> 12641 <div class="slot-title-line">
12642 <span class="slot-title" 12643 >How Not to Visualize Your Simulation Output 12644 Data</span 12645 > 12646 </div> 12647 <div class="slot-authors"> 12648 Jonas Genath (Ilmenau University of Technology) and 12649 Steffen Strassburger (Technische Universität 12650 Ilmenau) 12651 </div> 12652 <div class="slot-abstract"> 12653 <div> 12654 <a 12655 class="clickable no-decoration" 12656 id="vhsjs_view_268_1707793551_9105067" 12657 onclick="$('#vhsjs_view_268_1707793551_9105067').hide(); 12658 $('#vhsjs_hide_268_1707793551_9105067').show(); 12659 $('#267_1707793551_9104981').slideDown(function() { 12660 if (typeof Masonry === 'function') { 12661 $('.use_masonry').masonry(); 12662 }; 12663 12664 });" 12665 ><i class="fa fa-caret-right"></i> 12666 <span class="hover_link">Abstract</span></a 12667 ><a 12668 class="clickable no-decoration" 12669 id="vhsjs_hide_268_1707793551_9105067" 12670 onclick="$('#267_1707793551_9104981').hide(function() { 12671 if (typeof Masonry === 'function') { 12672 $('.use_masonry').masonry(); 12673 }; 12674 }); 12675 $('#vhsjs_hide_268_1707793551_9105067').hide(); 12676 $('#vhsjs_view_268_1707793551_9105067').show();" 12677 style="display: none" 12678 ><i class="fa fa-caret-down"></i> 12679 <span class="hover_link">Abstract</span></a 12680 > 12681 <div 12682 data-display-control="268_1707793551_9105067" 12683 id="267_1707793551_9104981" 12684 style="display: none" 12685 > 12686 <div class="arrow-slidedown"> 12687 <blockquote> 12688 Hybrid modeling and simulation studies combine 12689 well-defined methods from other disciplines with 12690 a simulation technique. Especially in the area 12691 of output data analysis of simulation studies, 12692 there is great potential for hybrid approaches 12693 that incorporate methods from machine learning 12694 and AI. For their successful application, the 12695 analytical capabilities of machine learning and 12696 AI must be combined with the interpretive 12697 capabilities of humans. In most cases, this 12698 connection is achieved through visualizations. 12699 As methods become more complicated, the demands 12700 on visualizations are increasing. In this paper, 12701 we conduct a data farming study and delve into 12702 the analysis of the result data. In doing so, we 12703 uncover typical errors in visualizations making 12704 the interpretation and evaluation of the data 12705 difficult or misleading. We then apply the 12706 concepts of visual analytics to these 12707 visualizations and derive general guidelines to 12708 help simulation users to analyze their 12709 simulation studies and present results 12710 unambiguously and clearly. 12711 </blockquote> 12712 </div> 12713 </div> 12714 </div> 12715 </div> 12716 <div class="slot-urls"></div> 12717 <a href="/wsc23papers/113.pdf" target="_blank">pdf</a 12718 ><br /> 12719 </div> 12720 <div class="slot-entry"> 12721 <a name="con355" tabindex="-1"></a> 12722 <div class="slot-title-line">
12723 <span class="slot-title" 12724 >Approximate Discrete-Event Method for Supervisory 12725 Control</span 12726 > 12727 </div> 12728 <div class="slot-authors"> 12729 Maaz Jamal and Gabriel Wainer (Carleton University) 12730 </div> 12731 <div class="slot-abstract"> 12732 <div> 12733 <a 12734 class="clickable no-decoration" 12735 id="vhsjs_view_270_1707793551_9125655" 12736 onclick="$('#vhsjs_view_270_1707793551_9125655').hide(); 12737 $('#vhsjs_hide_270_1707793551_9125655').show(); 12738 $('#269_1707793551_9125571').slideDown(function() { 12739 if (typeof Masonry === 'function') { 12740 $('.use_masonry').masonry(); 12741 }; 12742 12743 });" 12744 ><i class="fa fa-caret-right"></i> 12745 <span class="hover_link">Abstract</span></a 12746 ><a 12747 class="clickable no-decoration" 12748 id="vhsjs_hide_270_1707793551_9125655" 12749 onclick="$('#269_1707793551_9125571').hide(function() { 12750 if (typeof Masonry === 'function') { 12751 $('.use_masonry').masonry(); 12752 }; 12753 }); 12754 $('#vhsjs_hide_270_1707793551_9125655').hide(); 12755 $('#vhsjs_view_270_1707793551_9125655').show();" 12756 style="display: none" 12757 ><i class="fa fa-caret-down"></i> 12758 <span class="hover_link">Abstract</span></a 12759 > 12760 <div 12761 data-display-control="270_1707793551_9125655" 12762 id="269_1707793551_9125571" 12763 style="display: none" 12764 > 12765 <div class="arrow-slidedown"> 12766 <blockquote> 12767 Supervisory systems are used to and act when 12768 certain events are detected. Studying 12769 supervisory using formal Discrete Event 12770 Modelling & Simulation allows analyzing an 12771 application and then using the model to build 12772 the controllers. Supervisors can lead to a state 12773 space explosion if the model size increases, 12774 thus, reducing the state space complexity can 12775 expand the practicality of the model. We present 12776 a method based on Discrete Event System 12777 Specifications using an approximate method that 12778 reduces the state space complexity. The plant 12779 models and synthesized controllers can then be 12780 deployed on embedded hardware providing model 12781 continuity. We discuss the method and present a 12782 case study of a supervisory system. 12783 </blockquote> 12784 </div> 12785 </div> 12786 </div> 12787 </div> 12788 <div class="slot-urls"></div> 12789 <a href="/wsc23papers/114.pdf" target="_blank">pdf</a 12790 ><br /> 12791 </div> 12792 </div> 12793 <div class="session-entry"> 12794 <span class="session-event-type">Technical Session</span 12795 ><span class="type-track-spacer"> · </span 12796 ><span class="program-track">Hybrid Simulation</span><br /> 12797 <div class="session-title"> 12798 Hybrid Simulation Applications I 12799 </div> 12800 <div class="session-chair"> 12801 Chair: Navonil Mustafee (University of Exeter, The
12802 Business School)<br /> 12803 </div> 12804 <div class="slot-entry"> 12805 <a name="con174" tabindex="-1"></a> 12806 <div class="slot-title-line"> 12807 <span class="slot-title" 12808 >Smart Sports Predictions via Hybrid Simulation: NBA 12809 Case Study</span 12810 > 12811 </div> 12812 <div class="slot-authors"> 12813 Ignacio Erazo (Georgia Institute of Technology) 12814 </div> 12815 <div class="slot-abstract"> 12816 <div> 12817 <a 12818 class="clickable no-decoration" 12819 id="vhsjs_view_272_1707793551_9171655" 12820 onclick="$('#vhsjs_view_272_1707793551_9171655').hide(); 12821 $('#vhsjs_hide_272_1707793551_9171655').show(); 12822 $('#271_1707793551_9171476').slideDown(function() { 12823 if (typeof Masonry === 'function') { 12824 $('.use_masonry').masonry(); 12825 }; 12826 12827 });" 12828 ><i class="fa fa-caret-right"></i> 12829 <span class="hover_link">Abstract</span></a 12830 ><a 12831 class="clickable no-decoration" 12832 id="vhsjs_hide_272_1707793551_9171655" 12833 onclick="$('#271_1707793551_9171476').hide(function() { 12834 if (typeof Masonry === 'function') { 12835 $('.use_masonry').masonry(); 12836 }; 12837 }); 12838 $('#vhsjs_hide_272_1707793551_9171655').hide(); 12839 $('#vhsjs_view_272_1707793551_9171655').show();" 12840 style="display: none" 12841 ><i class="fa fa-caret-down"></i> 12842 <span class="hover_link">Abstract</span></a 12843 > 12844 <div 12845 data-display-control="272_1707793551_9171655" 12846 id="271_1707793551_9171476" 12847 style="display: none" 12848 > 12849 <div class="arrow-slidedown"> 12850 <blockquote> 12851 Increased data availability has stimulated the 12852 interest in studying sports prediction problems
12853 via analytical approaches; in particular, with 12854 machine learning and simulation. We characterize 12855 several models that have been proposed in the 12856 literature, all of which suffer from the same 12857 drawback: they cannot incorporate rational 12858 decision-making and strategies from 12859 teams/players effectively. We tackle this issue 12860 by proposing hybrid simulation logic that 12861 incorporates teams as agents, generalizing the 12862 models/methodologies that have been proposed in 12863 the past. We perform a case study on the NBA 12864 with two goals: i) study the quality of 12865 predictions when using only one predictive 12866 variable, and ii) study how much historical data 12867 should be kept to maximize prediction accuracy. 12868 Results indicate that there is an optimal range 12869 of data quantity and that studying what data and 12870 variables to include is of extreme importance. 12871 </blockquote> 12872 </div> 12873 </div> 12874 </div> 12875 </div> 12876 <div class="slot-urls"></div> 12877 <a href="/wsc23papers/115.pdf" target="_blank">pdf</a 12878 ><br /> 12879 </div> 12880 <div class="slot-entry"> 12881 <a name="con269" tabindex="-1"></a> 12882 <div class="slot-title-line"> 12883 <span class="slot-title" 12884 >Simulation Model to Forecast Gender Pension Wealth 12885 Gap in the Light of Demographic Changes</span 12886 > 12887 </div> 12888 <div class="slot-authors"> 12889 Bożena Mielczarek (Wroclaw University of Science 12890 and Technology) 12891 </div> 12892 <div class="slot-abstract"> 12893 <div> 12894 <a 12895 class="clickable no-decoration" 12896 id="vhsjs_view_274_1707793551_9193964" 12897 onclick="$('#vhsjs_view_274_1707793551_9193964').hide(); 12898 $('#vhsjs_hide_274_1707793551_9193964').show(); 12899 $('#273_1707793551_919388').slideDown(function() { 12900 if (typeof Masonry === 'function') { 12901 $('.use_masonry').masonry(); 12902 }; 12903 12904 });" 12905 ><i class="fa fa-caret-right"></i> 12906 <span class="hover_link">Abstract</span></a 12907 ><a 12908 class="clickable no-decoration" 12909 id="vhsjs_hide_274_1707793551_9193964" 12910 onclick="$('#273_1707793551_919388').hide(function() { 12911 if (typeof Masonry === 'function') { 12912 $('.use_masonry').masonry(); 12913 }; 12914 }); 12915 $('#vhsjs_hide_274_1707793551_9193964').hide(); 12916 $('#vhsjs_view_274_1707793551_9193964').show();" 12917 style="display: none" 12918 ><i class="fa fa-caret-down"></i> 12919 <span class="hover_link">Abstract</span></a 12920 > 12921 <div 12922 data-display-control="274_1707793551_9193964" 12923 id="273_1707793551_919388" 12924 style="display: none" 12925 > 12926 <div class="arrow-slidedown"> 12927 <blockquote> 12928 The ageing of the population has forced changes 12929 in many areas of social policy, including 12930 pension systems. Countries are reforming their 12931 retirement policies in such a way that the size 12932 of pension benefits depends on the total peri
12932od 12933 of employment, contributions made, and life 12934 expectancy. Due to the fact that in these types 12935 of system, employment plays a significant role 12936 in the accumulation of pension capital, a gender 12937 pay gap translates into a gender pension gap. In 12938 this article, we propose a hybrid simulation 12939 model to analyze the impact of long-term 12940 economic and demographic changes on the level of 12941 pension benefits when a worker retires, with a 12942 special focus on gender wealth pension gaps. The 12943 model combines demographic simulation conducted 12944 using a systems dynamics approach with discrete 12945 stochastic simulation by means of which we model 12946 the employment history of men and women. The 12947 model uses data from Polish statistical 12948 databases. 12949 </blockquote> 12950 </div> 12951 </div> 12952 </div> 12953 </div> 12954 <div class="slot-urls"></div> 12955 <a href="/wsc23papers/116.pdf" target="_blank">pdf</a 12956 ><br /> 12957 </div> 12958 <div class="slot-entry"> 12959 <a name="inv109" tabindex="-1"></a> 12960 <div class="slot-title-line"> 12961 <span class="slot-title" 12962 >Hybrid Simulation in Construction</span 12963 > 12964 </div> 12965 <div class="slot-authors"> 12966 Masoud Fakhimi (University of Surrey); Navonil Mustafee 12967 (University of Exeter, The Business School); and Tillal 12968 Eldabi (University of Bradford) 12969 </div> 12970 <div class="slot-abstract"> 12971 <div> 12972 <a 12973 class="clickable no-decoration" 12974 id="vhsjs_view_276_1707793551_9216435" 12975 onclick="$('#vhsjs_view_276_1707793551_9216435').hide(); 12976 $('#vhsjs_hide_276_1707793551_9216435').show(); 12977 $('#275_1707793551_9216354').slideDown(function() { 12978 if (typeof Masonry === 'function') { 12979 $('.use_masonry').masonry(); 12980 }; 12981 12982 });" 12983 ><i class="fa fa-caret-right"></i> 12984 <span class="hover_link">Abstract</span></a 12985 ><a 12986 class="clickable no-decoration" 12987 id="vhsjs_hide_276_1707793551_9216435" 12988 onclick="$('#275_1707793551_9216354').hide(function() { 12989 if (typeof Masonry === 'function') { 12990 $('.use_masonry').masonry(); 12991 }; 12992 }); 12993 $('#vhsjs_hide_276_1707793551_9216435').hide(); 12994 $('#vhsjs_view_276_1707793551_9216435').show();" 12995 style="display: none" 12996 ><i class="fa fa-caret-down"></i> 12997 <span class="hover_link">Abstract</span></a 12998 > 12999 <div 13000 data-display-control="276_1707793551_9216435" 13001 id="275_1707793551_9216354" 13002 style="display: none" 13003 > 13004 <div class="arrow-slidedown"> 13005 <blockquote> 13006 Hybrid Simulation (HS) is the application of 13007 multiple simulation techniques, for example, 13008 Discrete-event, Agent-based and System Dynamics, 13009 in the context of a single simulation study. HS 13010 is a growing area of research; numerous papers 13011 have delved into conceptualizations, frameworks, 13012 and case studies applied to specific application 13013 domains. The focus of our paper is on the 13014 construction domain. Through a systematic 13015 methodology for literature assessment, it 13016 presents a synthesis of the existing literature, 13017 providing insights on the choice of simulation 13018 technique, the context of its application, and 13019 the level of implementation, among others. 13020 Through an in-depth review of 36 relevant papers 13021 published over the past two decades, we 13022 contribute to a comprehensive understanding of 13023 the current state-of-the-art in HS as applied to 13024 Construction. The results of our investigation 13025 underscore the immense potential of HS in 13026 construction, with broad applicability spanning 13027 diverse areas such as structural analysis and 13028 building performance evaluation. 13029 </blockquote> 13030 </div> 13031 </div> 13032 </div> 13033 </div> 13034 <div class="slot-urls"></div> 13035 <a href="/wsc23papers/117.pdf" target="_blank">pdf</a 13036 ><br /> 13037 </div> 13038 </div> 13039 <div class="session-entry">
13040 <span class="session-event-type">Technical Session</span 13041 ><span class="type-track-spacer"> · </span 13042 ><span class="program-track">Hybrid Simulation</span><br /> 13043 <div class="session-title"> 13044 Hybrid Simulation Applications II 13045 </div> 13046 <div class="session-chair"> 13047 Chair: Tillal Eldabi (University of Bradford)<br /> 13048 </div> 13049 <div class="slot-entry"> 13050 <a name="con239" tabindex="-1"></a> 13051 <div class="slot-title-line"> 13052 <span class="slot-title" 13053 >Simulating Technician Populations with Tandem 13054 Analytic and Discrete Event Models</span 13055 > 13056 </div> 13057 <div class="slot-authors"> 13058 George Ryan Ambrose and Francois Alex Bourque (Defence 13059 Research and Development Canada) 13060 </div> 13061 <div class="slot-abstract"> 13062 <div> 13063 <a 13064 class="clickable no-decoration" 13065 id="vhsjs_view_278_1707793551_9258182" 13066 onclick="$('#vhsjs_view_278_1707793551_9258182').hide(); 13067 $('#vhsjs_hide_278_1707793551_9258182').show(); 13068 $('#277_1707793551_9258096').slideDown(function() { 13069 if (typeof Masonry === 'function') { 13070 $('.use_masonry').masonry(); 13071 }; 13072 13073 });" 13074 ><i class="fa fa-caret-right"></i> 13075 <span class="hover_link">Abstract</span></a 13076 ><a 13077 class="clickable no-decoration" 13078 id="vhsjs_hide_278_1707793551_9258182" 13079 onclick="$('#277_1707793551_9258096').hide(function() { 13080 if (typeof Masonry === 'function') { 13081 $('.use_masonry').masonry(); 13082 }; 13083 }); 13084 $('#vhsjs_hide_278_1707793551_9258182').hide(); 13085 $('#vhsjs_view_278_1707793551_9258182').show();" 13086 style="display: none" 13087 ><i class="fa fa-caret-down"></i> 13088 <span class="hover_link">Abstract</span></a 13089 > 13090 <div 13091 data-display-control="278_1707793551_9258182" 13092 id="277_1707793551_9258096" 13093 style="display: none" 13094 > 13095 <div class="arrow-slidedown"> 13096 <blockquote> 13097 Military workforce modelling is typically 13098 limited to either a series of analytic 13099 equations, or a simulation model. However, 13100 developing two such models in tandem has the 13101 benefit of cross-validation as well as the 13102 opportunity to explore problem space not easily 13103 accessed by a single approach. In particular, 13104 business rules for force employment are not 13105 easily described by closed-form equations while 13106 simulation models require exceedingly large 13107 computational resources to reach the asymptotic 13108 behaviour provided by analytic equations. This 13109 work leverages the benefits of both approaches 13110 to describe the population and career trends of 13111 technician individuals. As this career tends to 13112 have well defined training requirements, hence 13113 clear delineation between semi-functional 13114 apprentices and fully-functional journeymen, it 13115 is well suited to population modelling. Notional 13116 distributions for career parameters are assumed 13117 and the results for career progression and fleet 13118 readiness are compared. 13119 </blockquote> 13120 </div> 13121 </div> 13122 </div> 13123 </div> 13124 <div class="slot-urls"></div> 13125 <a href="/wsc23papers/118.pdf" target="_blank">pdf</a 13126 ><br /> 13127 </div> 13128 <div class="slot-entry"> 13129 <a name="inv172" tabindex="-1"></a> 13130 <div class="slot-title-line">
13131 <span class="slot-title" 13132 >πHyFlow: A Modular Process Interaction 13133 Worldview</span 13134 > 13135 </div> 13136 <div class="slot-authors"> 13137 Fernando Barros (University of Coimbra) 13138 </div> 13139 <div class="slot-abstract"> 13140 <div> 13141 <a 13142 class="clickable no-decoration" 13143 id="vhsjs_view_280_1707793551_927804" 13144 onclick="$('#vhsjs_view_280_1707793551_927804').hide(); 13145 $('#vhsjs_hide_280_1707793551_927804').show(); 13146 $('#279_1707793551_9277956').slideDown(function() { 13147 if (typeof Masonry === 'function') { 13148 $('.use_masonry').masonry(); 13149 }; 13150 13151 });" 13152 ><i class="fa fa-caret-right"></i> 13153 <span class="hover_link">Abstract</span></a 13154 ><a 13155 class="clickable no-decoration" 13156 id="vhsjs_hide_280_1707793551_927804" 13157 onclick="$('#279_1707793551_9277956').hide(function() { 13158 if (typeof Masonry === 'function') { 13159 $('.use_masonry').masonry(); 13160 }; 13161 }); 13162 $('#vhsjs_hide_280_1707793551_927804').hide(); 13163 $('#vhsjs_view_280_1707793551_927804').show();" 13164 style="display: none" 13165 ><i class="fa fa-caret-down"></i> 13166 <span class="hover_link">Abstract</span></a 13167 > 13168 <div 13169 data-display-control="280_1707793551_927804" 13170 id="279_1707793551_9277956" 13171 style="display: none" 13172 > 13173 <div class="arrow-slidedown"> 13174 <blockquote> 13175 Worldviews play a central role in M&S providing 13176 the basic constructs to describe simulation 13177 models. Three main worldviews have been defined: 13178 event scheduling, activity scanning, and process 13179 interaction (PI). The latter has been described 13180 in two flavors, one centered in the network of 13181 resources and other in the transitory 13182 transactions that flow in the network. In this 13183 paper we present a new M&S approach based on the 13184 πHYFLOW formalism that combines network and 13185 transaction PI, while keeping the support for 13186 modular and hierarchical models. We demonstrate 13187 πHYFLOW expressiveness by representing a 13188 hybrid production unit with a variable number of 13189 machines subjected to breakdowns. The hybrid 13190 model combines a fluid queue describing the 13191 work-in-progress, with discrete events modeling 13192 machines arrivals, departures, and breakdowns. 13193 Arrivals and departures of machines are achieved 13194 through modular communication, enabling model 13195 composition with other πHYFLOW components. 13196 </blockquote> 13197 </div> 13198 </div> 13199 </div> 13200 </div> 13201 <div class="slot-urls"></div> 13202 <a href="/wsc23papers/119.pdf" target="_blank">pdf</a 13203 ><br /> 13204 </div> 13205 </div> 13206 </div> 13207 <div class="centered"> 13208 <div class="top-link"><a href="#top">Return to Top</a></div> 13209 </div> 13210 <hr /> 13211 </div> 13212 <div class="area-section"> 13213 <div class="centered"> 13214 <a name="ptrack128" tabindex="-1"></a> 13215 <div class="section-title"> 13216 Logistics Supply Chains Transportation 13217 </div> 13218 </div> 13219 <div class="section-entry"> 13220 <div class="session-entry">
13221 <span class="session-event-type">Technical Session</span 13222 ><span class="type-track-spacer"> · </span 13223 ><span class="program-track" 13224 >Logistics Supply Chains Transportation</span 13225 ><br /> 13226 <div class="session-title">Automated Vehicles</div> 13227 <div class="session-chair"> 13228 Chair: Carles Serrat (Universitat Politècnica de 13229 Catalunya-BarcelonaTECH)<br /> 13230 </div> 13231 <div class="slot-entry"> 13232 <a name="inv112" tabindex="-1"></a> 13233 <div class="slot-title-line"> 13234 <span class="slot-title" 13235 >Simulating and Evaluating Internal Logistics 13236 Strategies for Suppliers in Just-in-Sequence Supply 13237 Systems in the Automotive Industry</span 13238 > 13239 </div> 13240 <div class="slot-authors"> 13241 Helen Christina Sand, Marvin Auf der Landwehr, and 13242 Christoph von Viebahn (Hochschule Hannover) 13243 </div> 13244 <div class="slot-abstract"> 13245 <div> 13246 <a 13247 class="clickable no-decoration" 13248 id="vhsjs_view_282_1707793551_93509" 13249 onclick="$('#vhsjs_view_282_1707793551_93509').hide(); 13250 $('#vhsjs_hide_282_1707793551_93509').show(); 13251 $('#281_1707793551_935082').slideDown(function() { 13252 if (typeof Masonry === 'function') { 13253 $('.use_masonry').masonry(); 13254 }; 13255 13256 });" 13257 ><i class="fa fa-caret-right"></i> 13258 <span class="hover_link">Abstract</span></a 13259 ><a 13260 class="clickable no-decoration" 13261 id="vhsjs_hide_282_1707793551_93509" 13262 onclick="$('#281_1707793551_935082').hide(function() { 13263 if (typeof Masonry === 'function') { 13264 $('.use_masonry').masonry(); 13265 }; 13266 }); 13267 $('#vhsjs_hide_282_1707793551_93509').hide(); 13268 $('#vhsjs_view_282_1707793551_93509').show();" 13269 style="display: none" 13270 ><i class="fa fa-caret-down"></i> 13271 <span class="hover_link">Abstract</span></a 13272 > 13273 <div 13274 data-display-control="282_1707793551_93509" 13275 id="281_1707793551_935082" 13276 style="display: none" 13277 > 13278 <div class="arrow-slidedown"> 13279 <blockquote> 13280 The reliability of just-in-sequence supply 13281 systems depends to a large extent on the 13282 efficiency of a supplier’s internal 13283 logistics distribution system. Thus, improving 13284 the logistics efficiency is a major objective 13285 for many suppliers in the automotive industry. 13286 In this paper, a discrete event simulation model 13287 is developed to evaluate the operational 13288 implications of different logistics strategies 13289 in just-in-sequence supply systems. Building 13290 upon the case of a major automotive supplier 13291 from Germany, the implications of various 13292 transportation resources and routing approaches 13293 are investigated and analyzed when it comes to 13294 the supply of components from an internal 13295 warehouse to the assembly lines. Experimental 13296 results show that the combined, 13297 load-carrier-specific use of forklifts, pallet 13298 trucks and tugger trains holds a high potential 13299 to achieve more efficient supply operations and 13300 meet different operational performance criteria 13301 such as downsizing the vehicle fleet, improving 13302 supply reliability and punctuality at the 13303 assembly lines, or minimizing warehouse traffic. 13304 </blockquote> 13305 </div> 13306 </div> 13307 </div> 13308 </div> 13309 <div class="slot-urls"></div> 13310 <a href="/wsc23papers/128.pdf" target="_blank">pdf</a 13311 ><br /> 13312 </div> 13313 <div class="slot-entry"> 13314 <a name="con150" tabindex="-1"></a> 13315 <div class="slot-title-line">
13316 <span class="slot-title" 13317 >Route Selection in Mixed Fleet Warehouses</span 13318 > 13319 </div> 13320 <div class="slot-authors"> 13321 Anna Rotondo (Irish Manufacturing Research) 13322 </div> 13323 <div class="slot-abstract"> 13324 <div> 13325 <a 13326 class="clickable no-decoration" 13327 id="vhsjs_view_284_1707793551_9371703" 13328 onclick="$('#vhsjs_view_284_1707793551_9371703').hide(); 13329 $('#vhsjs_hide_284_1707793551_9371703').show(); 13330 $('#283_1707793551_9371617').slideDown(function() { 13331 if (typeof Masonry === 'function') { 13332 $('.use_masonry').masonry(); 13333 }; 13334 13335 });" 13336 ><i class="fa fa-caret-right"></i> 13337 <span class="hover_link">Abstract</span></a 13338 ><a 13339 class="clickable no-decoration" 13340 id="vhsjs_hide_284_1707793551_9371703" 13341 onclick="$('#283_1707793551_9371617').hide(function() { 13342 if (typeof Masonry === 'function') { 13343 $('.use_masonry').masonry(); 13344 }; 13345 }); 13346 $('#vhsjs_hide_284_1707793551_9371703').hide(); 13347 $('#vhsjs_view_284_1707793551_9371703').show();" 13348 style="display: none" 13349 ><i class="fa fa-caret-down"></i> 13350 <span class="hover_link">Abstract</span></a 13351 > 13352 <div 13353 data-display-control="284_1707793551_9371703" 13354 id="283_1707793551_9371617" 13355 style="display: none" 13356 > 13357 <div class="arrow-slidedown"> 13358 <blockquote> 13359 Warehouse systems are progressively shifting 13360 towards mixed fleet models where automated and 13361 manually operated vehicles work together sharing 13362 the same floorspace. This is posing 13363 communication and co-ordination challenges from 13364 both a design and an operational perspective. 13365 Mixed fleet co-ordination is particularly 13366 challenging from a traffic control viewpoint due 13367 to the erratic behavior that human drivers may 13368 exhibit. In this work, an optimisation framework 13369 that aims at selecting the optimal route among 13370 candidate ones in a mixed fleet warehouse 13371 environment is developed. More specifically, the 13372 foundational deterministic components of the 13373 framework are described and an interactive 13374 dashboard used for verification purposes is 13375 presented. The development work of the 13376 stochastic component and the simulator is still 13377 ongoing. Initial feedback based on virtual 13378 testing conducted by an industrial partner 13379 suggests that a static optimisation approach 13380 based on historical traffic information may not 13381 lead to optimal choices when the human behavior 13382 is neglected. 13383 </blockquote> 13384 </div> 13385 </div> 13386 </div> 13387 </div> 13388 <div class="slot-urls"></div> 13389 <a href="/wsc23papers/129.pdf" target="_blank">pdf</a 13390 ><br /> 13391 </div> 13392 <div class="slot-entry"> 13393 <a name="con264" tabindex="-1"></a> 13394 <div class="slot-title-line">
13395 <span class="slot-title" 13396 >Modeling Autonomous Vehicle-Targeted Aggressive 13397 Merging Behaviors in Mixed Traffic Environment</span 13398 > 13399 </div> 13400 <div class="slot-authors"> 13401 JongIn Bae (Georgia Institute of Technology), Abhilasha 13402 Jairam Saroj (Oak Ridge National Laboratory), Wonho Suh 13403 (Hanyang University), and Michael P. Hunter and 13404 Angshuman Guin (Georgia Institute of Technology) 13405 </div> 13406 <div class="slot-abstract"> 13407 <div> 13408 <a 13409 class="clickable no-decoration" 13410 id="vhsjs_view_286_1707793551_9398165" 13411 onclick="$('#vhsjs_view_286_1707793551_9398165').hide(); 13412 $('#vhsjs_hide_286_1707793551_9398165').show(); 13413 $('#285_1707793551_9398084').slideDown(function() { 13414 if (typeof Masonry === 'function') { 13415 $('.use_masonry').masonry(); 13416 }; 13417 13418 });" 13419 ><i class="fa fa-caret-right"></i> 13420 <span class="hover_link">Abstract</span></a 13421 ><a 13422 class="clickable no-decoration" 13423 id="vhsjs_hide_286_1707793551_9398165" 13424 onclick="$('#285_1707793551_9398084').hide(function() { 13425 if (typeof Masonry === 'function') { 13426 $('.use_masonry').masonry(); 13427 }; 13428 }); 13429 $('#vhsjs_hide_286_1707793551_9398165').hide(); 13430 $('#vhsjs_view_286_1707793551_9398165').show();" 13431 style="display: none" 13432 ><i class="fa fa-caret-down"></i> 13433 <span class="hover_link">Abstract</span></a 13434 > 13435 <div 13436 data-display-control="286_1707793551_9398165" 13437 id="285_1707793551_9398084" 13438 style="display: none" 13439 > 13440 <div class="arrow-slidedown"> 13441 <blockquote> 13442 Advances in Autonomous Vehicle (AV) technology 13443 has fueled industry and research fields to 13444 dedicate significant effort to the study of the 13445 integration of AVs into the traffic network. 13446 This study focuses on the transition phase 13447 between all Human Driven Vehicles (HDVs) in the 13448 network to all AVs, where these different 13449 vehicle types coexist in a mixed traffic 13450 environment. This paper investigates the 13451 potential impacts of aggressive merging 13452 behaviors by human drivers on traffic 13453 performance in a mixed environment. For this, 13454 three vehicle types – AVs, HDVs, and 13455 Aggressive HDVs (AHDVs) are modeled in an 13456 open-source microscopic traffic simulation 13457 model, SUMO. In the developed simulation, the 13458 AHDVs are modeled to emulate aggressive merging 13459 behaviors in front of AVs at a merge section of 13460 a freeway exit ramp. Several experiments are 13461 used to study the impact of such behavior. 13462 Results show travel-time gains by AHDVs at the 13463 expense of AVs and HDVs. 13464 </blockquote> 13465 </div> 13466 </div> 13467 </div> 13468 </div> 13469 <div class="slot-urls"></div> 13470 <a href="/wsc23papers/130.pdf" target="_blank">pdf</a 13471 ><br /> 13472 </div> 13473 </div> 13474 <div class="session-entry">
13475 <span class="session-event-type">Technical Session</span 13476 ><span class="type-track-spacer"> · </span 13477 ><span class="program-track" 13478 >Logistics Supply Chains Transportation</span 13479 ><br /> 13480 <div class="session-title">New Approaches</div> 13481 <div class="session-chair"> 13482 Chair: Canan Gunes Corlu (Boston University)<br /> 13483 </div> 13484 <div class="slot-entry"> 13485 <a name="con245" tabindex="-1"></a> 13486 <div class="slot-title-line"> 13487 <span class="slot-title" 13488 >Estimating Parameters with Data Farming for 13489 Condition-Based Maintenance in a Digital Twin</span 13490 > 13491 </div> 13492 <div class="slot-authors"> 13493 Alexander Wuttke, Joachim Hunker, and Markus Rabe (TU 13494 Dortmund University) and Jan-Philipp Diepenbrock (IVA 13495 Schmetz GmbH) 13496 </div> 13497 <div class="slot-abstract"> 13498 <div> 13499 <a 13500 class="clickable no-decoration" 13501 id="vhsjs_view_288_1707793551_944561" 13502 onclick="$('#vhsjs_view_288_1707793551_944561').hide(); 13503 $('#vhsjs_hide_288_1707793551_944561').show(); 13504 $('#287_1707793551_9445527').slideDown(function() { 13505 if (typeof Masonry === 'function') { 13506 $('.use_masonry').masonry(); 13507 }; 13508 13509 });" 13510 ><i class="fa fa-caret-right"></i> 13511 <span class="hover_link">Abstract</span></a 13512 ><a 13513 class="clickable no-decoration" 13514 id="vhsjs_hide_288_1707793551_944561" 13515 onclick="$('#287_1707793551_9445527').hide(function() { 13516 if (typeof Masonry === 'function') { 13517 $('.use_masonry').masonry(); 13518 }; 13519 }); 13520 $('#vhsjs_hide_288_1707793551_944561').hide(); 13521 $('#vhsjs_view_288_1707793551_944561').show();" 13522 style="display: none" 13523 ><i class="fa fa-caret-down"></i> 13524 <span class="hover_link">Abstract</span></a 13525 > 13526 <div 13527 data-display-control="288_1707793551_944561" 13528 id="287_1707793551_9445527" 13529 style="display: none" 13530 > 13531 <div class="arrow-slidedown"> 13532 <blockquote> 13533 Nowadays, vast amounts of data can be collected 13534 by sensors and used for data-driven approaches. 13535 Digital twins provide a framework to exploit 13536 these data for solving various issues. For many 13537 companies in the industrial sector, machine 13538 maintenance is a significant issue. Maintenance 13539 is essential for high overall equipment 13540 efficiency, but it can also be c
13540ostly. 13541 Therefore, it should only be performed when 13542 necessary, based on the machine’s 13543 condition. Condition monitoring is used to 13544 assess a machine’s condition periodically, 13545 allowing for condition-based maintenance. In 13546 this paper, a simulation-based approach for 13547 parameter estimation is presented that 13548 contributes to condition-based maintenance. It 13549 introduces condition indicators for certain 13550 features of machines and demonstrates how to 13551 evaluate them using data farming, which employs 13552 simulation models as data generators. 13553 Additionally, the implementation of this 13554 approach in digital twins is discussed. 13555 </blockquote> 13556 </div> 13557 </div> 13558 </div> 13559 </div> 13560 <div class="slot-urls"></div> 13561 <a href="/wsc23papers/136.pdf" target="_blank">pdf</a 13562 ><br /> 13563 </div> 13564 <div class="slot-entry"> 13565 <a name="con350" tabindex="-1"></a> 13566 <div class="slot-title-line"> 13567 <span class="slot-title" 13568 >Approach for Classifying the Automatability of 13569 Verification and Validation Techniques</span 13570 > 13571 </div> 13572 <div class="slot-authors"> 13573 Katharina Langenbach and Markus Rabe (TU Dortmund 13574 University) 13575 </div> 13576 <div class="slot-abstract"> 13577 <div> 13578 <a 13579 class="clickable no-decoration" 13580 id="vhsjs_view_290_1707793551_9466534" 13581 onclick="$('#vhsjs_view_290_1707793551_9466534').hide(); 13582 $('#vhsjs_hide_290_1707793551_9466534').show(); 13583 $('#289_1707793551_9466453').slideDown(function() { 13584 if (typeof Masonry === 'function') { 13585 $('.use_masonry').masonry(); 13586 }; 13587 13588 });" 13589 ><i class="fa fa-caret-right"></i> 13590 <span class="hover_link">Abstract</span></a 13591 ><a 13592 class="clickable no-decoration" 13593 id="vhsjs_hide_290_1707793551_9466534" 13594 onclick="$('#289_1707793551_9466453').hide(function() { 13595 if (typeof Masonry === 'function') { 13596 $('.use_masonry').masonry(); 13597 }; 13598 }); 13599 $('#vhsjs_hide_290_1707793551_9466534').hide(); 13600 $('#vhsjs_view_290_1707793551_9466534').show();" 13601 style="display: none" 13602 ><i class="fa fa-caret-down"></i> 13603 <span class="hover_link">Abstract</span></a 13604 > 13605 <div 13606 data-display-control="290_1707793551_9466534" 13607 id="289_1707793551_9466453" 13608 style="display: none" 13609 > 13610 <div class="arrow-slidedown"> 13611 <blockquote> 13612 Simulation is a proven method in industry and 13613 research to constitute the basis for further 13614 decisions. Therefore, the credibility of its 13615 results is of major importance. Generally, 13616 simulation studies are guided by procedure 13617 models comprised of several phases with specific 13618 results. To assess the credibility, verification 13619 and validation (V&V) is used by applying V&V 13620 techniques to these phase results, which 13621 requires significant effort. Additionally, the 13622 amount of processed data increases and there is 13623 a growing desire for real-time-adjustable 13624 models, increasing the effort required for V&V 13625 while reducing the time available. One way to 13626 address these challenges is to automate V&V. For 13627 this purpose, the notions of automation and 13628 associated automation levels have to be 13629 transferred to the domain of V&V in order to 13630 assess and classify the automatability of 13631 individual V&V techniques. The effort for 13632 application of V&V techniques can be reduced 13633 while keeping or increasing the credibility of 13634 simulation. 13635 </blockquote> 13636 </div> 13637 </div> 13638 </div> 13639 </div> 13640 <div class="slot-urls"></div> 13641 <a href="/wsc23papers/138.pdf" target="_blank">pdf</a 13642 ><br /> 13643 </div> 13644 <div class="slot-entry"> 13645 <a name="con362" tabindex="-1"></a> 13646 <div class="slot-title-line">
13647 <span class="slot-title" 13648 >A Simulation-Based TDABC Model to Manage Supply Chain 13649 Costing: A Case Study</span 13650 > 13651 </div> 13652 <div class="slot-authors"> 13653 Siham Rahoui, John Crowe, and Amr Mahfouz (Technological 13654 University Dublin) 13655 </div> 13656 <div class="slot-abstract"> 13657 <div> 13658 <a 13659 class="clickable no-decoration" 13660 id="vhsjs_view_292_1707793551_9490407" 13661 onclick="$('#vhsjs_view_292_1707793551_9490407').hide(); 13662 $('#vhsjs_hide_292_1707793551_9490407').show(); 13663 $('#291_1707793551_9490268').slideDown(function() { 13664 if (typeof Masonry === 'function') { 13665 $('.use_masonry').masonry(); 13666 }; 13667 13668 });" 13669 ><i class="fa fa-caret-right"></i> 13670 <span class="hover_link">Abstract</span></a 13671 ><a 13672 class="clickable no-decoration" 13673 id="vhsjs_hide_292_1707793551_9490407" 13674 onclick="$('#291_1707793551_9490268').hide(function() { 13675 if (typeof Masonry === 'function') { 13676 $('.use_masonry').masonry(); 13677 }; 13678 }); 13679 $('#vhsjs_hide_292_1707793551_9490407').hide(); 13680 $('#vhsjs_view_292_1707793551_9490407').show();" 13681 style="display: none" 13682 ><i class="fa fa-caret-down"></i> 13683 <span class="hover_link">Abstract</span></a 13684 > 13685 <div 13686 data-display-control="292_1707793551_9490407" 13687 id="291_1707793551_9490268" 13688 style="display: none" 13689 > 13690 <div class="arrow-slidedown"> 13691 <blockquote>
13692 Effective management of supply chain costing is 13693 crucial for decision-making during times of 13694 disruption. It provides accurate cost 13695 indicators, enabling organizations to adapt to 13696 the risks of disruptions and mitigate their 13697 adverse effects. Supply chain costing literature 13698 has shown that traditional cost accounting 13699 approaches are inadequate in addressing the 13700 dynamic and complex nature of supply chain 13701 performance and the nonlinear behavior of the 13702 involved processes. Consequently, this paper 13703 presents a simulation-based supply chain costing 13704 framework that integrates discrete event 13705 simulation and time-driven activity-based 13706 costing to explore the dynamics of management 13707 accounting tools in a real context with all 13708 their complexities and interdependencies. The 13709 framework will be applied to the logistics 13710 function of an automotive supply chain to 13711 demonstrate the applicability of a static versus 13712 a dynamic time-driven activity-based costing 13713 model, their suitability to reflect the real 13714 operational performance of the supply chain and 13715 suggest ways to improve it. 13716 </blockquote> 13717 </div> 13718 </div> 13719 </div> 13720 </div> 13721 <div class="slot-urls"></div> 13722 <a href="/wsc23papers/137.pdf" target="_blank">pdf</a 13723 ><br /> 13724 </div> 13725 </div> 13726 <div class="session-entry"> 13727 <span class="session-event-type">Technical Session</span 13728 ><span class="type-track-spacer"> · </span 13729 ><span class="program-track" 13730 >Logistics Supply Chains Transportation</span 13731 ><br /> 13732 <div class="session-title"> 13733 Freight and Complex Supply Chains 13734 </div> 13735 <div class="session-chair"> 13736 Chair: Xueping Li (University of Tennessee)<br /> 13737 </div> 13738 <div class="slot-entry"> 13739 <a name="con242" tabindex="-1"></a> 13740 <div class="slot-title-line"> 13741 <span class="slot-title" 13742 >A Deep Q-Network Based on Radial Basis Functions for 13743 Multi-Echelon Inventory Management</span 13744 > 13745 </div> 13746 <div class="slot-authors"> 13747 Liqiang Cheng and Jun Luo (Shanghai Jiao Tong 13748 University), Weiwei Fan (Tongji University), and Yidong 13749 Zhang and Yuan Li (Alibaba) 13750 </div> 13751 <div class="slot-abstract"> 13752 <div> 13753 <a 13754 class="clickable no-decoration" 13755 id="vhsjs_view_294_1707793551_9541116" 13756 onclick="$('#vhsjs_view_294_1707793551_9541116').hide(); 13757 $('#vhsjs_hide_294_1707793551_9541116').show(); 13758 $('#293_1707793551_954103').slideDown(function() { 13759 if (typeof Masonry === 'function') { 13760 $('.use_masonry').masonry(); 13761 }; 13762 13763 });" 13764 ><i class="fa fa-caret-right"></i>
13765 <span class="hover_link">Abstract</span></a 13766 ><a 13767 class="clickable no-decoration" 13768 id="vhsjs_hide_294_1707793551_9541116" 13769 onclick="$('#293_1707793551_954103').hide(function() { 13770 if (typeof Masonry === 'function') { 13771 $('.use_masonry').masonry(); 13772 }; 13773 }); 13774 $('#vhsjs_hide_294_1707793551_9541116').hide(); 13775 $('#vhsjs_view_294_1707793551_9541116').show();" 13776 style="display: none" 13777 ><i class="fa fa-caret-down"></i> 13778 <span class="hover_link">Abstract</span></a 13779 > 13780 <div 13781 data-display-control="294_1707793551_9541116" 13782 id="293_1707793551_954103" 13783 style="display: none" 13784 > 13785 <div class="arrow-slidedown"> 13786 <blockquote> 13787 This paper addresses a multi-echelon inventory 13788 management problem with a complex network 13789 topology where deriving optimal ordering 13790 decisions is difficult. Deep reinforcement 13791 learning (DRL) has recently shown potential in 13792 solving such problems, while designing the 13793 neural networks in DRL remains a challenge. In 13794 order to address this, a DRL model is developed 13795 whose Q-network is based on radial basis 13796 functions. The approach can be more easily 13797 constructed compared to classic DRL models based 13798 on neural networks, thus alleviating the 13799 computational burden of hyperparameter tuning. 13800 Through a series of simulation experiments, the 13801 superior performance of this approach is 13802 demonstrated compared to the simple base-stock 13803 policy, producing a better policy in the 13804 multi-echelon system and competitive performance 13805 in the serial system where the base-stock policy 13806 is optimal. In addition, the approach 13807 outperforms current DRL approaches. 13808 </blockquote> 13809 </div> 13810 </div> 13811 </div> 13812 </div> 13813 <div class="slot-urls"></div> 13814 <a href="/wsc23papers/131.pdf" target="_blank">pdf</a 13815 ><br /> 13816 </div> 13817 <div class="slot-entry"> 13818 <a name="con155" tabindex="-1"></a> 13819 <div class="slot-title-line"> 13820 <span class="slot-title" 13821 >Simulation-based Cost Modeling to Measure the Effect 13822 of Automated Trucks in Inter-terminal Container 13823 Transportation</span 13824 > 13825 </div> 13826 <div class="slot-authors"> 13827 Ann-Kathrin Lange, Johannes Hinckeldeyn, Hendrik Rose, 13828 Nicole Nellen, and Michaela Grafelmann (Hamburg 13829 University of Technology) 13830 </div> 13831 <div class="slot-abstract"> 13832 <div> 13833 <a 13834 class="clickable no-decoration" 13835 id="vhsjs_view_296_1707793551_9565318" 13836 onclick="$('#vhsjs_view_296_1707793551_9565318').hide(); 13837 $('#vhsjs_hide_296_1707793551_9565318').show(); 13838 $('#295_1707793551_9565234').slideDown(function() { 13839 if (typeof Masonry === 'function') { 13840 $('.use_masonry').masonry(); 13841 }; 13842 13843 });" 13844 ><i class="fa fa-caret-right"></i>
13845 <span class="hover_link">Abstract</span></a 13846 ><a 13847 class="clickable no-decoration" 13848 id="vhsjs_hide_296_1707793551_9565318" 13849 onclick="$('#295_1707793551_9565234').hide(function() { 13850 if (typeof Masonry === 'function') { 13851 $('.use_masonry').masonry(); 13852 }; 13853 }); 13854 $('#vhsjs_hide_296_1707793551_9565318').hide(); 13855 $('#vhsjs_view_296_1707793551_9565318').show();" 13856 style="display: none" 13857 ><i class="fa fa-caret-down"></i> 13858 <span class="hover_link">Abstract</span></a 13859 > 13860 <div 13861 data-display-control="296_1707793551_9565318" 13862 id="295_1707793551_9565234" 13863 style="display: none" 13864 > 13865 <div class="arrow-slidedown"> 13866 <blockquote> 13867 Container transports within ports are 13868 characterized by mostly manual trucks and many 13869 handling operations in relatively small areas. 13870 Accordingly, they incur a disproportionately 13871 large cost in maritime transport chains. One way 13872 to reduce these costs is to use automated trucks 13873 in a port-internal transport system. Such 13874 systems have only been used on terminals, but 13875 not within whole ports. Thus, it is important to 13876 determine the design parameters of such 13877 transport systems. Discrete-event simulation is 13878 particularly suitable for investigating planned 13879 systems and controls in logistics. However, the 13880 costs of such systems are usually neglected. 13881 Therefore, a simulation-based cost model is used 13882 in this study to determine the 13883 cost-effectiveness of automated trucking 13884 systems. It is shown which factors possess the 13885 greatest influence on the cost-effectiveness of 13886 port-internal container transports. Furthermore, 13887 it can be estimated for the first time which 13888 cost savings can be achieved by using automated 13889 trucks for port-internal container transports. 13890 </blockquote> 13891 </div> 13892 </div> 13893 </div> 13894 </div> 13895 <div class="slot-urls"></div> 13896 <a href="/wsc23papers/132.pdf" target="_blank">pdf</a 13897 ><br /> 13898 </div> 13899 <div class="slot-entry"> 13900 <a name="con147" tabindex="-1"></a> 13901 <div class="slot-title-line"> 13902 <span class="slot-title" 13903 >Large Scale Logistics Network Simulation and Its 13904 Application in JD Logistics</span 13905 > 13906 </div> 13907 <div class="slot-authors"> 13908 Sheng Liu (Institute of Automation) and Xiaotian Zhuang, 13909 Liang Yan, Yu Wang, and Shengnan Wu (Jingdong Logistics) 13910 </div> 13911 <div class="slot-abstract"> 13912 <div> 13913 <a 13914 class="clickable no-decoration" 13915 id="vhsjs_view_298_1707793551_9588277" 13916 onclick="$('#vhsjs_view_298_1707793551_9588277').hide(); 13917 $('#vhsjs_hide_298_1707793551_9588277').show(); 13918 $('#297_1707793551_95882').slideDown(function() { 13919 if (typeof Masonry === 'function') { 13920 $('.use_masonry').masonry(); 13921 }; 13922 13923 });" 13924 ><i class="fa fa-caret-right"></i>
13925 <span class="hover_link">Abstract</span></a 13926 ><a 13927 class="clickable no-decoration" 13928 id="vhsjs_hide_298_1707793551_9588277" 13929 onclick="$('#297_1707793551_95882').hide(function() { 13930 if (typeof Masonry === 'function') { 13931 $('.use_masonry').masonry(); 13932 }; 13933 }); 13934 $('#vhsjs_hide_298_1707793551_9588277').hide(); 13935 $('#vhsjs_view_298_1707793551_9588277').show();" 13936 style="display: none" 13937 ><i class="fa fa-caret-down"></i> 13938 <span class="hover_link">Abstract</span></a 13939 > 13940 <div 13941 data-display-control="298_1707793551_9588277" 13942 id="297_1707793551_95882" 13943 style="display: none" 13944 > 13945 <div class="arrow-slidedown"> 13946 <blockquote> 13947 This paper proposes a large-scale logistics 13948 network simulation method to reduce package 13949 delivery delay and package loss caused by the 13950 sudden increase of package transportation demand 13951 during large-scale promotion activities such as 13952 11.11 and 6.18. We develop a large-scale 13953 logistics network simulation software for a 13954 large logistics enterprise. According to its 13955 actual logistics network, we establish its 13956 equivalent virtual logistics network in the 13957 simulation software. Then we simulate and adjust 13958 the virtual logistics network in advance. At 13959 last we regulate the actual logistics network 13960 according to the virtual network. As a result, 13961 we reduce the transportation time, the 13962 transportation distance, and the transportation 13963 costs for the logistics enterprise. The 13964 simulation software can complete the simulation 13965 of 500 million package distribution of a month 13966 in less than 30 minutes on a personal computer. 13967 </blockquote> 13968 </div> 13969 </div> 13970 </div> 13971 </div> 13972 <div class="slot-urls"></div> 13973 <a href="/wsc23papers/133.pdf" target="_blank">pdf</a 13974 ><br /> 13975 </div> 13976 </div> 13977 <div class="session-entry"> 13978 <span class="session-event-type">Technical Session</span 13979 ><span class="type-track-spacer"> · </span 13980 ><span class="program-track" 13981 >Logistics Supply Chains Transportation</span 13982 ><br /> 13983 <div class="session-title">Hybrid Models</div> 13984 <div class="session-chair"> 13985 Chair: Sahil Belsare (Walmart, Inc. USA; Northeastern 13986 University)<br /> 13987 </div> 13988 <div class="slot-entry"> 13989 <a name="inv110" tabindex="-1"></a> 13990 <div class="slot-title-line"> 13991 <span class="slot-title" 13992 >An Integrated System Dynamics and Discrete Event 13993 Supply Chain Simulation Framework for Supply Chain 13994 Resilience with Non-stationary Pandemic Demand</span 13995 > 13996 </div> 13997 <div class="slot-authors"> 13998 Mustafa Camur (GE Research); Chin-Yuan Tseng (Georgia 13999 Institute of Technology); Aristotelis E. Thanos (GE 14000 Research); Chelsea C. White (Georgia Institute of 14001 Technology); Walter Yund (GE Research); and Eleftherios 14002 Iakovou (Texas A&M University, Texas A&M Energy 14003 Institute) 14004 </div> 14005 <div class="slot-abstract"> 14006 <div> 14007 <a 14008 class="clickable no-decoration" 14009 id="vhsjs_view_300_1707793551_9637952" 14010 onclick="$('#vhsjs_view_300_1707793551_9637952').hide(); 14011 $('#vhsjs_hide_300_1707793551_9637952').show(); 14012 $('#299_1707793551_9637868').slideDown(function() { 14013 if (typeof Masonry === 'function') { 14014 $('.use_masonry').masonry(); 14015 }; 14016 14017 });" 14018 ><i class="fa fa-caret-right"></i>
14019 <span class="hover_link">Abstract</span></a 14020 ><a 14021 class="clickable no-decoration" 14022 id="vhsjs_hide_300_1707793551_9637952" 14023 onclick="$('#299_1707793551_9637868').hide(function() { 14024 if (typeof Masonry === 'function') { 14025 $('.use_masonry').masonry(); 14026 }; 14027 }); 14028 $('#vhsjs_hide_300_1707793551_9637952').hide(); 14029 $('#vhsjs_view_300_1707793551_9637952').show();" 14030 style="display: none" 14031 ><i class="fa fa-caret-down"></i> 14032 <span class="hover_link">Abstract</span></a 14033 > 14034 <div 14035 data-display-control="300_1707793551_9637952" 14036 id="299_1707793551_9637868" 14037 style="display: none" 14038 > 14039 <div class="arrow-slidedown"> 14040 <blockquote> 14041 COVID-19 resulted in some of the largest supply 14042 chain disruptions in recent history. To mitigate 14043 the impact of future disruptions, we propose an 14044 integrated hybrid simulation framework to couple 14045 nonstationary demand signals from an event like 14046 COVID-19 with a model of an end-to-e
14046nd supply 14047 chain. We first create a system dynamics 14048 susceptible-infected-recovered (SIR) model, 14049 augmenting a classic epidemiological model to 14050 create a realistic portrayal of demand patterns 14051 for oxygen concentrators (OC). Informed by this 14052 granular demand signal, we then create a supply 14053 chain discrete event simulation model of OC 14054 sourcing, manufacturing, and distribution to 14055 test production augmentation policies to satisfy 14056 this increased demand. This model utilizes 14057 publicly available data, engineering teardowns 14058 of OCs, and a supply chain illumination to 14059 identify suppliers. Our findings indicate that 14060 this coupled approach can use realistic demand 14061 during a disruptive event to enable rapid 14062 recommendations of policies for increased supply 14063 chain resilience with controlled cost. 14064 </blockquote> 14065 </div> 14066 </div> 14067 </div> 14068 </div> 14069 <div class="slot-urls"></div> 14070 <a href="/wsc23papers/134.pdf" target="_blank">pdf</a 14071 ><br /> 14072 </div> 14073 <div class="slot-entry"> 14074 <a name="con319" tabindex="-1"></a> 14075 <div class="slot-title-line"> 14076 <span class="slot-title" 14077 >Integrating a Mode Choice Model into Agent-based 14078 Simulation for Freight Transport Planning and 14079 Decarbonization Analysis</span 14080 > 14081 </div> 14082 <div class="slot-authors"> 14083 Senlei Wang, Dhanan Sarwo Utomo, and Philip Greening 14084 (Heriot-Watt University) 14085 </div> 14086 <div class="slot-abstract"> 14087 <div> 14088 <a 14089 class="clickable no-decoration" 14090 id="vhsjs_view_302_1707793551_9659855" 14091 onclick="$('#vhsjs_view_302_1707793551_9659855').hide(); 14092 $('#vhsjs_hide_302_1707793551_9659855').show(); 14093 $('#301_1707793551_9659777').slideDown(function() { 14094 if (typeof Masonry === 'function') { 14095 $('.use_masonry').masonry(); 14096 }; 14097 14098 });" 14099 ><i class="fa fa-caret-right"></i> 14100 <span class="hover_link">Abstract</span></a 14101 ><a 14102 class="clickable no-decoration" 14103 id="vhsjs_hide_302_1707793551_9659855" 14104 onclick="$('#301_1707793551_9659777').hide(function() { 14105 if (typeof Masonry === 'function') { 14106 $('.use_masonry').masonry(); 14107 }; 14108 }); 14109 $('#vhsjs_hide_302_1707793551_9659855').hide(); 14110 $('#vhsjs_view_302_1707793551_9659855').show();" 14111 style="display: none" 14112 ><i class="fa fa-caret-down"></i> 14113 <span class="hover_link">Abstract</span></a 14114 > 14115 <div 14116 data-display-control="302_1707793551_9659855" 14117 id="301_1707793551_9659777" 14118 style="display: none" 14119 > 14120 <div class="arrow-slidedown"> 14121 <blockquote> 14122 This paper presents a framework for integrating 14123 a discrete mode choice model with agent-based 14124 simulation. The integrated framework provides a 14125 more realistic representation of long-haul 14126 freight transport and is applied to the 14127 real-world scenarios of moving freight from 14128 ports to inland destinations via road, rail, and 14129 inland waterways. It incorporates a mode choice 14130 component that captures demand shifts between 14131 modes in response to different different policy 14132 and vehicle technology interventions. The 14133 objective is to investigate the financial and 14134 environmental impacts of introducing new vehicle 14135 technologies and associated energy sources under 14136 different future scenarios in a UK multimodal 14137 freight system. 14138 </blockquote> 14139 </div> 14140 </div> 14141 </div> 14142 </div> 14143 <div class="slot-urls"></div> 14144 <a href="/wsc23papers/135.pdf" target="_blank">pdf</a 14145 ><br /> 14146 </div> 14147 </div> 14148 <div class="session-entry">
14149 <span class="session-event-type">Technical Session</span 14150 ><span class="type-track-spacer"> · </span 14151 ><span class="program-track" 14152 >Logistics Supply Chains Transportation</span 14153 ><br /> 14154 <div class="session-title">Production Planning</div> 14155 <div class="session-chair"> 14156 Chair: Katharina Langenbach (TU Dortmund University)<br /> 14157 </div> 14158 <div class="slot-entry"> 14159 <a name="con197" tabindex="-1"></a> 14160 <div class="slot-title-line"> 14161 <span class="slot-title" 14162 >Improving Buffer Storage Performance in Ceramic Tile 14163 Industry via Simulation</span 14164 > 14165 </div> 14166 <div class="slot-authors"> 14167 Marco Taccini (University of Modena and Reggio Emilia); 14168 Giulia Dotti (University of Modena and Reggio Emilia, 14169 Marco Biagi Foundation); Manuel Iori (University of 14170 Modena and Reggio Emilia); and Anand Subramanian 14171 (Universidade Federal da Paraíba) 14172 </div> 14173 <div class="slot-abstract"> 14174 <div> 14175 <a 14176 class="clickable no-decoration" 14177 id="vhsjs_view_304_1707793551_9716818" 14178 onclick="$('#vhsjs_view_304_1707793551_9716818').hide(); 14179 $('#vhsjs_hide_304_1707793551_9716818').show(); 14180 $('#303_1707793551_971674').slideDown(function() { 14181 if (typeof Masonry === 'function') { 14182 $('.use_masonry').masonry(); 14183 }; 14184 14185 });" 14186 ><i class="fa fa-caret-right"></i> 14187 <span class="hover_link">Abstract</span></a 14188 ><a 14189 class="clickable no-decoration" 14190 id="vhsjs_hide_304_1707793551_9716818" 14191 onclick="$('#303_1707793551_971674').hide(function() { 14192 if (typeof Masonry === 'function') { 14193 $('.use_masonry').masonry(); 14194 }; 14195 }); 14196 $('#vhsjs_hide_304_1707793551_9716818').hide(); 14197 $('#vhsjs_view_304_1707793551_9716818').show();" 14198 style="display: none" 14199 ><i class="fa fa-caret-down"></i> 14200 <span class="hover_link">Abstract</span></a 14201 > 14202 <div 14203 data-display-control="304_1707793551_9716818" 14204 id="303_1707793551_971674" 14205 style="display: none" 14206 > 14207 <div class="arrow-slidedown"> 14208 <blockquote> 14209 This study aims at identifying the best strategy 14210 to temporarily store products within a buffer 14211 area in an Italian ceramic tile company. The 14212 storage policy is analyzed to maximize the 14213 storage capacity, facilitate operators' 14214 activities, and, consequently, improve the 14215 warehouse logistics performance. A discrete 14216 event simulation was conducted using Salabim, a 14217 Python based open-source software, in order to 14218 determine the best policy. We compare the 14219 performance of the current storage policy, based 14220 on technical production properties of products, 14221 and a newly proposed one, based on products' 14222 downstream destination. The results suggested 14223 that the proposed strategy significantly 14224 improves the performance of the buffer area 14225 management. The approach can be applied to 14226 different applications, contributing to the 14227 literature on simulation-based decision-making 14228 in material management. Furthermore, the study 14229 provides a functional case study showing the 14230 potential and achievable results of Salabim for 14231 modeling complex systems. 14232 </blockquote> 14233 </div> 14234 </div> 14235 </div> 14236 </div> 14237 <div class="slot-urls"></div> 14238 <a href="/wsc23papers/139.pdf" target="_blank">pdf</a 14239 ><br /> 14240 </div> 14241 <div class="slot-entry"> 14242 <a name="con135" tabindex="-1"></a> 14243 <div class="slot-title-line">
14244 <span class="slot-title" 14245 >Simulating the Impact of Forecast related Overbooking 14246 and Underbooking Behavior on MRP Planning and a 14247 Reorder Point System</span 14248 > 14249 </div> 14250 <div class="slot-authors"> 14251 Wolfgang Seiringer and Klaus Altendorfer (University of 14252 Applied Sciences Upper Austria) and Thomas Felberbauer 14253 (University of Applied Sciences St. Pölten) 14254 </div> 14255 <div class="slot-abstract"> 14256 <div> 14257 <a 14258 class="clickable no-decoration" 14259 id="vhsjs_view_306_1707793551_97483" 14260 onclick="$('#vhsjs_view_306_1707793551_97483').hide(); 14261 $('#vhsjs_hide_306_1707793551_97483').show(); 14262 $('#305_1707793551_9748216').slideDown(function() { 14263 if (typeof Masonry === 'function') { 14264 $('.use_masonry').masonry(); 14265 }; 14266 14267 });" 14268 ><i class="fa fa-caret-right"></i> 14269 <span class="hover_link">Abstract</span></a 14270 ><a 14271 class="clickable no-decoration" 14272 id="vhsjs_hide_306_1707793551_97483" 14273 onclick="$('#305_1707793551_9748216').hide(function() { 14274 if (typeof Masonry === 'function') { 14275 $('.use_masonry').masonry(); 14276 }; 14277 }); 14278 $('#vhsjs_hide_306_1707793551_97483').hide(); 14279 $('#vhsjs_view_306_1707793551_97483').show();" 14280 style="display: none" 14281 ><i class="fa fa-caret-down"></i> 14282 <span class="hover_link">Abstract</span></a 14283 > 14284 <div 14285 data-display-control="306_1707793551_97483" 14286 id="305_1707793551_9748216" 14287 style="display: none" 14288 > 14289 <div class="arrow-slidedown"> 14290 <blockquote> 14291 Production Planning and its parameterization is 14292 critical to fulfil customer demands and to 14293 successfully react on changes in high volatile 14294 markets. Therefore, demand updates should be 14295 considered to improve production planning. In 14296 this paper the performance of two production 14297 planning methods MRP (Material Requirements 14298 Planning) and RPS (Reorder Point System) are 14299 compared in a multi-item single stage system 14300 where customer orders are updated in a rolling 14301 horizon manner. Applying a simulation study, we 14302 investigate the performance of MRP and RPS for 14303 biased and unbiased forecast information and 14304 discuss the difference in the optimal planning 14305 parameters. The study shows that for a 14306 production system with underbooking and low 14307 demand uncertainty, RPS method is superior, in 14308 all other scenarios MRP outperforms RPS. For 14309 overbooking scenarios, the results show that MRP 14310 leads to overall cost improvements ranging from 14311 8% to 30%. 14312 </blockquote> 14313 </div> 14314 </div> 14315 </div> 14316 </div> 14317 <div class="slot-urls"></div> 14318 <a href="/wsc23papers/140.pdf" target="_blank">pdf</a 14319 ><br /> 14320 </div> 14321 <div class="slot-entry"> 14322 <a name="con226" tabindex="-1"></a> 14323 <div class="slot-title-line">
14324 <span class="slot-title" 14325 >Pick Order Assignment and Order Batching Strategy for 14326 Robotic Mobile Fulfilment System Warehouse</span 14327 > 14328 </div> 14329 <div class="slot-authors"> 14330 Shuo-Yan Chou, Aisyahna Nurul Mauliddina, Anindhita 14331 Dewabharata, and Ferani Eva Zulvia (National Taiwan 14332 University of Science and Technology) 14333 </div> 14334 <div class="slot-abstract"> 14335 <div> 14336 <a 14337 class="clickable no-decoration" 14338 id="vhsjs_view_308_1707793551_9771905" 14339 onclick="$('#vhsjs_view_308_1707793551_9771905').hide(); 14340 $('#vhsjs_hide_308_1707793551_9771905').show(); 14341 $('#307_1707793551_977182').slideDown(function() { 14342 if (typeof Masonry === 'function') { 14343 $('.use_masonry').masonry(); 14344 }; 14345 14346 });" 14347 ><i class="fa fa-caret-right"></i> 14348 <span class="hover_link">Abstract</span></a 14349 ><a 14350 class="clickable no-decoration" 14351 id="vhsjs_hide_308_1707793551_9771905" 14352 onclick="$('#307_1707793551_977182').hide(function() { 14353 if (typeof Masonry === 'function') { 14354 $('.use_masonry').masonry(); 14355 }; 14356 }); 14357 $('#vhsjs_hide_308_1707793551_9771905').hide(); 14358 $('#vhsjs_view_308_1707793551_9771905').show();" 14359 style="display: none" 14360 ><i class="fa fa-caret-down"></i> 14361 <span class="hover_link">Abstract</span></a 14362 > 14363 <div 14364 data-display-control="308_1707793551_9771905" 14365 id="307_1707793551_977182" 14366 style="display: none" 14367 > 14368 <div class="arrow-slidedown"> 14369 <blockquote> 14370 This study aims to optimize the order 14371 fulfillment process in a Robotic Mobile 14372 Fulfilment System warehouse by improving the 14373 order batching and the pick order assignment in 14374 order-picking activities using a simulation 14375 approach. The order-to-station assignment 14376 considers the association between the new order 14377 and the in-progress order at the station instead 14378 of random assignment. The proposed model aims to 14379 maximize the total throughput, maximize the 14380 pile-on value, and minimize the required number 14381 of pods. The proposed model is compared with a 14382 baseline scenario. The result shows that the 14383 proposed model significantly decreases the 14384 number of required pods by 40%, increases the 14385 pile-on by 60%, and increases the throughput by 14386 4%. This result proves that the proposed 14387 strategy can improve the efficiency of the 14388 order-picking process by ensuring every order 14389 and/or batch of orders always goes to the 14390 picking station with the most similar order. 14391 </blockquote> 14392 </div> 14393 </div> 14394 </div> 14395 </div> 14396 <div class="slot-urls"></div> 14397 <a href="/wsc23papers/141.pdf" target="_blank">pdf</a 14398 ><br /> 14399 </div> 14400 </div> 14401 <div class="session-entry">
14402 <span class="session-event-type">Technical Session</span 14403 ><span class="type-track-spacer"> · </span 14404 ><span class="program-track" 14405 >Logistics Supply Chains Transportation</span 14406 ><br /> 14407 <div class="session-title">Risks and Resilience</div> 14408 <div class="session-chair"> 14409 Chair: Joachim Hunker (Technische Universität 14410 Dortmund)<br /> 14411 </div> 14412 <div class="slot-entry"> 14413 <a name="cea106" tabindex="-1"></a> 14414 <div class="slot-title-line"> 14415 <span class="slot-title" 14416 >A Supply Chain Resilience Case Study Linking Key 14417 Resilience Areas with Process Mining</span 14418 > 14419 </div> 14420 <div class="slot-authors"> 14421 Frank Schätter, Florian Haas, and Frank Morelli 14422 (Pforzheim University of Applied Sciences) 14423 </div> 14424 <div class="slot-abstract"> 14425 <div> 14426 <a 14427 class="clickable no-decoration" 14428 id="vhsjs_view_310_1707793551_9824488" 14429 onclick="$('#vhsjs_view_310_1707793551_9824488').hide(); 14430 $('#vhsjs_hide_310_1707793551_9824488').show(); 14431 $('#309_1707793551_9824402').slideDown(function() { 14432 if (typeof Masonry === 'function') { 14433 $('.use_masonry').masonry(); 14434 }; 14435 14436 });" 14437 ><i class="fa fa-caret-right"></i> 14438 <span class="hover_link">Abstract</span></a 14439 ><a 14440 class="clickable no-decoration" 14441 id="vhsjs_hide_310_1707793551_9824488" 14442 onclick="$('#309_1707793551_9824402').hide(function() { 14443 if (typeof Masonry === 'function') { 14444 $('.use_masonry').masonry(); 14445 }; 14446 }); 14447 $('#vhsjs_hide_310_1707793551_9824488').hide(); 14448 $('#vhsjs_view_310_1707793551_9824488').show();" 14449 style="display: none" 14450 ><i class="fa fa-caret-down"></i> 14451 <span class="hover_link">Abstract</span></a 14452 > 14453 <div 14454 data-display-control="310_1707793551_9824488" 14455 id="309_1707793551_9824402" 14456 style="display: none" 14457 > 14458 <div class="arrow-slidedown"> 14459 <blockquote> 14460 At a time when supply chain disruptions are on 14461 the rise, supply chain managers are often 14462 overwhelmed by a simple question: How resilient 14463 is my supply chain and how can the status quo be 14464 improved? We present a case study of a 14465 manufacturing company in Central Europe that 14466 uses a two-step approach to help managers answer 14467 these questions. In the first stage, Key 14468 Resilience Areas (KRAs) are applied to 14469 transactional data to identify critical elements 14470 of the supply chain and their potential impacts. 14471 In the second stage, process mining is used to 14472 analyze the root causes of the identified 14473 impacts. In the case study, we reveal vulnerable 14474 locations and relevant product characteristics 14475 of the material flows of the company's inbound 14476 network, and process mining is used to analyze 14477 why, for example, a single sourcing strategy was 14478 chosen for a critical supplier. 14479 </blockquote> 14480 </div> 14481 </div> 14482 </div> 14483 </div> 14484 <div class="slot-urls"></div> 14485 <a href="/wsc23papers/cea106.pdf" target="_blank">pdf</a 14486 ><br /> 14487 </div> 14488 <div class="slot-entry"> 14489 <a name="cea118" tabindex="-1"></a> 14490 <div class="slot-title-line">
14491 <span class="slot-title" 14492 >Conceptualizing Resilience in Supply Chain 14493 Simulation</span 14494 > 14495 </div> 14496 <div class="slot-authors"> 14497 Simon Taylor, Anastasia Anagnostou, and Kate Mintram 14498 (Brunel University London) and Ed Hua, Andreas Tolk, 14499 Mark Pfaff, and David Mendonca (MITRE Corporation) 14500 </div> 14501 <div class="slot-abstract"> 14502 <div> 14503 <a 14504 class="clickable no-decoration" 14505 id="vhsjs_view_312_1707793551_9847841" 14506 onclick="$('#vhsjs_view_312_1707793551_9847841').hide(); 14507 $('#vhsjs_hide_312_1707793551_9847841').show(); 14508 $('#311_1707793551_984776').slideDown(function() { 14509 if (typeof Masonry === 'function') { 14510 $('.use_masonry').masonry(); 14511 }; 14512 14513 });" 14514 ><i class="fa fa-caret-right"></i> 14515 <span class="hover_link">Abstract</span></a 14516 ><a 14517 class="clickable no-decoration" 14518 id="vhsjs_hide_312_1707793551_9847841" 14519 onclick="$('#311_1707793551_984776').hide(function() { 14520 if (typeof Masonry === 'function') { 14521 $('.use_masonry').masonry(); 14522 }; 14523 }); 14524 $('#vhsjs_hide_312_1707793551_9847841').hide(); 14525 $('#vhsjs_view_312_1707793551_9847841').show();" 14526 style="display: none" 14527 ><i class="fa fa-caret-down"></i> 14528 <span class="hover_link">Abstract</span></a 14529 > 14530 <div 14531 data-display-control="312_1707793551_9847841" 14532 id="311_1707793551_984776" 14533 style="display: none" 14534 > 14535 <div class="arrow-slidedown"> 14536 <blockquote> 14537 Supply chains (SCs) collaborate in production 14538 and consumption across the world. SC management 14539 techniques attempt to optimize and balance 14540 supply chain operations. SC simulation can help 14541 support this by exploring “what-if” 14542 scenarios across key performance indicators, 14543 particularly when SCs are subject to potentially 14544 disruptive events. Resilience is the capacity 14545 for an enterprise to survive, adapt, and grow in 14546 the face of turbulent change. Change engenders 14547 SC vulnerabilities and management control 14548 attempts to create SC capabilities to address 14549 them. We are investigating the feasibility of 14550 creating a generic SC Simulation framework that 14551 represents sources of vulnerability and 14552 resilience and allows decision makers to explore 14553 potential capabilities to address them. This 14554 article reports progress on the first step of 14555 this study towards the creation of a conceptual 14556 model of SC resilience. 14557 </blockquote> 14558 </div> 14559 </div> 14560 </div> 14561 </div> 14562 <div class="slot-urls"></div> 14563 <a href="/wsc23papers/cea118.pdf" target="_blank">pdf</a 14564 ><br /> 14565 </div> 14566 <div class="slot-entry"> 14567 <a name="con280" tabindex="-1"></a> 14568 <div class="slot-title-line">
14569 <span class="slot-title" 14570 >Building and Operating Resilient Transportation Yards 14571 Using Simulation</span 14572 > 14573 </div> 14574 <div class="slot-authors"> 14575 Hafsa Binte Mohsin, Jae Yong Lee, and Vamshi Krishna 14576 Suvarna (Amazon) 14577 </div> 14578 <div class="slot-abstract"> 14579 <div> 14580 <a 14581 class="clickable no-decoration" 14582 id="vhsjs_view_314_1707793551_9870322" 14583 onclick="$('#vhsjs_view_314_1707793551_9870322').hide(); 14584 $('#vhsjs_hide_314_1707793551_9870322').show(); 14585 $('#313_1707793551_9870238').slideDown(function() { 14586 if (typeof Masonry === 'function') { 14587 $('.use_masonry').masonry(); 14588 }; 14589 14590 });" 14591 ><i class="fa fa-caret-right"></i> 14592 <span class="hover_link">Abstract</span></a 14593 ><a 14594 class="clickable no-decoration" 14595 id="vhsjs_hide_314_1707793551_9870322" 14596 onclick="$('#313_1707793551_9870238').hide(function() { 14597 if (typeof Masonry === 'function') { 14598 $('.use_masonry').masonry(); 14599 }; 14600 }); 14601 $('#vhsjs_hide_314_1707793551_9870322').hide(); 14602 $('#vhsjs_view_314_1707793551_9870322').show();" 14603 style="display: none" 14604 ><i class="fa fa-caret-down"></i> 14605 <span class="hover_link">Abstract</span></a 14606 > 14607 <div 14608 data-display-control="314_1707793551_9870322" 14609 id="313_1707793551_9870238" 14610 style="display: none" 14611 > 14612 <div class="arrow-slidedown"> 14613 <blockquote> 14614 Developing a comprehensive model is an effective 14615 approach for gaining insight into and analyzing 14616 complex systems such as transportation yards. 14617 Following this approach, a data-driven 14618 agent-based simulation model has been developed 14619 for transportation yards at Amazon which 14620 captures the features and processes of the 14621 system. By simulating different scenarios and 14622 using simulation output performance indicators 14623 like yard/parking slip/dock door utilization, 14624 entry/exit gate queue, and late departure count, 14625 this model helps to identify potential 14626 bottlenecks, inefficiencies, and risks in the 14627 system. This information is used for strategic 14628 decision making and/or improving the system. 14629 Furthermore, the user can find ways to increase 14630 the yards’ daily maximum volume process 14631 capacities through multiple 14632 ‘what-if’ scenarios. This model is 14633 performed with mean absolute error (MAE) and 14634 root mean square error (RMSE) of 6% and 7% 14635 respectively. This paper presents the overview, 14636 current use cases and future works for 14637 improvement of the simulation model. 14638 </blockquote> 14639 </div> 14640 </div> 14641 </div> 14642 </div> 14643 <div class="slot-urls"></div> 14644 <a href="/wsc23papers/142.pdf" target="_blank">pdf</a 14645 ><br /> 14646 </div> 14647 </div> 14648 <div class="session-entry">
14649 <span class="session-event-type">Technical Session</span 14650 ><span class="type-track-spacer"> · </span 14651 ><span class="program-track" 14652 >Logistics Supply Chains Transportation</span 14653 ><br /> 14654 <div class="session-title">Traffic Simulation</div> 14655 <div class="session-chair"> 14656 Chair: Dave Goldsman (Georgia Institute of Technology)<br /> 14657 </div> 14658 <div class="slot-entry"> 14659 <a name="con133" tabindex="-1"></a> 14660 <div class="slot-title-line"> 14661 <span class="slot-title" 14662 >Optimizing Arterial Traffic Signal Settings: Shotgun 14663 Version for Simultaneous Perturbation Stochastic 14664 Approximation Approach</span 14665 > 14666 </div> 14667 <div class="slot-authors"> 14668 Yen-Hsiang Chen and Michael Franciudi Hartono (National 14669 Taiwan University) 14670 </div> 14671 <div class="slot-abstract"> 14672 <div> 14673 <a 14674 class="clickable no-decoration" 14675 id="vhsjs_view_316_1707793551_992489" 14676 onclick="$('#vhsjs_view_316_1707793551_992489').hide(); 14677 $('#vhsjs_hide_316_1707793551_992489').show(); 14678 $('#315_1707793551_992481').slideDown(function() { 14679 if (typeof Masonry === 'function') { 14680 $('.use_masonry').masonry(); 14681 }; 14682 14683 });" 14684 ><i class="fa fa-caret-right"></i> 14685 <span class="hover_link">Abstract</span></a 14686 ><a 14687 class="clickable no-decoration" 14688 id="vhsjs_hide_316_1707793551_992489" 14689 onclick="$('#315_1707793551_992481').hide(function() { 14690 if (typeof Masonry === 'function') { 14691 $('.use_masonry').masonry(); 14692 }; 14693 }); 14694 $('#vhsjs_hide_316_1707793551_992489').hide(); 14695 $('#vhsjs_view_316_1707793551_992489').show();" 14696 style="display: none" 14697 ><i class="fa fa-caret-down"></i> 14698 <span class="hover_link">Abstract</span></a 14699 > 14700 <div 14701 data-display-control="316_1707793551_992489" 14702 id="315_1707793551_992481" 14703 style="display: none" 14704 > 14705 <div class="arrow-slidedown"> 14706 <blockquote> 14707 The recent advancement in hardware computation 14708 speed has allowed stochastic microscopic traffic 14709 simulators to be embedded in signal optimization 14710 systems. In this study, stochastic perturbation 14711 simulation approximations (SPSA), an efficient 14712 difference-typed gradient-based searching, has 14713 been applied in the signal solver of a signal 14714 optimization system due to (i) its lower 14715 required total number of replications and (ii) 14716 the capability to conduct variance reduction 14717 technique (VRT). The case study has shown that 14718 the objective value, in terms of road 14719 users’ delay, indeed improves over 14720 iterations. Since the gradient-based method may 14721 be trapped in the local optimal, this study has 14722 further applied the shotgun mechanism that 14723 allows better solutions in the subject stage to 14724 proceed to the next stage. By further offering 14725 the shotgun process, the quality of the solution 14726 can be further improved. 14727 </blockquote> 14728 </div> 14729 </div> 14730 </div> 14731 </div> 14732 <div class="slot-urls"></div> 14733 <a href="/wsc23papers/144.pdf" target="_blank">pdf</a 14734 ><br /> 14735 </div> 14736 <div class="slot-entry"> 14737 <a name="con329" tabindex="-1"></a> 14738 <div class="slot-title-line">
14739 <span class="slot-title" 14740 >Breaking Through the Traffic Congestion: Asynchronous 14741 Time Series Data Integration and XGBOOST for Accurate 14742 Traffic Density Prediction</span 14743 > 14744 </div> 14745 <div class="slot-authors"> 14746 Eloi Garcia, Carles Serrat, and Fatos Xhafa (Universitat 14747 Politècnica de Catalunya-BarcelonaTECH) 14748 </div> 14749 <div class="slot-abstract"> 14750 <div> 14751 <a 14752 class="clickable no-decoration" 14753 id="vhsjs_view_318_1707793551_9947648" 14754 onclick="$('#vhsjs_view_318_1707793551_9947648').hide(); 14755 $('#vhsjs_hide_318_1707793551_9947648').show(); 14756 $('#317_1707793551_994757').slideDown(function() { 14757 if (typeof Masonry === 'function') { 14758 $('.use_masonry').masonry(); 14759 }; 14760 14761 });" 14762 ><i class="fa fa-caret-right"></i> 14763 <span class="hover_link">Abstract</span></a 14764 ><a 14765 class="clickable no-decoration" 14766 id="vhsjs_hide_318_1707793551_9947648" 14767 onclick="$('#317_1707793551_994757').hide(function() { 14768 if (typeof Masonry === 'function') { 14769 $('.use_masonry').masonry(); 14770 }; 14771 }); 14772 $('#vhsjs_hide_318_1707793551_9947648').hide(); 14773 $('#vhsjs_view_318_1707793551_9947648').show();" 14774 style="display: none" 14775 ><i class="fa fa-caret-down"></i> 14776 <span class="hover_link">Abstract</span></a 14777 > 14778 <div 14779 data-display-control="318_1707793551_9947648" 14780 id="317_1707793551_994757" 14781 style="display: none" 14782 > 14783 <div class="arrow-slidedown"> 14784 <blockquote> 14785 The proliferation of data collection from smart 14786 cities has resulted in an exponential growth in 14787 the volume of measurements available for 14788 analysis. However, collecting all parameters 14789 concurrently at the same location is not 14790 feasible due to the complex nature of the real 14791 world. We present an innovative methodology that 14792 enriches asynchronous time series data from a 14793 variety of sources to facilitate data enrichment 14794 and city-wide behavior simulation. A case study 14795 on OpenDataBCN attests to the efficacy of this 14796 approach via an XGBoost model, predicated on 14797 geographical coordinates and timestamp 14798 disparities. The consolidation of data from 14799 different sources improves the richness and 14800 granularity of information at disposal for 14801 analysis, thereby revealing previously hidden 14802 patterns and relationships, exhibiting new 14803 insights and underscoring the potential of this 14804 methodology for sustainable and efficient data 14805 enrichment processes as well as new 14806 possibilities for simulation based on smart city 14807 datasets. 14808 </blockquote> 14809 </div> 14810 </div> 14811 </div> 14812 </div> 14813 <div class="slot-urls"></div> 14814 <a href="/wsc23papers/145.pdf" target="_blank">pdf</a 14815 ><br /> 14816 </div> 14817 </div> 14818 <div class="session-entry">
14819 <span class="session-event-type">Technical Session</span 14820 ><span class="type-track-spacer"> · </span 14821 ><span class="program-track" 14822 >Logistics Supply Chains Transportation</span 14823 ><br /> 14824 <div class="session-title">Simheuristic Approaches</div> 14825 <div class="session-chair"> 14826 Chair: Michael Kuhl (Rochester Institute of Technology)<br /> 14827 </div> 14828 <div class="slot-entry"> 14829 <a name="inv120" tabindex="-1"></a> 14830 <div class="slot-title-line"> 14831 <span class="slot-title" 14832 >A Dynamic Forecast Demand Scenario Analysis to Design 14833 an Automated Parcel Lockers Network in Pamplona 14834 (Spain) Using a Simulation-Optimization Model</span 14835 > 14836 </div> 14837 <div class="slot-authors"> 14838 Irene Izco (Public University of Navarre); Adrian 14839 Serrano-Hernandez and Javier Faulin (Public University 14840 of Navarre, Institute of Smart Cities); and Bartosz 14841 Sawik (AGH University of Science and Technology) 14842 </div> 14843 <div class="slot-abstract"> 14844 <div> 14845 <a 14846 class="clickable no-decoration" 14847 id="vhsjs_view_320_1707793551_9996877" 14848 onclick="$('#vhsjs_view_320_1707793551_9996877').hide(); 14849 $('#vhsjs_hide_320_1707793551_9996877').show(); 14850 $('#319_1707793551_9996793').slideDown(function() { 14851 if (typeof Masonry === 'function') { 14852 $('.use_masonry').masonry(); 14853 }; 14854 14855 });" 14856 ><i class="fa fa-caret-right"></i> 14857 <span class="hover_link">Abstract</span></a 14858 ><a 14859 class="clickable no-decoration" 14860 id="vhsjs_hide_320_1707793551_9996877" 14861 onclick="$('#319_1707793551_9996793').hide(function() { 14862 if (typeof Masonry === 'function') { 14863 $('.use_masonry').masonry(); 14864 }; 14865 }); 14866 $('#vhsjs_hide_320_1707793551_9996877').hide(); 14867 $('#vhsjs_view_320_1707793551_9996877').show();" 14868 style="display: none" 14869 ><i class="fa fa-caret-down"></i> 14870 <span class="hover_link">Abstract</span></a 14871 > 14872 <div 14873 data-display-control="320_1707793551_9996877" 14874 id="319_1707793551_9996793" 14875 style="display: none" 14876 > 14877 <div class="arrow-slidedown"> 14878 <blockquote> 14879 The disruptions experienced by the last mile 14880 delivery processes during the SARS-CoV-2 14881 pandemic have inevitably raised the dilemma of 14882 alternative last mile approaches in Urban 14883 Logistics (UL). Self-Collection Delivery Systems 14884 (SCDS) suppose an improvement for both courier 14885 companies and customers, providing flexibility 14886 of time-windows and reducing overall mileage, 14887 delivery time and, gas emissions. Drawing a 14888 distinction from previous works involving hybrid 14889 modeling for automated parcel lockers (APL) 14890 network design, this study integrates a System 14891 Dynamics Simulation Model (SDSM) to forecast 14892 e-commerce demand in Pamplona (Spain), and 14893 considers the scalability of the model for other 14894 cities. A bi-criteria Facility Location Problem 14895 (FLP) is proposed and solved with an 14896 ε-constraint method, where ε is 14897 defined as the level of coverage of the total 14898 demand, and four different cases of demand 14899 coverage are run. The simulation and demand 14900 forecast was carried out using Anylogic 14901 software, being CPLEX the optimization solver. 14902 </blockquote> 14903 </div> 14904 </div> 14905 </div> 14906 </div> 14907 <div class="slot-urls"></div> 14908 <a href="/wsc23papers/146.pdf" target="_blank">pdf</a 14909 ><br /> 14910 </div> 14911 <div class="slot-entry"> 14912 <a name="inv116" tabindex="-1"></a> 14913 <div class="slot-title-line">
14914 <span class="slot-title" 14915 >A Demand Modeling Pipeline for an Agent-Based Traffic 14916 Simulation of the City of Barcelona</span 14917 > 14918 </div> 14919 <div class="slot-authors"> 14920 Jonas Fuentes Leon (Universitat Oberta de Catalunya, 14921 Spindox Spain); Francesca Giancola (Spindox S.p.A.; 14922 DIAG, Sapienza University of Rome); and Andrea 14923 Boccolucci and Mattia Neroni (Spindox S.p.A.) 14924 </div> 14925 <div class="slot-abstract"> 14926 <div> 14927 <a 14928 class="clickable no-decoration" 14929 id="vhsjs_view_322_1707793552_0019572" 14930 onclick="$('#vhsjs_view_322_1707793552_0019572').hide(); 14931 $('#vhsjs_hide_322_1707793552_0019572').show(); 14932 $('#321_1707793552_0019488').slideDown(function() { 14933 if (typeof Masonry === 'function') { 14934 $('.use_masonry').masonry(); 14935 }; 14936 14937 });" 14938 ><i class="fa fa-caret-right"></i> 14939 <span class="hover_link">Abstract</span></a 14940 ><a 14941 class="clickable no-decoration" 14942 id="vhsjs_hide_322_1707793552_0019572" 14943 onclick="$('#321_1707793552_0019488').hide(function() { 14944 if (typeof Masonry === 'function') { 14945 $('.use_masonry').masonry(); 14946 }; 14947 }); 14948 $('#vhsjs_hide_322_1707793552_0019572').hide(); 14949 $('#vhsjs_view_322_1707793552_0019572').show();" 14950 style="display: none" 14951 ><i class="fa fa-caret-down"></i> 14952 <span class="hover_link">Abstract</span></a 14953 > 14954 <div 14955 data-display-control="322_1707793552_0019572" 14956 id="321_1707793552_0019488" 14957 style="display: none" 14958 > 14959 <div class="arrow-slidedown"> 14960 <blockquote> 14961 The growth of urban population and the 14962 proliferation of mobility options in big cities 14963 are adding to the complexity of comprehending 14964 how people move about and how efficiently they 14965 do it. Understanding how traffic patterns change 14966 throughout the day is essential for legislators, 14967 public administrations, and other stakeholders, 14968 as it has a direct impact on citizens' quality 14969 of life by, for instance, increasing greenhouse 14970 gas emissions and noise pollution. In this 14971 context, simulation becomes an essential tool 14972 for grasping the emerging dynamics of urban 14973 transportation, citizens' mobility patterns, and 14974 traffic flow bottlenecks. This work presents a 14975 complete data modelling pipeline for generating 14976 the population, network and transportation 14977 demand that is fed to a multi-modal traffic 14978 simulation of the city of Barcelona using MATSim 14979 and open-access statistical data sources. The 14980 model is calibrated, the results are obtained, 14981 and future applications of the developed tool 14982 are outlined. 14983 </blockquote> 14984 </div> 14985 </div> 14986 </div> 14987 </div> 14988 <div class="slot-urls"></div> 14989 <a href="/wsc23papers/147.pdf" target="_blank">pdf</a 14990 ><br /> 14991 </div> 14992 </div> 14993 <div class="session-entry">
14994 <span class="session-event-type">Technical Session</span 14995 ><span class="type-track-spacer"> · </span 14996 ><span class="program-track" 14997 >Logistics Supply Chains Transportation</span 14998 ><br /> 14999 <div class="session-title">Yard Management</div> 15000 <div class="session-chair"> 15001 Chair: Klaus Altendorfer (Upper Austrian University of 15002 Applied Science)<br /> 15003 </div> 15004 <div class="slot-entry"> 15005 <a name="con106" tabindex="-1"></a> 15006 <div class="slot-title-line"> 15007 <span class="slot-title" 15008 >Cloud-Based Hybrid Simulation Model For Optimizing 15009 Warehouse Yard Operations</span 15010 > 15011 </div> 15012 <div class="slot-authors"> 15013 Mohammed Farhan, Pascalin Ngoko, Farouq Halawa, and 15014 Raashid Mohammed (Amazon) 15015 </div> 15016 <div class="slot-abstract"> 15017 <div> 15018 <a 15019 class="clickable no-decoration" 15020 id="vhsjs_view_324_1707793552_0089338" 15021 onclick="$('#vhsjs_view_324_1707793552_0089338').hide(); 15022 $('#vhsjs_hide_324_1707793552_0089338').show(); 15023 $('#323_1707793552_0089257').slideDown(function() { 15024 if (typeof Masonry === 'function') { 15025 $('.use_masonry').masonry(); 15026 }; 15027 15028 });" 15029 ><i class="fa fa-caret-right"></i> 15030 <span class="hover_link">Abstract</span></a 15031 ><a 15032 class="clickable no-decoration" 15033 id="vhsjs_hide_324_1707793552_0089338" 15034 onclick="$('#323_1707793552_0089257').hide(function() { 15035 if (typeof Masonry === 'function') { 15036 $('.use_masonry').masonry(); 15037 }; 15038 }); 15039 $('#vhsjs_hide_324_1707793552_0089338').hide(); 15040 $('#vhsjs_view_324_1707793552_0089338').show();" 15041 style="display: none" 15042 ><i class="fa fa-caret-down"></i> 15043 <span class="hover_link">Abstract</span></a 15044 > 15045 <div 15046 data-display-control="324_1707793552_0089338" 15047 id="323_1707793552_0089257" 15048 style="display: none" 15049 > 15050 <div class="arrow-slidedown"> 15051 <blockquote> 15052 Fulfillment centers in the E-commerce industry 15053 are highly complex systems that houses inventory 15054 and fulfill customer orders. One of the key 15055 processes at these centers involves translating 15056 customer demands into trucks and yard 15057 operations. Truck yards with operational issues 15058 can create delays in customer orders. In this 15059 paper, we show how a scalable cloud-based hybrid 15060 simulation model is used to improve yard 15061 operations, optimize flow and design, and 15062 forecast yard congestion. Cloud experimentation 15063 along with automated database connectivity 15064 allows any user to run simulation analyses to 15065 derive data driven operational decisions. We 15066 tested the model on two real world case studies, 15067 which results in cost savings for the 15068 organization. This paper also proposes a robust 15069 automated framework for setting simulation 15070 validation benchmarks and measuring model 15071 accuracy. 15072 </blockquote> 15073 </div> 15074 </div> 15075 </div> 15076 </div> 15077 <div class="slot-urls"></div> 15078 <a href="/wsc23papers/148.pdf" target="_blank">pdf</a 15079 ><br /> 15080 </div> 15081 <div class="slot-entry"> 15082 <a name="con157" tabindex="-1"></a> 15083 <div class="slot-title-line">
15084 <span class="slot-title" 15085 >Simulation-Based Analysis of Improvements in Vehicle 15086 Routing with Time Windows Using a One-sided VCG 15087 Mechanism for the Reallocation of Unfavorable Time 15088 Windows</span 15089 > 15090 </div> 15091 <div class="slot-authors"> 15092 Felix Roeper and Ralf Elbert (Technische 15093 Universität Darmstadt) 15094 </div> 15095 <div class="slot-abstract"> 15096 <div> 15097 <a 15098 class="clickable no-decoration" 15099 id="vhsjs_view_326_1707793552_0110252" 15100 onclick="$('#vhsjs_view_326_1707793552_0110252').hide(); 15101 $('#vhsjs_hide_326_1707793552_0110252').show(); 15102 $('#325_1707793552_0110166').slideDown(function() { 15103 if (typeof Masonry === 'function') { 15104 $('.use_masonry').masonry(); 15105 }; 15106 15107 });" 15108 ><i class="fa fa-caret-right"></i> 15109 <span class="hover_link">Abstract</span></a 15110 ><a 15111 class="clickable no-decoration" 15112 id="vhsjs_hide_326_1707793552_0110252" 15113 onclick="$('#325_1707793552_0110166').hide(function() { 15114 if (typeof Masonry === 'function') { 15115 $('.use_masonry').masonry(); 15116 }; 15117 }); 15118 $('#vhsjs_hide_326_1707793552_0110252').hide(); 15119 $('#vhsjs_view_326_1707793552_0110252').show();" 15120 style="display: none" 15121 ><i class="fa fa-caret-down"></i> 15122 <span class="hover_link">Abstract</span></a 15123 > 15124 <div 15125 data-display-control="326_1707793552_0110252" 15126 id="325_1707793552_0110166" 15127 style="display: none" 15128 > 15129 <div class="arrow-slidedown"> 15130 <blockquote> 15131 In road freight transport, booking unfavorable 15132 time windows (TW) through time window management 15133 systems (TWMS) for loading or unloading trucks 15134 at the loading dock often leads to avoidable 15135 long tours. Therefore, this paper investigates, 15136 based on an agent-based simulation framework, 15137 the efficiency gains and improvements in vehicle 15138 routing with TW constraints that can be achieved 15139 by a reallocation of unfavorable TWs using a 15140 one-sided Vickrey-Clarke-Groves mechanism. A 15141 branch-and-cut algorithm is used to evaluate the 15142 value of a TW in the context of a pickup and 15143 delivery problem with time windows and to 15144 generate a bid for the auction. A winner 15145 determination problem is solved for conducting 15146 the auction. We show that a reallocation of 15147 unfavorable TWs leads to distance savings for 15148 the considered tours of the auction winners of 15149 13% on average. Further, we can show that the 15150 TWMS provider can benefit by operating the 15151 mechanism on an electronic marketplace. 15152 </blockquote> 15153 </div> 15154 </div> 15155 </div> 15156 </div> 15157 <div class="slot-urls"></div> 15158 <a href="/wsc23papers/149.pdf" target="_blank">pdf</a 15159 ><br /> 15160 </div> 15161 <div class="slot-entry"> 15162 <a name="con278" tabindex="-1"></a> 15163 <div class="slot-title-line">
15164 <span class="slot-title" 15165 >Crossstacks: A Dataset and a Simulative Study of 15166 Storage Allocation Strategies for Cross-Docking 15167 Block-Stacking Warehouses</span 15168 > 15169 </div> 15170 <div class="slot-authors"> 15171 Alexandru Rinciog (TU Dortmund University), Natalia 15172 Ogorelysheva (Fraunhofer IML), Jakob Pfrommer (TU 15173 Dortmund University), Anna Vasileva (Fraunhofer IML), 15174 and Hardik Rathod and Anne Meyer (TU Dortmund 15175 University) 15176 </div> 15177 <div class="slot-abstract"> 15178 <div> 15179 <a 15180 class="clickable no-decoration" 15181 id="vhsjs_view_328_1707793552_0135155" 15182 onclick="$('#vhsjs_view_328_1707793552_0135155').hide(); 15183 $('#vhsjs_hide_328_1707793552_0135155').show(); 15184 $('#327_1707793552_0135074').slideDown(function() { 15185 if (typeof Masonry === 'function') { 15186 $('.use_masonry').masonry(); 15187 }; 15188 15189 });" 15190 ><i class="fa fa-caret-right"></i> 15191 <span class="hover_link">Abstract</span></a 15192 ><a 15193 class="clickable no-decoration" 15194 id="vhsjs_hide_328_1707793552_0135155" 15195 onclick="$('#327_1707793552_0135074').hide(function() { 15196 if (typeof Masonry === 'function') { 15197 $('.use_masonry').masonry(); 15198 }; 15199 }); 15200 $('#vhsjs_hide_328_1707793552_0135155').hide(); 15201 $('#vhsjs_view_328_1707793552_0135155').show();" 15202 style="display: none" 15203 ><i class="fa fa-caret-down"></i> 15204 <span class="hover_link">Abstract</span></a 15205 > 15206 <div 15207 data-display-control="328_1707793552_0135155" 15208 id="327_1707793552_0135074" 15209 style="display: none" 15210 > 15211 <div class="arrow-slidedown"> 15212 <blockquote> 15213 Cross-docking is a warehousing strategy that 15214 (ideally) moves goods from inbound docks 15215 directly to outbound docks. In reality, goods 15216 often need to be temporarily stored. 15217 Cross-docking is typically set up as a 15218 block-stacking warehouse (BSW), where goods are 15219 stored directly on the ground. Autonomous mobile 15220 robots (AMRs) could significantly reduce BSW 15221 costs. To deploy AMR systems to BSWs, five 15222 interlaced decision problems, including the 15223 storage location assignment problem (SLAP), need 15224 to be solved. Because of the combinatorial 15225 complexity of BSWs, and the absence of pertinent 15226 use case data and fitting simulation software, 15227 this is a challenging task. This work seeks to 15228 alleviate these gaps by (1) extending SLAPStack, 15229 a fine-grained open-source BSW simulation 15230 framework to accommodate cross-docking, (2) 15231 providing CROSSStacks, a real-world 15232 cross-docking dataset, and (3) evaluating two 15233 dual command cycle SLAP strategies as of yet 15234 untested for BSWs. One of the approaches 15235 outperforms a naive cross-docking SLAP strategy. 15236 </blockquote> 15237 </div> 15238 </div> 15239 </div> 15240 </div> 15241 <div class="slot-urls"></div> 15242 <a href="/wsc23papers/150.pdf" target="_blank">pdf</a 15243 ><br /> 15244 </div> 15245 </div> 15246 <div class="session-entry">
15247 <span class="session-event-type">Technical Session</span 15248 ><span class="type-track-spacer"> · </span 15249 ><span class="program-track" 15250 >Logistics Supply Chains Transportation</span 15251 ><br /> 15252 <div class="session-title"> 15253 Simulation with Reinforcement Learning 15254 </div> 15255 <div class="session-chair"> 15256 Chair: Steffen Strassburger (Technische Universität 15257 Ilmenau)<br /> 15258 </div> 15259 <div class="slot-entry"> 15260 <a name="con292" tabindex="-1"></a> 15261 <div class="slot-title-line"> 15262 <span class="slot-title" 15263 >Multi-Agent Proximal Policy Optimization for a 15264 Deadlock Capable Transport System in a 15265 Simulation-Based Learning Environment</span 15266 > 15267 </div> 15268 <div class="slot-authors"> 15269 Marcel Müller (Otto von Guericke University 15270 Magdeburg); Lorena Silvana Reyes Rubiano (RWTH Aachen 15271 University, Universidad de La Sabana); and Tobias 15272 Reggelin and Hartmut Zadek (Otto von Guericke University 15273 Magdeburg) 15274 </div> 15275 <div class="slot-abstract"> 15276 <div> 15277 <a 15278 class="clickable no-decoration" 15279 id="vhsjs_view_330_1707793552_019319" 15280 onclick="$('#vhsjs_view_330_1707793552_019319').hide(); 15281 $('#vhsjs_hide_330_1707793552_019319').show(); 15282 $('#329_1707793552_019311').slideDown(function() { 15283 if (typeof Masonry === 'function') { 15284 $('.use_masonry').masonry(); 15285 }; 15286 15287 });" 15288 ><i class="fa fa-caret-right"></i> 15289 <span class="hover_link">Abstract</span></a 15290 ><a 15291 class="clickable no-decoration" 15292 id="vhsjs_hide_330_1707793552_019319" 15293 onclick="$('#329_1707793552_019311').hide(function() { 15294 if (typeof Masonry === 'function') { 15295 $('.use_masonry').masonry(); 15296 }; 15297 }); 15298 $('#vhsjs_hide_330_1707793552_019319').hide(); 15299 $('#vhsjs_view_330_1707793552_019319').show();" 15300 style="display: none" 15301 ><i class="fa fa-caret-down"></i> 15302 <span class="hover_link">Abstract</span></a 15303 > 15304 <div 15305 data-display-control="330_1707793552_019319" 15306 id="329_1707793552_019311" 15307 style="display: none" 15308 > 15309 <div class="arrow-slidedown"> 15310 <blockquote> 15311 In this paper, we explore the potential of 15312 multi-agent reinforcement learning (MARL) for 15313 managing the driving behavior of autonomous 15314 guided vehicles (AGVs) in production logistics 15315 environments with single-lane tracks, where 15316 deadlocks pose a significant challenge. We build 15317 upon previous work and adopt a MARL approach 15318 using the Proximal Policy Optimization (PPO) 15319 algorithm. We conduct a thorough hyperparameter 15320 search and investigate the impact of varying 15321 numbers of agents on the performance of the 15322 AGVs. Our results demonstrate the effectiveness 15323 of the MARL approach in addressing deadlocks and 15324 coordinating AGV behavior, as well as the 15325 scalability of the learned policy to different 15326 numbers of agents. The Bayesian optimization 15327 process and increased iteration count contribute 15328 to improved performance and more stable learning 15329 curves. 15330 </blockquote> 15331 </div> 15332 </div> 15333 </div> 15334 </div> 15335 <div class="slot-urls"></div> 15336 <a href="/wsc23papers/151.pdf" target="_blank">pdf</a 15337 ><br /> 15338 </div> 15339 <div class="slot-entry"> 15340 <a name="con348" tabindex="-1"></a> 15341 <div class="slot-title-line">
15342 <span class="slot-title" 15343 >Simulation Analysis of a Reinforcement-Learning-Based 15344 Warehouse Dispatching Method Considering Due Date and 15345 Travel Distance</span 15346 > 15347 </div> 15348 <div class="slot-authors"> 15349 Sriparvathi Shaji Bhattathiri, Ankita Tondwalkar, 15350 Michael E. Kuhl, and Andres Kwasinski (Rochester 15351 Institute of Technology) 15352 </div> 15353 <div class="slot-abstract"> 15354 <div> 15355 <a 15356 class="clickable no-decoration" 15357 id="vhsjs_view_332_1707793552_0217657" 15358 onclick="$('#vhsjs_view_332_1707793552_0217657').hide(); 15359 $('#vhsjs_hide_332_1707793552_0217657').show(); 15360 $('#331_1707793552_0217576').slideDown(function() { 15361 if (typeof Masonry === 'function') { 15362 $('.use_masonry').masonry(); 15363 }; 15364 15365 });" 15366 ><i class="fa fa-caret-right"></i> 15367 <span class="hover_link">Abstract</span></a 15368 ><a 15369 class="clickable no-decoration" 15370 id="vhsjs_hide_332_1707793552_0217657" 15371 onclick="$('#331_1707793552_0217576').hide(function() { 15372 if (typeof Masonry === 'function') { 15373 $('.use_masonry').masonry(); 15374 }; 15375 }); 15376 $('#vhsjs_hide_332_1707793552_0217657').hide(); 15377 $('#vhsjs_view_332_1707793552_0217657').show();" 15378 style="display: none" 15379 ><i class="fa fa-caret-down"></i> 15380 <span class="hover_link">Abstract</span></a 15381 > 15382 <div 15383 data-display-control="332_1707793552_0217657" 15384 id="331_1707793552_0217576" 15385 style="display: none" 15386 > 15387 <div class="arrow-slidedown"> 15388 <blockquote> 15389 As the adoption of autonomous mobile robots in 15390 warehouses and other industrial environments 15391 continues to increase, there is a need for 15392 methods that can effectively dispatch robots to 15393 meet system demand. Real-time dispatching of 15394 autonomous mobile robots can be very complex, 15395 but simple rule-based methods are typically used 15396 for this task. In this paper, a 15397 reinforcement-learning-based dispatching method 15398 for intralogistics (RLDI) is proposed. RLDI is 15399 warehouse layout independent and takes into 15400 consideration task due dates and the travel 15401 distance. The algorithm is trained and tested in 15402 a simulation environment that represents a small 15403 warehouse. Monte Carlo simulation analysis is 15404 used to explore the capabilities and limitations 15405 of the established RLDI. The performance of the 15406 method is compared to the shortest distance 15407 dispatching rule in single and multi-agent 15408 environments under various levels of due date 15409 tightness. Experimental results demonstrate the
15410 potential for using reinforcement learning 15411 methods for warehouse dispatching. 15412 </blockquote> 15413 </div> 15414 </div> 15415 </div> 15416 </div> 15417 <div class="slot-urls"></div> 15418 <a href="/wsc23papers/152.pdf" target="_blank">pdf</a 15419 ><br /> 15420 </div> 15421 <div class="slot-entry"> 15422 <a name="con140" tabindex="-1"></a> 15423 <div class="slot-title-line"> 15424 <span class="slot-title" 15425 >Purpose in the Machine: Do Traffic Simulators Produce 15426 Distributionally Equivalent Outcomes for Reinforcement 15427 Learning Applications?</span 15428 > 15429 </div> 15430 <div class="slot-authors"> 15431 Rex Chen, Kathleen M. Carley, Fei Fang, and Norman Sadeh 15432 (Carnegie Mellon University) 15433 </div> 15434 <div class="slot-abstract"> 15435 <div> 15436 <a 15437 class="clickable no-decoration" 15438 id="vhsjs_view_334_1707793552_024086" 15439 onclick="$('#vhsjs_view_334_1707793552_024086').hide(); 15440 $('#vhsjs_hide_334_1707793552_024086').show(); 15441 $('#333_1707793552_0240777').slideDown(function() { 15442 if (typeof Masonry === 'function') { 15443 $('.use_masonry').masonry(); 15444 }; 15445 15446 });" 15447 ><i class="fa fa-caret-right"></i> 15448 <span class="hover_link">Abstract</span></a 15449 ><a 15450 class="clickable no-decoration" 15451 id="vhsjs_hide_334_1707793552_024086" 15452 onclick="$('#333_1707793552_0240777').hide(function() { 15453 if (typeof Masonry === 'function') { 15454 $('.use_masonry').masonry(); 15455 }; 15456 }); 15457 $('#vhsjs_hide_334_1707793552_024086').hide(); 15458 $('#vhsjs_view_334_1707793552_024086').show();" 15459 style="display: none" 15460 ><i class="fa fa-caret-down"></i> 15461 <span class="hover_link">Abstract</span></a 15462 > 15463 <div 15464 data-display-control="334_1707793552_024086" 15465 id="333_1707793552_0240777" 15466 style="display: none" 15467 > 15468 <div class="arrow-slidedown"> 15469 <blockquote> 15470 Traffic simulators are used to generate data for 15471 learning in intelligent transportation systems 15472 (ITSs). A key question is to what extent their 15473 modelling assumptions affect the capabilities of 15474 ITSs to adapt to various scenarios when deployed 15475 in the real world. This work focuses on two 15476 simulators commonly used to train reinforcement 15477 learning (RL) agents for traffic applications, 15478 CityFlow and SUMO. A controlled virtual 15479 experiment varying driver behavior and 15480 simulation scale finds evidence against 15481 distributional equivalence in RL-relevant 15482 measures from these simulators, with the root 15483 mean squared error and KL divergence being 15484 significantly greater than 0 for all assessed 15485 measures. While granular real-world validation 15486 generally remains infeasible, these findings 15487 suggest that traffic simulators are not a deus 15488 ex machina for RL training: understanding the 15489 impacts of inter-simulator differences is 15490 necessary to train and deploy RL-based ITSs. 15491 </blockquote> 15492 </div> 15493 </div> 15494 </div> 15495 </div> 15496 <div class="slot-urls"></div> 15497 <a href="/wsc23papers/153.pdf" target="_blank">pdf</a 15498 ><br /> 15499 </div> 15500 </div> 15501 <div class="session-entry">
15502 <span class="session-event-type">Technical Session</span 15503 ><span class="type-track-spacer"> · </span 15504 ><span class="program-track" 15505 >Logistics Supply Chains Transportation</span 15506 ><br /> 15507 <div class="session-title"> 15508 Simulation-Optimization with Uncertainty 15509 </div> 15510 <div class="session-chair"> 15511 Chair: Javier Faulin (Public University of Navarre, 15512 Institute of Smart Cities)<br /> 15513 </div> 15514 <div class="slot-entry"> 15515 <a name="con241" tabindex="-1"></a> 15516 <div class="slot-title-line"> 15517 <span class="slot-title" 15518 >Solving the Multi-Allocation p-Hub Median Problem 15519 with Stochastic Travel Times: A Simheuristic 15520 Approach</span 15521 > 15522 </div> 15523 <div class="slot-authors"> 15524 Niklas Jost (TU Dortmund), Majsa Ammouriova (Universitat 15525 Oberta de Catalunya), Aleksandra Grochala (TU Dortmund), 15526 Angel Juan (Universitat Polit`ecnica de Val`encia), and 15527 Christin Schumacher (TU Dortmund) 15528 </div> 15529 <div class="slot-abstract"> 15530 <div> 15531 <a 15532 class="clickable no-decoration" 15533 id="vhsjs_view_336_1707793552_0284088" 15534 onclick="$('#vhsjs_view_336_1707793552_0284088').hide(); 15535 $('#vhsjs_hide_336_1707793552_0284088').show(); 15536 $('#335_1707793552_0284004').slideDown(function() { 15537 if (typeof Masonry === 'function') { 15538 $('.use_masonry').masonry(); 15539 }; 15540 15541 });" 15542 ><i class="fa fa-caret-right"></i> 15543 <span class="hover_link">Abstract</span></a 15544 ><a 15545 class="clickable no-decoration" 15546 id="vhsjs_hide_336_1707793552_0284088" 15547 onclick="$('#335_1707793552_0284004').hide(function() { 15548 if (typeof Masonry === 'function') { 15549 $('.use_masonry').masonry(); 15550 }; 15551 }); 15552 $('#vhsjs_hide_336_1707793552_0284088').hide(); 15553 $('#vhsjs_view_336_1707793552_0284088').show();" 15554 style="display: none" 15555 ><i class="fa fa-caret-down"></i> 15556 <span class="hover_link">Abstract</span></a 15557 > 15558 <div 15559 data-display-control="336_1707793552_0284088" 15560 id="335_1707793552_0284004" 15561 style="display: none" 15562 > 15563 <div class="arrow-slidedown"> 15564 <blockquote> 15565 The p-hub median problems (pHMPs) are a 15566 well-researched topic within the fields of 15567 Operations Research and Industrial Engineering. 15568 These problems have been found to have a wide 15569 range of practical applications in various areas 15570 such as logistics, retailing, and Internet 15571 computing. These applications have made pHMPs an 15572 important area of study, leading to numerous 15573 research efforts aimed at solving different 15574 variations of the problem. This paper presents a 15575 simheuristic algorithm for solving the
15576 uncapacitated version of the pHMP with 15577 stochastic travel times. The proposed approach 15578 combines simulation with biased-randomized 15579 heuristics to generate high-quality solutions 15580 quickly. The proposed method is validated by 15581 testing it on huge benchmark instances, which 15582 include stochastic travel times. The results 15583 demonstrate the efficiency of the proposed 15584 approach for this particular problem variation. 15585 The simulation-optimization approach provides a 15586 promising solution to a practical problem that 15587 arises in many real-world applications. 15588 </blockquote> 15589 </div> 15590 </div> 15591 </div> 15592 </div> 15593 <div class="slot-urls"></div> 15594 <a href="/wsc23papers/154.pdf" target="_blank">pdf</a 15595 ><br /> 15596 </div> 15597 <div class="slot-entry"> 15598 <a name="con170" tabindex="-1"></a> 15599 <div class="slot-title-line"> 15600 <span class="slot-title" 15601 >Simulation-based Analysis of Onshore Wind Farm 15602 Installation Strategies</span 15603 > 15604 </div> 15605 <div class="slot-authors"> 15606 Daniel Rippel, Sebastian Eberlein, Stephan Oelker, and 15607 Michael Lütjen (BIBA - Bremer Institut für 15608 Produktion und Logistik GmbH at the University of 15609 Bremen) and Michael Freitag (BIBA - Bremer Institut 15610 für Produktion und Logistik GmbH at the University 15611 of Bremen, University of Bremen) 15612 </div> 15613 <div class="slot-abstract"> 15614 <div> 15615 <a 15616 class="clickable no-decoration" 15617 id="vhsjs_view_338_1707793552_0309825" 15618 onclick="$('#vhsjs_view_338_1707793552_0309825').hide(); 15619 $('#vhsjs_hide_338_1707793552_0309825').show(); 15620 $('#337_1707793552_0309741').slideDown(function() { 15621 if (typeof Masonry === 'function') { 15622 $('.use_masonry').masonry(); 15623 }; 15624 15625 });" 15626 ><i class="fa fa-caret-right"></i> 15627 <span class="hover_link">Abstract</span></a 15628 ><a 15629 class="clickable no-decoration" 15630 id="vhsjs_hide_338_1707793552_0309825" 15631 onclick="$('#337_1707793552_0309741').hide(function() { 15632 if (typeof Masonry === 'function') { 15633 $('.use_masonry').masonry(); 15634 }; 15635 }); 15636 $('#vhsjs_hide_338_1707793552_0309825').hide(); 15637 $('#vhsjs_view_338_1707793552_0309825').show();" 15638 style="display: none" 15639 ><i class="fa fa-caret-down"></i> 15640 <span class="hover_link">Abstract</span></a 15641 > 15642 <div 15643 data-display-control="338_1707793552_0309825" 15644 id="337_1707793552_0309741" 15645 style="display: none" 15646 > 15647 <div class="arrow-slidedown"> 15648 <blockquote> 15649 Wind energy constitutes a main contributor to 15650 clean and renewable energy. While the offshore 15651 sector received much attention from research and 15652 industry, onshore wind farms still make up the 15653 largest share of installation projects. Thereby, 15654 onshore installations retain similar wind speed 15655 restrictions as their offshore counterparts but 15656 additionally introduce limits and wait time 15657 restrictions between installation operations. 15658 This article proposes extending a planning 15659 method initially designed for offshore wind 15660 farms to cover these additional requirements and 15661 proposes a simulation model capable of 15662 evaluating the resulting plans. The results show 15663 that the extended approach prevents violations 15664 of these requirements, mitigates the influence 15665 of weather forecast uncertainties, and provides 15666 efficient plans for installation operations. 15667 </blockquote> 15668 </div> 15669 </div> 15670 </div> 15671 </div> 15672 <div class="slot-urls"></div> 15673 <a href="/wsc23papers/155.pdf" target="_blank">pdf</a 15674 ><br /> 15675 </div> 15676 <div class="slot-entry"> 15677 <a name="con283" tabindex="-1"></a> 15678 <div class="slot-title-line">
15679 <span class="slot-title" 15680 >A Two-Stage Stochastic Model for Drone Delivery 15681 System with Uncertainty in Customer Demands</span 15682 > 15683 </div> 15684 <div class="slot-authors"> 15685 Xudong Wang, Gerald Jones, and Xueping Li (University of 15686 Tennessee, Knoxville) 15687 </div> 15688 <div class="slot-abstract"> 15689 <div> 15690 <a 15691 class="clickable no-decoration" 15692 id="vhsjs_view_340_1707793552_0333295" 15693 onclick="$('#vhsjs_view_340_1707793552_0333295').hide(); 15694 $('#vhsjs_hide_340_1707793552_0333295').show(); 15695 $('#339_1707793552_0333207').slideDown(function() { 15696 if (typeof Masonry === 'function') { 15697 $('.use_masonry').masonry(); 15698 }; 15699 15700 });" 15701 ><i class="fa fa-caret-right"></i> 15702 <span class="hover_link">Abstract</span></a 15703 ><a 15704 class="clickable no-decoration" 15705 id="vhsjs_hide_340_1707793552_0333295" 15706 onclick="$('#339_1707793552_0333207').hide(function() { 15707 if (typeof Masonry === 'function') { 15708 $('.use_masonry').masonry(); 15709 }; 15710 }); 15711 $('#vhsjs_hide_340_1707793552_0333295').hide(); 15712 $('#vhsjs_view_340_1707793552_0333295').show();" 15713 style="display: none" 15714 ><i class="fa fa-caret-down"></i> 15715 <span class="hover_link">Abstract</span></a 15716 > 15717 <div 15718 data-display-control="340_1707793552_0333295" 15719 id="339_1707793552_0333207" 15720 style="display: none" 15721 > 15722 <div class="arrow-slidedown"> 15723 <blockquote> 15724 Drone delivery is a popular logistics method for 15725 e-commerce businesses due to its efficie
15725ncy and 15726 convenience, especially for last-mile delivery 15727 and emergency situations in areas with poor 15728 infrastructure. However, the uncertainty of 15729 customer demands can affect transportation costs 15730 in the long run, making it vital to design an 15731 effective delivery system. To tackle this issue, 15732 we propose a two-stage stochastic model that 15733 minimizes the sum of fixed and expected 15734 operating costs. The first stage minimizes the 15735 total cost of the delivery system, including the 15736 facilities fixed costs and expected operating 15737 costs, while the second stage arranges drones' 15738 routes according to simulated demands to 15739 estimate the minimal expected transportation 15740 cost and penalty cost. Since this stoch
15740astic 15741 programming has infinite scenarios, we deploy a 15742 sample average approximation method to estimate 15743 its bounds. Additionally, we use a heuristic 15744 simulation framework to find a satisfactory 15745 solution in an acceptable time. 15746 </blockquote> 15747 </div> 15748 </div> 15749 </div> 15750 </div> 15751 <div class="slot-urls"></div> 15752 <a href="/wsc23papers/156.pdf" target="_blank">pdf</a 15753 ><br /> 15754 </div> 15755 </div> 15756 </div> 15757 <div class="centered"> 15758 <div class="top-link"><a href="#top">Return to Top</a></div> 15759 </div> 15760 <hr /> 15761 </div> 15762 <div class="area-section"> 15763 <div class="centered"> 15764 <a name="ptrack123" tabindex="-1"></a> 15765 <div class="section-title"> 15766 Manufacturing and Industry 4.0 15767 </div> 15768 </div> 15769 <div class="centered track-chair"> 15770 <span class="track-chair-role" 15771 >Track Coordinator - Manufacturing and Industry 4.0: </span 15772 ><span class="track-chair-names" 15773 >Alp Akcay (Eindhoven University of Technology), Christoph 15774 Laroque (University of Applied Sciences Zwickau), Guodong 15775 Shao (National Institute of Standards and Technology)</span 15776 > 15777 </div> 15778 <div class="section-entry"> 15779 <div class="session-entry"> 15780 <span class="session-event-type">Technical Session</span 15781 ><span class="type-track-spacer"> · </span 15782 ><span class="program-track" 15783 >Manufacturing and Industry 4.0</span 15784 ><br /> 15785 <div class="session-title"> 15786 Panel: Maintenance and Operations of Manufacturing Digital 15787 Twins 15788 </div> 15789 <div class="session-chair"> 15790 Chair: Alp Akcay (Eindhoven University of Technology)<br /> 15791 </div> 15792 <div class="slot-entry"> 15793 <a name="inv206" tabindex="-1"></a> 15794 <div class="slot-title-line"> 15795 <span class="slot-title" 15796 >Maintenance and Operations of Manufacturing Digital 15797 Twins</span 15798 > 15799 </div> 15800 <div class="slot-authors"> 15801 Alp Akcay (Eindhoven University of Technology), Stephan 15802 Biller (Purdue University), Boon Ping Gan (D-SIMLAB 15803 Technologies Pte Ltd), Christoph Laroque (University of 15804 Applied Sciences Zwickau), and Guodong Shao (National 15805 Institute of Standards and Technology) 15806 </div> 15807 <div class="slot-abstract"> 15808 <div> 15809 <a 15810 class="clickable no-decoration" 15811 id="vhsjs_view_342_1707793552_0393229" 15812 onclick="$('#vhsjs_view_342_1707793552_0393229').hide(); 15813 $('#vhsjs_hide_342_1707793552_0393229').show(); 15814 $('#341_1707793552_0393143').slideDown(function() { 15815 if (typeof Masonry === 'function') { 15816 $('.use_masonry').masonry(); 15817 }; 15818 15819 });" 15820 ><i class="fa fa-caret-right"></i> 15821 <span class="hover_link">Abstract</span></a 15822 ><a 15823 class="clickable no-decoration" 15824 id="vhsjs_hide_342_1707793552_0393229" 15825 onclick="$('#341_1707793552_0393143').hide(function() { 15826 if (typeof Masonry === 'function') { 15827 $('.use_masonry').masonry(); 15828 }; 15829 }); 15830 $('#vhsjs_hide_342_1707793552_0393229').hide(); 15831 $('#vhsjs_view_342_1707793552_0393229').show();" 15832 style="display: none" 15833 ><i class="fa fa-caret-down"></i>
15834 <span class="hover_link">Abstract</span></a 15835 > 15836 <div 15837 data-display-control="342_1707793552_0393229" 15838 id="341_1707793552_0393143" 15839 style="display: none" 15840 > 15841 <div class="arrow-slidedown"> 15842 <blockquote> 15843 Digital twins have become an important element 15844 in smart manufacturing. As any other product, 15845 digital twins also have a lifecycle, starting 15846 from specifying the requirements of the digital 15847 twins until their decommissioning. As part of 15848 the Manufacturing and Industry 4.0 track of the 15849 Winter Simulation Conference (WSC), the purpose 15850 of this panel is to discuss the state of the art 15851 in digital twins with a special emphasis on the 15852 operations and maintenance of manufacturing 15853 digital twins during their lifecycles. The 15854 panelists come from academia, industry, and 15855 government with experience in the digital-twin 15856 landscape of the manufacturing industry in the 15857 United States, Europe, and Asia. This paper 15858 provides a collection of the statements from 15859 each panelist with the objective of initiating a 15860 deeper discussion during the panel session and 15861 inspiring researchers in the simulation 15862 community with their perspectives on the use of 15863 digital twins for smart manufacturing. 15864 </blockquote> 15865 </div> 15866 </div> 15867 </div> 15868 </div> 15869 <div class="slot-urls"></div> 15870 <a href="/wsc23papers/157.pdf" target="_blank">pdf</a 15871 ><br /> 15872 </div> 15873 </div> 15874 <div class="session-entry"> 15875 <span class="session-event-type">Technical Session</span 15876 ><span class="type-track-spacer"> · </span 15877 ><span class="program-track" 15878 >Manufacturing and Industry 4.0</span 15879 ><br /> 15880 <div class="session-title"> 15881 Biomanufacturing and Process Industry 15882 </div> 15883 <div class="session-chair"> 15884 Chair: Daniel Seufferth (Universität der Bundeswehr 15885 München)<br /> 15886 </div> 15887 <div class="slot-entry"> 15888 <a name="inv208" tabindex="-1"></a> 15889 <div class="slot-title-line"> 15890 <span class="slot-title" 15891 >Stochastic Molecular Reaction Queueing Network 15892 Modeling for In Vitro Transcription Process</span 15893 > 15894 </div> 15895 <div class="slot-authors"> 15896 Keqi Wang, Wei Xie, and Hua Zheng (Northeastern 15897 University) 15898 </div> 15899 <div class="slot-abstract"> 15900 <div> 15901 <a 15902 class="clickable no-decoration" 15903 id="vhsjs_view_344_1707793552_0439746" 15904 onclick="$('#vhsjs_view_344_1707793552_0439746').hide(); 15905 $('#vhsjs_hide_344_1707793552_0439746').show(); 15906 $('#343_1707793552_0439663').slideDown(function() { 15907 if (typeof Masonry === 'function') { 15908 $('.use_masonry').masonry(); 15909 }; 15910 15911 });" 15912 ><i class="fa fa-caret-right"></i>
15913 <span class="hover_link">Abstract</span></a 15914 ><a 15915 class="clickable no-decoration" 15916 id="vhsjs_hide_344_1707793552_0439746" 15917 onclick="$('#343_1707793552_0439663').hide(function() { 15918 if (typeof Masonry === 'function') { 15919 $('.use_masonry').masonry(); 15920 }; 15921 }); 15922 $('#vhsjs_hide_344_1707793552_0439746').hide(); 15923 $('#vhsjs_view_344_1707793552_0439746').show();" 15924 style="display: none" 15925 ><i class="fa fa-caret-down"></i> 15926 <span class="hover_link">Abstract</span></a 15927 > 15928 <div 15929 data-display-control="344_1707793552_0439746" 15930 id="343_1707793552_0439663" 15931 style="display: none" 15932 > 15933 <div class="arrow-slidedown"> 15934 <blockquote> 15935 To facilitate a rapid response to pandemic 15936 threats, this paper focuses on developing a 15937 mechanistic simulation model for in vitro 15938 transcription (IVT) process, a crucial step in 15939 mRNA vaccine manufacturing. To enhance 15940 production and support industry 4.0, this model 15941 is proposed to improve the prediction and 15942 analysis of IVT enzymatic reaction network. It 15943 incorporates a novel stochastic molecular 15944 reaction queueing network with a regulatory 15945 kinetic model characterizing the effect of 15946 bioprocess state variables on reaction rates. 15947 The empirical study demonstrates that the 15948 proposed model has a promising performance under 15949 different production conditions and it could 15950 offer potential improvements in mRNA product 15951 quality and yield. 15952 </blockquote> 15953 </div> 15954 </div> 15955 </div> 15956 </div> 15957 <div class="slot-urls"></div> 15958 <a href="/wsc23papers/158.pdf" target="_blank">pdf</a 15959 ><br /> 15960 </div> 15961 <div class="slot-entry"> 15962 <a name="con186" tabindex="-1"></a> 15963 <div class="slot-title-line"> 15964 <span class="slot-title" 15965 >Rolling-Horizon Simulation Optimization for a 15966 Multi-Objective Biomanufacturing Scheduling 15967 Problem</span 15968 > 15969 </div> 15970 <div class="slot-authors"> 15971 Kim van den Houten, Mathijs de Weerdt, and David Tax 15972 (Delft University of Technology); Esteban Freydell 15973 (DSM); and Eva Christopoulou and Alessandro Nati 15974 (Systems Navigator) 15975 </div> 15976 <div class="slot-abstract"> 15977 <div> 15978 <a 15979 class="clickable no-decoration" 15980 id="vhsjs_view_346_1707793552_0466397" 15981 onclick="$('#vhsjs_view_346_1707793552_0466397').hide(); 15982 $('#vhsjs_hide_346_1707793552_0466397').show(); 15983 $('#345_1707793552_046632').slideDown(function() { 15984 if (typeof Masonry === 'function') { 15985 $('.use_masonry').masonry(); 15986 }; 15987 15988 });" 15989 ><i class="fa fa-caret-right"></i>
15990 <span class="hover_link">Abstract</span></a 15991 ><a 15992 class="clickable no-decoration" 15993 id="vhsjs_hide_346_1707793552_0466397" 15994 onclick="$('#345_1707793552_046632').hide(function() { 15995 if (typeof Masonry === 'function') { 15996 $('.use_masonry').masonry(); 15997 }; 15998 }); 15999 $('#vhsjs_hide_346_1707793552_0466397').hide(); 16000 $('#vhsjs_view_346_1707793552_0466397').show();" 16001 style="display: none" 16002 ><i class="fa fa-caret-down"></i> 16003 <span class="hover_link">Abstract</span></a 16004 > 16005 <div 16006 data-display-control="346_1707793552_0466397" 16007 id="345_1707793552_046632" 16008 style="display: none" 16009 > 16010 <div class="arrow-slidedown"> 16011 <blockquote> 16012 We study a highly complex scheduling problem 16013 that requires the generation and optimization of 16014 production schedules for a multi-product 16015 biomanufacturing system with continuous and 16016 batch processes. There are two main objectives 16017 here; makespan and lateness, which are combined 16018 into a cost function that is a weighted sum. An 16019 additional complexity comes from long horizons 16020 considered (up to a full year), yielding problem 16021 instances with more than 200 jobs, each 16022 consisting of multiple tasks that must be 16023 executed in the factory. We investigate whether 16024 a rolling-horizon principle is more efficient 16025 than a global strategy. We evaluate how cost 16026 function weights for makespan and lateness 16027 should be set in a rolling-horizon approach 16028 where deadlines are used for subproblem 16029 definition. We show that the rolling-horizon 16030 strategy outperforms a global search, evaluated 16031 on problem instances of a real biomanufacturing 16032 system, and we show that this result generalizes 16033 to problem instances of a synthetic factory. 16034 </blockquote> 16035 </div> 16036 </div> 16037 </div> 16038 </div> 16039 <div class="slot-urls"></div> 16040 <a href="/wsc23papers/159.pdf" target="_blank">pdf</a 16041 ><br /> 16042 </div> 16043 <div class="slot-entry"> 16044 <a name="cea156" tabindex="-1"></a> 16045 <div class="slot-title-line"> 16046 <span class="slot-title" 16047 >From Simulation To Real-Time Digital Twin and AI - 16048 Implementation in a Food Manufacturing Plant</span 16049 > 16050 </div> 16051 <div class="slot-authors"> 16052 Hosni Adra (CreateASoft, Inc) 16053 </div> 16054 <div class="slot-abstract"> 16055 <div> 16056 <a 16057 class="clickable no-decoration" 16058 id="vhsjs_view_348_1707793552_0486555" 16059 onclick="$('#vhsjs_view_348_1707793552_0486555').hide(); 16060 $('#vhsjs_hide_348_1707793552_0486555').show(); 16061 $('#347_1707793552_0486476').slideDown(function() { 16062 if (typeof Masonry === 'function') { 16063 $('.use_masonry').masonry(); 16064 }; 16065 16066 });" 16067 ><i class="fa fa-caret-right"></i>
16068 <span class="hover_link">Abstract</span></a 16069 ><a 16070 class="clickable no-decoration" 16071 id="vhsjs_hide_348_1707793552_0486555" 16072 onclick="$('#347_1707793552_0486476').hide(function() { 16073 if (typeof Masonry === 'function') { 16074 $('.use_masonry').masonry(); 16075 }; 16076 }); 16077 $('#vhsjs_hide_348_1707793552_0486555').hide(); 16078 $('#vhsjs_view_348_1707793552_0486555').show();" 16079 style="display: none" 16080 ><i class="fa fa-caret-down"></i> 16081 <span class="hover_link">Abstract</span></a 16082 > 16083 <div 16084 data-display-control="348_1707793552_0486555" 16085 id="347_1707793552_0486476" 16086 style="display: none" 16087 > 16088 <div class="arrow-slidedown"> 16089 <blockquote> 16090 Data-Driven simulation models are valuable tools 16091 to improve the accuracy of the models and enable 16092 them to transition to real-time predictive 16093 analytics tools. Adding AI (Artificial 16094 Intelligence) and ML (Machine Learning) enables 16095 those model to provide feedback and real-time 16096 optimization in un-attended environment. This 16097 paper details the steps and benefits that were 16098 used to implement such system in a large filling 16099 and packaging manufacturing setting, from 16100 initial randomized models to full real-time 16101 digital twin systems. Final models were used to 16102 optimize (real-time and offline) changeover, CIP 16103 (Clean in Place), production, filling lines, and 16104 material handling. 16105 </blockquote> 16106 </div> 16107 </div> 16108 </div> 16109 </div> 16110 <div class="slot-urls"></div> 16111 <a href="/wsc23papers/cea156.pdf" target="_blank">pdf</a 16112 ><br /> 16113 </div> 16114 </div> 16115 <div class="session-entry"> 16116 <span class="session-event-type">Technical Session</span 16117 ><span class="type-track-spacer"> · </span 16118 ><span class="program-track" 16119 >Manufacturing and Industry 4.0</span 16120 ><br /> 16121 <div class="session-title"> 16122 Deep Reinforcement Learning Applications 16123 </div> 16124 <div class="session-chair"> 16125 Chair: Alp Akcay (Eindhoven University of Technology)<br /> 16126 </div> 16127 <div class="slot-entry"> 16128 <a name="con127" tabindex="-1"></a> 16129 <div class="slot-title-line"> 16130 <span class="slot-title" 16131 >Semiconductor Fab Scheduling with Self-Supervised and 16132 Reinforcement Learning</span 16133 > 16134 </div> 16135 <div> 16136 <span class="BAP award" 16137 >Best Contributed Applied Paper - Finalist</span 16138 > 16139 </div> 16140 <div class="slot-authors"> 16141 Pierre Tassel and Benjamin Kovács 16142 (Alpen-Adria-Universität Klagenfurt); Martin Gebser 16143 (Alpen-Adria-Universität Klagenfurt, Graz 16144 University of Technology); Konstantin Schekotihin 16145 (Alpen-Adria-Universität Klagenfurt); and Patrick 16146 Stöckermann and Georg Seidel (Infineon Technologies 16147 AG) 16148 </div> 16149 <div class="slot-abstract"> 16150 <div> 16151 <a 16152 class="clickable no-decoration" 16153 id="vhsjs_view_350_1707793552_0551672" 16154 onclick="$('#vhsjs_view_350_1707793552_0551672').hide(); 16155 $('#vhsjs_hide_350_1707793552_0551672').show(); 16156 $('#349_1707793552_0551488').slideDown(function() { 16157 if (typeof Masonry === 'function') { 16158 $('.use_masonry').masonry(); 16159 }; 16160 16161 });" 16162 ><i class="fa fa-caret-right"></i>
16163 <span class="hover_link">Abstract</span></a 16164 ><a 16165 class="clickable no-decoration" 16166 id="vhsjs_hide_350_1707793552_0551672" 16167 onclick="$('#349_1707793552_0551488').hide(function() { 16168 if (typeof Masonry === 'function') { 16169 $('.use_masonry').masonry(); 16170 }; 16171 }); 16172 $('#vhsjs_hide_350_1707793552_0551672').hide(); 16173 $('#vhsjs_view_350_1707793552_0551672').show();" 16174 style="display: none" 16175 ><i class="fa fa-caret-down"></i> 16176 <span class="hover_link">Abstract</span></a 16177 > 16178 <div 16179 data-display-control="350_1707793552_0551672" 16180 id="349_1707793552_0551488" 16181 style="display: none" 16182 > 16183 <div class="arrow-slidedown"> 16184 <blockquote> 16185 Semiconductor manufacturing is a complex, costly 16186 process involving a long sequence of operations 16187 on limited, expensive equipment. Recent chip 16188 shortages and their impacts have highlighted the 16189 importance of semiconductors in the global 16190 supply chains and how reliant on those our daily 16191 lives are. Due to the investment cost, 16192 environmental impact, and time scale needed to 16193 build new factories, it is difficult to ramp up 16194 production when demand spikes. This work 16195 introduces a method to successfully learn to 16196 schedule a semiconductor manufacturing facility 16197 more efficiently using deep reinforcement and 16198 self-supervised learning. We propose the first 16199 adaptive scheduling approach to handle complex, 16200 continuous, stochastic, dynamic, modern 16201 semiconductor manufacturing models. Our method 16202 outperforms the traditional hierarchical 16203 dispatching strategies typically used in 16204 semiconductor manufacturing plants, 16205 substantially reducing each order’s 16206 tardiness and time until completion. 16207 Consequently, our method yields a better 16208 allocation of resources in the semiconductor 16209 manufacturing process. 16210 </blockquote> 16211 </div> 16212 </div> 16213 </div> 16214 </div> 16215 <div class="slot-urls"></div> 16216 <a href="/wsc23papers/160.pdf" target="_blank">pdf</a 16217 ><br /> 16218 </div> 16219 <div class="slot-entry"> 16220 <a name="cea131" tabindex="-1"></a> 16221 <div class="slot-title-line"> 16222 <span class="slot-title" 16223 >Deep Reinforcement Learning with Discrete-event 16224 Simulation for Steel Plate Stacking Problem</span 16225 > 16226 </div> 16227 <div class="slot-authors"> 16228 SaeNal Sung and SookYoung Son (HD Korea Shipbuilding & 16229 Offshore Engineering); Young-in Cho, Hee-chang Yoon, and 16230 Jong Hun Woo (Seoul National University); and Jong-Ho 16231 Nam (Korea Maritime and Ocean University) 16232 </div> 16233 <div class="slot-abstract"> 16234 <div> 16235 <a 16236 class="clickable no-decoration" 16237 id="vhsjs_view_352_1707793552_0574236" 16238 onclick="$('#vhsjs_view_352_1707793552_0574236').hide(); 16239 $('#vhsjs_hide_352_1707793552_0574236').show(); 16240 $('#351_1707793552_0574155').slideDown(function() { 16241 if (typeof Masonry === 'function') { 16242 $('.use_masonry').masonry(); 16243 }; 16244 16245 });" 16246 ><i class="fa fa-caret-right"></i>
16247 <span class="hover_link">Abstract</span></a 16248 ><a 16249 class="clickable no-decoration" 16250 id="vhsjs_hide_352_1707793552_0574236" 16251 onclick="$('#351_1707793552_0574155').hide(function() { 16252 if (typeof Masonry === 'function') { 16253 $('.use_masonry').masonry(); 16254 }; 16255 }); 16256 $('#vhsjs_hide_352_1707793552_0574236').hide(); 16257 $('#vhsjs_view_352_1707793552_0574236').show();" 16258 style="display: none" 16259 ><i class="fa fa-caret-down"></i> 16260 <span class="hover_link">Abstract</span></a 16261 > 16262 <div 16263 data-display-control="352_1707793552_0574236" 16264 id="351_1707793552_0574155" 16265 style="display: none" 16266 > 16267 <div class="arrow-slidedown"> 16268 <blockquote> 16269 In shipyards, newly supplied steel plates from 16270 steel-making companies are stored in steel 16271 stockyards until they are retrieved according to 16272 the pre-determined cutting schedule. Steel 16273 plates are grouped into lots, and all steel 16274 plates of the identical lot are retrieved and 16275 transported into the cutting workshop at the 16276 same time. In this study, we developed the 16277 two-stage stacking algorithm to minimize the 16278 workload of overhead cranes for the rehandling 16279 work in the retrieval process. In the proposed 16280 algorithm, a reinforcement learning-based agent 16281 which learns the stacking policy in the 16282 simulation environment determines the initial 16283 stacking location of the steel plates only 16284 considering the cutting schedule. After the 16285 initial arrangement of steel plates is created, 16286 steel plates are reshuffled using the simulated 16287 annealing considering both the cutting schedule 16288 and lot information. 16289 </blockquote> 16290 </div> 16291 </div> 16292 </div> 16293 </div> 16294 <div class="slot-urls"></div> 16295 <a href="/wsc23papers/cea131.pdf" target="_blank">pdf</a 16296 ><br /> 16297 </div> 16298 <div class="slot-entry"> 16299 <a name="cea158" tabindex="-1"></a> 16300 <div class="slot-title-line"> 16301 <span class="slot-title" 16302 >Digital Twins and Deep Reinforcement Learning for 16303 Online Optimization of Scheduling Problems</span 16304 > 16305 </div> 16306 <div class="slot-authors"> 16307 Bulent Soykan and Ghaith Rabadi (University of Central 16308 Florida) 16309 </div> 16310 <div class="slot-abstract"> 16311 <div> 16312 <a 16313 class="clickable no-decoration" 16314 id="vhsjs_view_354_1707793552_0595815" 16315 onclick="$('#vhsjs_view_354_1707793552_0595815').hide(); 16316 $('#vhsjs_hide_354_1707793552_0595815').show(); 16317 $('#353_1707793552_0595732').slideDown(function() { 16318 if (typeof Masonry === 'function') { 16319 $('.use_masonry').masonry(); 16320 }; 16321 16322 });" 16323 ><i class="fa fa-caret-right"></i>
16324 <span class="hover_link">Abstract</span></a 16325 ><a 16326 class="clickable no-decoration" 16327 id="vhsjs_hide_354_1707793552_0595815" 16328 onclick="$('#353_1707793552_0595732').hide(function() { 16329 if (typeof Masonry === 'function') { 16330 $('.use_masonry').masonry(); 16331 }; 16332 }); 16333 $('#vhsjs_hide_354_1707793552_0595815').hide(); 16334 $('#vhsjs_view_354_1707793552_0595815').show();" 16335 style="display: none" 16336 ><i class="fa fa-caret-down"></i> 16337 <span class="hover_link">Abstract</span></a 16338 > 16339 <div 16340 data-display-control="354_1707793552_0595815" 16341 id="353_1707793552_0595732" 16342 style="display: none" 16343 > 16344 <div class="arrow-slidedown"> 16345 <blockquote> 16346 This paper presents an approach that combines 16347 data-driven digital twins (DTs) and deep 16348 reinforcement learning (DRL) to address the 16349 challenges of online optimization of scheduling 16350 problems, focusing specifically on the classic 16351 job shop scheduling problem. Traditional 16352 approaches to solving such problems often 16353 encounter limitations in handling uncertainties 16354 and dynamic environments. In this study, we 16355 explore the integration of DTs and DRL to 16356 enhance decision-making in scheduling problems. 16357 We investigate the adaptability of a Graph 16358 Neural Network model within the DRL framework, 16359 enabling the agent to learn optimal scheduling 16360 policies through interactions with the DT. The 16361 potential of this convergence to tackle modern 16362 scheduling complexities offers insights into the 16363 future of operations management. 16364 </blockquote> 16365 </div> 16366 </div> 16367 </div> 16368 </div> 16369 <div class="slot-urls"></div> 16370 <a href="/wsc23papers/cea158.pdf" target="_blank">pdf</a 16371 ><br /> 16372 </div> 16373 </div> 16374 <div class="session-entry"> 16375 <span class="session-event-type">Technical Session</span 16376 ><span class="type-track-spacer"> · </span 16377 ><span class="program-track" 16378 >Manufacturing and Industry 4.0</span 16379 ><br /> 16380 <div class="session-title">Manufacturing Operations</div> 16381 <div class="session-chair"> 16382 Chair: Klaus Altendorfer (Upper Austrian University of 16383 Applied Science)<br /> 16384 </div> 16385 <div class="slot-entry"> 16386 <a name="con130" tabindex="-1"></a> 16387 <div class="slot-title-line"> 16388 <span class="slot-title" 16389 >Modeling and Simulation for the Operative Service 16390 Delivery Planning in the Context of Product-Service 16391 Systems</span 16392 > 16393 </div> 16394 <div class="slot-authors"> 16395 Enes Alp (Ruhr-Universität Bochum); Michael Herzog 16396 (Centre for the Engineering of Smart Product-Service 16397 Systems (ZESS)); Furkan Ercan (Ruhr-Universität 16398 Bochum); and Bernd Kuhlenkötter 16399 (Ruhr-Universität Bochum, Centre for the 16400 Engineering of Smart Product-Service Systems (ZESS)) 16401 </div> 16402 <div class="slot-abstract"> 16403 <div> 16404 <a 16405 class="clickable no-decoration" 16406 id="vhsjs_view_356_1707793552_0639436" 16407 onclick="$('#vhsjs_view_356_1707793552_0639436').hide(); 16408 $('#vhsjs_hide_356_1707793552_0639436').show(); 16409 $('#355_1707793552_0639353').slideDown(function() { 16410 if (typeof Masonry === 'function') { 16411 $('.use_masonry').masonry(); 16412 }; 16413 16414 });" 16415 ><i class="fa fa-caret-right"></i>
16416 <span class="hover_link">Abstract</span></a 16417 ><a 16418 class="clickable no-decoration" 16419 id="vhsjs_hide_356_1707793552_0639436" 16420 onclick="$('#355_1707793552_0639353').hide(function() { 16421 if (typeof Masonry === 'function') { 16422 $('.use_masonry').masonry(); 16423 }; 16424 }); 16425 $('#vhsjs_hide_356_1707793552_0639436').hide(); 16426 $('#vhsjs_view_356_1707793552_0639436').show();" 16427 style="display: none" 16428 ><i class="fa fa-caret-down"></i> 16429 <span class="hover_link">Abstract</span></a 16430 > 16431 <div 16432 data-display-control="356_1707793552_0639436" 16433 id="355_1707793552_0639353" 16434 style="display: none" 16435 > 16436 <div class="arrow-slidedown"> 16437 <blockquote> 16438 Accelerated with the developments in the context 16439 of Industry 4.0, a new trend has established 16440 itself in the manufacturing industry within the 16441 last two decades. Companies started to offer 16442 integrated solutions such as Product-Service 16443 Systems (PSS). While the provision of PSS 16444 enables benefits like business model innovation 16445 or strengthening competitiveness, the 16446 exploitation of these benefits depends heavily 16447 on the decisions in the operative service 16448 delivery planning. This, however, is a complex 16449 task due to the huge solution space. Analytical 16450 methods reach their limitations when trying to 16451 find the optimal solution. Though different 16452 optimization algorithms were elaborated for this 16453 problem, the evaluation of their solutions is 16454 overly simplified, and thus, their 16455 expressiveness for the uncertain and dynamic 16456 reality remains questionable. This paper 16457 addresses these issues by demonstrating the 16458 modeling of an adaptive simulation model that 16459 can be used to gain a realistic evaluation of 16460 operative service delivery plans in PSS. 16461 </blockquote> 16462 </div> 16463 </div> 16464 </div> 16465 </div> 16466 <div class="slot-urls"></div> 16467 <a href="/wsc23papers/161.pdf" target="_blank">pdf</a 16468 ><br /> 16469 </div> 16470 <div class="slot-entry"> 16471 <a name="con168" tabindex="-1"></a> 16472 <div class="slot-title-line"> 16473 <span class="slot-title" 16474 >Simulation-Based Energy Reduction for a Lead-Acid 16475 Battery Production with Stochastic Maturation and 16476 Drying Processes</span 16477 > 16478 </div> 16479 <div class="slot-authors"> 16480 Balwin Bokor and Klaus Altendorfer (University of 16481 Applied Sciences Upper Austria) 16482 </div> 16483 <div class="slot-abstract"> 16484 <div> 16485 <a 16486 class="clickable no-decoration" 16487 id="vhsjs_view_358_1707793552_0662572" 16488 onclick="$('#vhsjs_view_358_1707793552_0662572').hide(); 16489 $('#vhsjs_hide_358_1707793552_0662572').show(); 16490 $('#357_1707793552_0662484').slideDown(function() { 16491 if (typeof Masonry === 'function') { 16492 $('.use_masonry').masonry(); 16493 }; 16494 16495 });" 16496 ><i class="fa fa-caret-right"></i>
16497 <span class="hover_link">Abstract</span></a 16498 ><a 16499 class="clickable no-decoration" 16500 id="vhsjs_hide_358_1707793552_0662572" 16501 onclick="$('#357_1707793552_0662484').hide(function() { 16502 if (typeof Masonry === 'function') { 16503 $('.use_masonry').masonry(); 16504 }; 16505 }); 16506 $('#vhsjs_hide_358_1707793552_0662572').hide(); 16507 $('#vhsjs_view_358_1707793552_0662572').show();" 16508 style="display: none" 16509 ><i class="fa fa-caret-down"></i> 16510 <span class="hover_link">Abstract</span></a 16511 > 16512 <div 16513 data-display-control="358_1707793552_0662572" 16514 id="357_1707793552_0662484" 16515 style="display: none" 16516 > 16517 <div class="arrow-slidedown"> 16518 <blockquote> 16519 The reduction of carbon dioxide emissions is a 16520 major goal of the European Union and energy 16521 storage is a core aspect to reach this goal. 16522 However, the production of lead-acid batteries 16523 is very energy consuming. Based on a case 16524 company production system and data, we develop a 16525 simulation model for the most energy-intensive 16526 lead-acid battery production steps, i.e., 16527 ripening and drying of lead plates. As both 16528 processes have some non-controllable stochastic 16529 aspects, the planned process times for both 16530 steps are a crucial factor for overall energy 16531 consumption. Too low or too high planned process 16532 times either lead to energy wasting for 16533 re-warm-up or to unnecessary energy consumption 16534 during processing. Simulation results reveal a 16535 significant energy reduction potential when 16536 optimizing planned process times, which 16537 increases when process uncertainty decreases. In 16538 addition, also the post-maturation and 16539 post-drying times are found to have a high 16540 influence on overall energy consumption. 16541 </blockquote> 16542 </div> 16543 </div> 16544 </div> 16545 </div> 16546 <div class="slot-urls"></div> 16547 <a href="/wsc23papers/162.pdf" target="_blank">pdf</a 16548 ><br /> 16549 </div> 16550 <div class="slot-entry"> 16551 <a name="cea150" tabindex="-1"></a> 16552 <div class="slot-title-line"> 16553 <span class="slot-title" 16554 >LNG CCS (Cargo Containment System) Manufacturing 16555 System using IoT Data and Schedule Simulation</span 16556 > 16557 </div> 16558 <div class="slot-authors"> 16559 Yonghee Kim and Eunsun Jeong (HDKSOE) 16560 </div> 16561 <div class="slot-abstract"> 16562 <div> 16563 <a 16564 class="clickable no-decoration" 16565 id="vhsjs_view_360_1707793552_068333" 16566 onclick="$('#vhsjs_view_360_1707793552_068333').hide(); 16567 $('#vhsjs_hide_360_1707793552_068333').show(); 16568 $('#359_1707793552_0683243').slideDown(function() { 16569 if (typeof Masonry === 'function') { 16570 $('.use_masonry').masonry(); 16571 }; 16572 16573 });" 16574 ><i class="fa fa-caret-right"></i>
16575 <span class="hover_link">Abstract</span></a 16576 ><a 16577 class="clickable no-decoration" 16578 id="vhsjs_hide_360_1707793552_068333" 16579 onclick="$('#359_1707793552_0683243').hide(function() { 16580 if (typeof Masonry === 'function') { 16581 $('.use_masonry').masonry(); 16582 }; 16583 }); 16584 $('#vhsjs_hide_360_1707793552_068333').hide(); 16585 $('#vhsjs_view_360_1707793552_068333').show();" 16586 style="display: none" 16587 ><i class="fa fa-caret-down"></i> 16588 <span class="hover_link">Abstract</span></a 16589 > 16590 <div 16591 data-display-control="360_1707793552_068333" 16592 id="359_1707793552_0683243" 16593 style="display: none" 16594 > 16595 <div class="arrow-slidedown"> 16596 <blockquote> 16597 Compared to other manufacturing industries, the 16598 shipbuilding industry has high uncertainties and 16599 volatility in resources such as manpower, space, 16600 and equipment. The labor-intensive, expansive 16601 yard spaces, and enclosed working areas of the 16602 shipbuilding industry make it difficult to 16603 aggregate and analyze data. The research effort 16604 presented in this extended abstract focuses on 16605 gathering production data using IoT technology 16606 and schedule simulation for the intent of 16607 reduction in uncertainty of project management. 16608 The gathered data from automated equipment can 16609 be employed to monitor production performance 16610 and conduct data-driven production management. 16611 It is possible to prevent from decreasing 16612 production performance and excluding input of 16613 batch production performance unrelated to actual 16614 work information. In addition, we use simulation 16615 to find the optimal solution for the purpose of 16616 load leveling in the process of establishing an 16617 LNG CCS manufacturing plan. 16618 </blockquote> 16619 </div> 16620 </div> 16621 </div> 16622 </div> 16623 <div class="slot-urls"></div> 16624 <a href="/wsc23papers/cea150.pdf" target="_blank">pdf</a 16625 ><br /> 16626 </div> 16627 </div> 16628 <div class="session-entry"> 16629 <span class="session-event-type">Technical Session</span 16630 ><span class="type-track-spacer"> · </span 16631 ><span class="program-track" 16632 >Manufacturing and Industry 4.0</span 16633 ><br /> 16634 <div class="session-title"> 16635 Manufacturing Intralogistics 16636 </div> 16637 <div class="session-chair"> 16638 Chair: Nitish Singh (Eindhoven University of 16639 Technology)<br /> 16640 </div> 16641 <div class="slot-entry"> 16642 <a name="con289" tabindex="-1"></a> 16643 <div class="slot-title-line"> 16644 <span class="slot-title" 16645 >Simulation-Based AGV Management with a Linear 16646 Dispatching Rule</span 16647 > 16648 </div> 16649 <div class="slot-authors"> 16650 Nitish Singh, Jeroen B.H.C. Didden, Alp Akcay, Tugce 16651 Martagan, and Ivo J.B.F. Adan (Eindhoven University of 16652 Technology) 16653 </div> 16654 <div class="slot-abstract"> 16655 <div> 16656 <a 16657 class="clickable no-decoration" 16658 id="vhsjs_view_362_1707793552_072798" 16659 onclick="$('#vhsjs_view_362_1707793552_072798').hide(); 16660 $('#vhsjs_hide_362_1707793552_072798').show(); 16661 $('#361_1707793552_0727897').slideDown(function() { 16662 if (typeof Masonry === 'function') { 16663 $('.use_masonry').masonry(); 16664 }; 16665 16666 });" 16667 ><i class="fa fa-caret-right"></i>
16668 <span class="hover_link">Abstract</span></a 16669 ><a 16670 class="clickable no-decoration" 16671 id="vhsjs_hide_362_1707793552_072798" 16672 onclick="$('#361_1707793552_0727897').hide(function() { 16673 if (typeof Masonry === 'function') { 16674 $('.use_masonry').masonry(); 16675 }; 16676 }); 16677 $('#vhsjs_hide_362_1707793552_072798').hide(); 16678 $('#vhsjs_view_362_1707793552_072798').show();" 16679 style="display: none" 16680 ><i class="fa fa-caret-down"></i> 16681 <span class="hover_link">Abstract</span></a 16682 > 16683 <div 16684 data-display-control="362_1707793552_072798" 16685 id="361_1707793552_0727897" 16686 style="display: none" 16687 > 16688 <div class="arrow-slidedown"> 16689 <blockquote> 16690 This paper considers the problem of real-time 16691 dispatching of a fleet of heterogeneous 16692 automated guided vehicles (AGVs) with battery 16693 constraints. The AGV fleet is heterogeneous in 16694 terms of material handling capabilities; some 16695 can tow loads, some can lift loads while others 16696 manipulate loads with the assistance of a 16697 robotic arm. Transport requests arrive in 16698 real-time and include a soft time window, with 16699 late delivery incurring tardiness costs. 16700 Transport requests need to be assigned to a 16701 capable AGV based on required material handling 16702 capabilities with the objective to minimize a 16703 weighted sum of tardiness costs of transport 16704 requests and travel costs of AGVs. In this 16705 paper, an AGV-specific linear dispatching rule 16706 (LDR) learning approach is proposed to assign 16707 AGVs to randomly arriving transport requests in 16708 real time over a finite horizon. The proposed 16709 approach is compared with a heuristic policy 16710 from practice by using real-world data provided 16711 by our industry partner. 16712 </blockquote> 16713 </div> 16714 </div> 16715 </div> 16716 </div> 16717 <div class="slot-urls"></div> 16718 <a href="/wsc23papers/163.pdf" target="_blank">pdf</a 16719 ><br /> 16720 </div> 16721 <div class="slot-entry"> 16722 <a name="cea157" tabindex="-1"></a> 16723 <div class="slot-title-line"> 16724 <span class="slot-title" 16725 >Analysis of Autonomous Mobile Robots in Warehousing 16726 Using a Digital Twin Simulation</span 16727 > 16728 </div> 16729 <div class="slot-authors"> 16730 Michael Sellen (CreateASoft, Inc) 16731 </div> 16732 <div class="slot-abstract"> 16733 <div> 16734 <a 16735 class="clickable no-decoration" 16736 id="vhsjs_view_364_1707793552_0748446" 16737 onclick="$('#vhsjs_view_364_1707793552_0748446').hide(); 16738 $('#vhsjs_hide_364_1707793552_0748446').show(); 16739 $('#363_1707793552_0748365').slideDown(function() { 16740 if (typeof Masonry === 'function') { 16741 $('.use_masonry').masonry(); 16742 }; 16743 16744 });" 16745 ><i class="fa fa-caret-right"></i>
16746 <span class="hover_link">Abstract</span></a 16747 ><a 16748 class="clickable no-decoration" 16749 id="vhsjs_hide_364_1707793552_0748446" 16750 onclick="$('#363_1707793552_0748365').hide(function() { 16751 if (typeof Masonry === 'function') { 16752 $('.use_masonry').masonry(); 16753 }; 16754 }); 16755 $('#vhsjs_hide_364_1707793552_0748446').hide(); 16756 $('#vhsjs_view_364_1707793552_0748446').show();" 16757 style="display: none" 16758 ><i class="fa fa-caret-down"></i> 16759 <span class="hover_link">Abstract</span></a 16760 > 16761 <div 16762 data-display-control="364_1707793552_0748446" 16763 id="363_1707793552_0748365" 16764 style="display: none" 16765 > 16766 <div class="arrow-slidedown"> 16767 <blockquote> 16768 The continued acceleration of e-commerce growth 16769 present a challenge for fulfillment centers to 16770 manage growing SKU counts and increased demand 16771 volatility while continuing to satisfy customer 16772 delivery expectations and maintain control over 16773 costs. Many fulfillment centers are turning to 16774 automated solutions such as Autonomous Mobile 16775 Robots in an effort to increase throughput and 16776 efficiency from existing facilities. AMRs move 16777 throughout the warehouse environment 16778 guidance-free and can be deployed bringing goods 16779 to person, bulk material movement and can work 16780 collaboratively with employees for picking 16781 applications. For warehouse operations 16782 management teams and AMR solution providers, 16783 identifying the optimum fleet size and 16784 deployment logic for current and projected 16785 demand is a crucial step in a successful 16786 adoption of this technology. Data-Driven 16787 modelling and simulation can be a useful asset 16788 when evaluating different solutions and 16789 requirements before installation as well as 16790 identifying opportunities for increased 16791 efficiency or expansion in existing operations. 16792 </blockquote> 16793 </div> 16794 </div> 16795 </div> 16796 </div> 16797 <div class="slot-urls"></div> 16798 <a href="/wsc23papers/cea157.pdf" target="_blank">pdf</a 16799 ><br /> 16800 </div> 16801 <div class="slot-entry"> 16802 <a name="con146" tabindex="-1"></a> 16803 <div class="slot-title-line"> 16804 <span class="slot-title" 16805 >Sequential Decision-Making Framework for Robotic 16806 Mobile Fulfillment System-Based Automated Kitting 16807 System</span 16808 > 16809 </div> 16810 <div class="slot-authors"> 16811 Jaeung Lee, Sungwook Jang, and Young Jae Jang (Korea 16812 Advanced Institute of Science and Technology) and Yooeui 16813 Jin, Il Kyu Lim, Seungmin Jeong, and Eoksu Sim (Global 16814 Technology Research Samsung Electronics) 16815 </div> 16816 <div class="slot-abstract"> 16817 <div> 16818 <a 16819 class="clickable no-decoration" 16820 id="vhsjs_view_366_1707793552_0773947" 16821 onclick="$('#vhsjs_view_366_1707793552_0773947').hide(); 16822 $('#vhsjs_hide_366_1707793552_0773947').show(); 16823 $('#365_1707793552_0773861').slideDown(function() { 16824 if (typeof Masonry === 'function') { 16825 $('.use_masonry').masonry(); 16826 }; 16827 16828 });" 16829 ><i class="fa fa-caret-right"></i>
16830 <span class="hover_link">Abstract</span></a 16831 ><a 16832 class="clickable no-decoration" 16833 id="vhsjs_hide_366_1707793552_0773947" 16834 onclick="$('#365_1707793552_0773861').hide(function() { 16835 if (typeof Masonry === 'function') { 16836 $('.use_masonry').masonry(); 16837 }; 16838 }); 16839 $('#vhsjs_hide_366_1707793552_0773947').hide(); 16840 $('#vhsjs_view_366_1707793552_0773947').show();" 16841 style="display: none" 16842 ><i class="fa fa-caret-down"></i> 16843 <span class="hover_link">Abstract</span></a 16844 > 16845 <div 16846 data-display-control="366_1707793552_0773947" 16847 id="365_1707793552_0773861" 16848 style="display: none" 16849 > 16850 <div class="arrow-slidedown"> 16851 <blockquote> 16852 In a flexible production line capable of 16853 producing various product types within a single 16854 assembly line, an efficient parts supply is 16855 critical. The kitting feeding policy, 16856 implemented in the flexible production line, 16857 aims to kit and supply the necessary parts to 16858 the production line without delay. This study 16859 investigates the kitting feeding operation for 16860 Samsung Electronics’ surface-mount device 16861 production line. To facilitate the timely supply 16862 of parts required for surface-mount device 16863 production, Samsung Electronics introduced a 16864 robotic mobile fulfillment system-based 16865 automated kitting system. This research proposes 16866 a sequential decision-making framework to 16867 address the kitting operation optimization 16868 problem, as well as a kitting scheduling 16869 algorithm within the proposed framework. A 16870 simulation environment has been implemented to 16871 verify the performance of the proposed framework 16872 and algorithm through a series of experiments. 16873 The experimental results indicate that the 16874 proposed framework enhances operational 16875 performance and maintains stability, even as the 16876 problem size expands. 16877 </blockquote> 16878 </div> 16879 </div> 16880 </div> 16881 </div> 16882 <div class="slot-urls"></div> 16883 <a href="/wsc23papers/164.pdf" target="_blank">pdf</a 16884 ><br /> 16885 </div> 16886 </div> 16887 <div class="session-entry"> 16888 <span class="session-event-type">Technical Session</span 16889 ><span class="type-track-spacer"> · </span 16890 ><span class="program-track" 16891 >Manufacturing and Industry 4.0</span 16892 ><br /> 16893 <div class="session-title"> 16894 Case Studies in Manufacturing I 16895 </div> 16896 <div class="session-chair"> 16897 Chair: David T. Sturrock (Simio LLC)<br /> 16898 </div> 16899 <div class="slot-entry"> 16900 <a name="cea139" tabindex="-1"></a> 16901 <div class="slot-title-line">
16902 <span class="slot-title" 16903 >Simulation of SKU Slotting in Lift Truck 16904 Manufacturing Facility Warehouse: Raymond Corporation, 16905 Iowa</span 16906 > 16907 </div> 16908 <div class="slot-authors"> 16909 Jay Amer (University of Tennessee, Knoxville; N. J. 16910 Malin); Xueping Li (University of Tennessee, Knoxville); 16911 and Michael Bambino (N. J. Malin) 16912 </div> 16913 <div class="slot-abstract"> 16914 <div> 16915 <a 16916 class="clickable no-decoration" 16917 id="vhsjs_view_368_1707793552_0829773" 16918 onclick="$('#vhsjs_view_368_1707793552_0829773').hide(); 16919 $('#vhsjs_hide_368_1707793552_0829773').show(); 16920 $('#367_1707793552_0829692').slideDown(function() { 16921 if (typeof Masonry === 'function') { 16922 $('.use_masonry').masonry(); 16923 }; 16924 16925 });" 16926 ><i class="fa fa-caret-right"></i> 16927 <span class="hover_link">Abstract</span></a 16928 ><a 16929 class="clickable no-decoration" 16930 id="vhsjs_hide_368_1707793552_0829773" 16931 onclick="$('#367_1707793552_0829692').hide(function() { 16932 if (typeof Masonry === 'function') { 16933 $('.use_masonry').masonry(); 16934 }; 16935 }); 16936 $('#vhsjs_hide_368_1707793552_0829773').hide(); 16937 $('#vhsjs_view_368_1707793552_0829773').show();" 16938 style="display: none" 16939 ><i class="fa fa-caret-down"></i> 16940 <span class="hover_link">Abstract</span></a 16941 > 16942 <div 16943 data-display-control="368_1707793552_0829773" 16944 id="367_1707793552_0829692" 16945 style="display: none" 16946 > 16947 <div class="arrow-slidedown"> 16948 <blockquote> 16949 This objective of this simulation was to 16950 estimate the impact of optimizing parts slotting 16951 on picking throughput within the existing 16952 Raymond Corporation lift truck manufacturing 16953 facility warehouse in Iowa. The simulation 16954 demonstrated that slotting can results in a 16955 67.89% increase in picking throughput. This 16956 increase exceeded production requirements and 16957 eliminated the need to outsource picking. 16958 </blockquote> 16959 </div> 16960 </div> 16961 </div> 16962 </div> 16963 <div class="slot-urls"></div> 16964 <a href="/wsc23papers/cea139.pdf" target="_blank">pdf</a 16965 ><br /> 16966 </div> 16967 <div class="slot-entry"> 16968 <a name="cea163" tabindex="-1"></a> 16969 <div class="slot-title-line"> 16970 <span class="slot-title" 16971 >Simulating the Material Delivery Process for an 16972 Automotive Body Shop</span 16973 > 16974 </div> 16975 <div class="slot-authors"> 16976 Joseph Hugan (TriMech, LLC) 16977 </div> 16978 <div class="slot-abstract"> 16979 <div> 16980 <a 16981 class="clickable no-decoration" 16982 id="vhsjs_view_370_1707793552_0849595" 16983 onclick="$('#vhsjs_view_370_1707793552_0849595').hide(); 16984 $('#vhsjs_hide_370_1707793552_0849595').show(); 16985 $('#369_1707793552_0849514').slideDown(function() { 16986 if (typeof Masonry === 'function') { 16987 $('.use_masonry').masonry(); 16988 }; 16989 16990 });" 16991 ><i class="fa fa-caret-right"></i>
16992 <span class="hover_link">Abstract</span></a 16993 ><a 16994 class="clickable no-decoration" 16995 id="vhsjs_hide_370_1707793552_0849595" 16996 onclick="$('#369_1707793552_0849514').hide(function() { 16997 if (typeof Masonry === 'function') { 16998 $('.use_masonry').masonry(); 16999 }; 17000 }); 17001 $('#vhsjs_hide_370_1707793552_0849595').hide(); 17002 $('#vhsjs_view_370_1707793552_0849595').show();" 17003 style="display: none" 17004 ><i class="fa fa-caret-down"></i> 17005 <span class="hover_link">Abstract</span></a 17006 > 17007 <div 17008 data-display-control="370_1707793552_0849595" 17009 id="369_1707793552_0849514" 17010 style="display: none" 17011 > 17012 <div class="arrow-slidedown"> 17013 <blockquote> 17014 Increasing product customization and a continual 17015 need for higher productivity has led to more 17016 complex automotive vehicles being built in more 17017 compressed spaces. The material delivery 17018 networks supporting these processes have also 17019 had to adapt to deliver a wider variety of parts 17020 in smaller packaging at an increasing frequency. 17021 The author will discuss the development and 17022 analysis of an automotive delivery network 17023 simulation with a focus on delivery times, the 17024 resources required, the data model used to drive 17025 the simulation and the analytical techniques 17026 used during the project. The presentation will 17027 also include a discussion on the model 17028 construction, the time required to construct the 17029 model, and the challenges encountered in the 17030 project. 17031 </blockquote> 17032 </div> 17033 </div> 17034 </div> 17035 </div> 17036 <div class="slot-urls"></div> 17037 <a href="/wsc23papers/cea163.pdf" target="_blank">pdf</a 17038 ><br /> 17039 </div> 17040 <div class="slot-entry"> 17041 <a name="cea103" tabindex="-1"></a> 17042 <div class="slot-title-line"> 17043 <span class="slot-title" 17044 >An Integrated System of Scheduling and Digital Twins 17045 for Ore Transportation Inside-Outside Steelworks</span 17046 > 17047 </div> 17048 <div class="slot-authors"> 17049 Shun Yamamoto and Akira Kumano (JFE Steel Corporation) 17050 </div> 17051 <div class="slot-abstract"> 17052 <div> 17053 <a 17054 class="clickable no-decoration" 17055 id="vhsjs_view_372_1707793552_0870142" 17056 onclick="$('#vhsjs_view_372_1707793552_0870142').hide(); 17057 $('#vhsjs_hide_372_1707793552_0870142').show(); 17058 $('#371_1707793552_0870059').slideDown(function() { 17059 if (typeof Masonry === 'function') { 17060 $('.use_masonry').masonry(); 17061 }; 17062 17063 });" 17064 ><i class="fa fa-caret-right"></i> 17065 <span class="hover_link">Abstract</span></a 17066 ><a 17067 class="clickable no-decoration" 17068 id="vhsjs_hide_372_1707793552_0870142" 17069 onclick="$('#371_1707793552_0870059').hide(function() { 17070 if (typeof Masonry === 'function') { 17071 $('.use_masonry').masonry(); 17072 }; 17073 }); 17074 $('#vhsjs_hide_372_1707793552_0870142').hide(); 17075 $('#vhsjs_view_372_1707793552_0870142').show();" 17076 style="display: none" 17077 ><i class="fa fa-caret-down"></i>
17078 <span class="hover_link">Abstract</span></a 17079 > 17080 <div 17081 data-display-control="372_1707793552_0870142" 17082 id="371_1707793552_0870059" 17083 style="display: none" 17084 > 17085 <div class="arrow-slidedown"> 17086 <blockquote> 17087 JFE Steel Corporation has developed an ore 17088 logistics optimizer to reduce transportation 17089 costs. Because the Japanese steel industry 17090 imports large quantities of raw materials, the 17091 huge cost of ship freight and demurrage fees has 17092 become a problem. This work presents the ore 17093 carrier scheduler which was developed using 17094 metaheuristics methods to minimize logistics 17095 costs. A strategy of consolidating various iron 17096 ore brands at a junction spot that super-large 17097 carriers can enter is suggested. A digital twin 17098 that represents the stockyard in the steelworks 17099 is developed using a discrete simulator to 17100 verify the feasibility of operations, confirming 17101 the possibility of reducing costs by more than 17102 10 % by utilizing this system. 17103 </blockquote> 17104 </div> 17105 </div> 17106 </div> 17107 </div> 17108 <div class="slot-urls"></div> 17109 <a href="/wsc23papers/cea103.pdf" target="_blank">pdf</a 17110 ><br /> 17111 </div> 17112 </div> 17113 <div class="session-entry"> 17114 <span class="session-event-type">Technical Session</span 17115 ><span class="type-track-spacer"> · </span 17116 ><span class="program-track" 17117 >Manufacturing and Industry 4.0</span 17118 ><br /> 17119 <div class="session-title">Predictive Maintenance</div> 17120 <div class="session-chair"> 17121 Chair: Christoph Laroque (University of Applied Sciences 17122 Zwickau)<br /> 17123 </div> 17124 <div class="slot-entry"> 17125 <a name="con235" tabindex="-1"></a> 17126 <div class="slot-title-line"> 17127 <span class="slot-title" 17128 >Simulation-Based Evaluation of Imperfect Predictive 17129 Maintenance Models in Discrete Manufacturing: A 17130 Procedure Model and Case Study</span 17131 > 17132 </div> 17133 <div class="slot-authors"> 17134 Clemens Gutschi, Nikolaus Furian, and Siegfried Voessner 17135 (Graz University of Technology) 17136 </div> 17137 <div class="slot-abstract"> 17138 <div> 17139 <a 17140 class="clickable no-decoration" 17141 id="vhsjs_view_374_1707793552_0913124" 17142 onclick="$('#vhsjs_view_374_1707793552_0913124').hide(); 17143 $('#vhsjs_hide_374_1707793552_0913124').show(); 17144 $('#373_1707793552_0913033').slideDown(function() { 17145 if (typeof Masonry === 'function') { 17146 $('.use_masonry').masonry(); 17147 }; 17148 17149 });" 17150 ><i class="fa fa-caret-right"></i> 17151 <span class="hover_link">Abstract</span></a 17152 ><a 17153 class="clickable no-decoration" 17154 id="vhsjs_hide_374_1707793552_0913124" 17155 onclick="$('#373_1707793552_0913033').hide(function() { 17156 if (typeof Masonry === 'function') { 17157 $('.use_masonry').masonry(); 17158 }; 17159 }); 17160 $('#vhsjs_hide_374_1707793552_0913124').hide(); 17161 $('#vhsjs_view_374_1707793552_0913124').show();" 17162 style="display: none" 17163 ><i class="fa fa-caret-down"></i>
17164 <span class="hover_link">Abstract</span></a 17165 > 17166 <div 17167 data-display-control="374_1707793552_0913124" 17168 id="373_1707793552_0913033" 17169 style="display: none" 17170 > 17171 <div class="arrow-slidedown"> 17172 <blockquote> 17173 The performance and reliability of production 17174 systems is greatly affected by sudden 17175 breakdowns. In order to avoid these unforeseen 17176 interruptions, predictive maintenance (PdM) 17177 systems are being widely used to predict 17178 failures and prevent outages by maintenance. The 17179 performance of PdM systems however depend 17180 heavily on precision and recall of prediction 17181 results. In the worst case, missing or false 17182 alarms can actually worsen the performance of an 17183 production system instead of improving it. We 17184 present a new procedural model which 17185 specifically focus on the imperfection of such 17186 PdM systems and estimate the impact of this 17187 unwanted property on the performance and 17188 economic aspects of a production system. The 17189 model is presented in all steps needed for 17190 implementation and evaluation and demonstrated 17191 in a realistic use case examining an interlinked 17192 production system with a simulation-based 17193 approach. 17194 </blockquote> 17195 </div> 17196 </div> 17197 </div> 17198 </div> 17199 <div class="slot-urls"></div> 17200 <a href="/wsc23papers/165.pdf" target="_blank">pdf</a 17201 ><br /> 17202 </div> 17203 <div class="slot-entry"> 17204 <a name="inv106" tabindex="-1"></a> 17205 <div class="slot-title-line"> 17206 <span class="slot-title" 17207 >Data-Driven Smart Maintenance Decision Analysis: A 17208 Drone Factory Demonstrator Combining Digital Twins and 17209 Adapted AHP</span 17210 > 17211 </div> 17212 <div class="slot-authors"> 17213 Paulo Victor Lopes (Aeronautics Institute of Technology) 17214 and Siyuan Chen, Juan Pablo González Sánchez, 17215 Ebru Turanoglu Bekar, Jon Bokrantz, and Anders Skoogh 17216 (Chalmers University of Technology) 17217 </div> 17218 <div class="slot-abstract"> 17219 <div> 17220 <a 17221 class="clickable no-decoration" 17222 id="vhsjs_view_376_1707793552_093721" 17223 onclick="$('#vhsjs_view_376_1707793552_093721').hide(); 17224 $('#vhsjs_hide_376_1707793552_093721').show(); 17225 $('#375_1707793552_0937126').slideDown(function() { 17226 if (typeof Masonry === 'function') { 17227 $('.use_masonry').masonry(); 17228 }; 17229 17230 });" 17231 ><i class="fa fa-caret-right"></i> 17232 <span class="hover_link">Abstract</span></a 17233 ><a 17234 class="clickable no-decoration" 17235 id="vhsjs_hide_376_1707793552_093721" 17236 onclick="$('#375_1707793552_0937126').hide(function() { 17237 if (typeof Masonry === 'function') { 17238 $('.use_masonry').masonry(); 17239 }; 17240 }); 17241 $('#vhsjs_hide_376_1707793552_093721').hide(); 17242 $('#vhsjs_view_376_1707793552_093721').show();" 17243 style="display: none" 17244 ><i class="fa fa-caret-down"></i>
17245 <span class="hover_link">Abstract</span></a 17246 > 17247 <div 17248 data-display-control="376_1707793552_093721" 17249 id="375_1707793552_0937126" 17250 style="display: none" 17251 > 17252 <div class="arrow-slidedown"> 17253 <blockquote> 17254 The concept of Digital Twins has gained 17255 significant attention in recent years due to its 17256 potential for improving the performance of 17257 production systems. One promising area for 17258 Digital Twins is Smart Maintenance, enabling the 17259 simulation of different strategies without 17260 disrupting operations in the real system. This 17261 study proposes a high-level framework to 17262 integrate Digital Twins to support Smart 17263 Maintenance data-driven decision making in 17264 production lines. We implement, then, a case 17265 study of a lab scale drone factory to 17266 demonstrate how the production line performance 17267 evaluation is made under different what-if 17268 maintenance scenarios. The effects of this Smart 17269 Maintenance decision analysis approach were 17270 evaluated according to Key Performance 17271 Indicators from literature. The identified 17272 contributions are: (i) Digital Twin demonstrator 17273 focused on smart maintenance; (ii) 17274 implementation of smart maintenance data-driven 17275 decision analysis concepts; (iii) design and 17276 evaluation of what-if maintenance scenarios. 17277 </blockquote> 17278 </div> 17279 </div> 17280 </div> 17281 </div> 17282 <div class="slot-urls"></div> 17283 <a href="/wsc23papers/166.pdf" target="_blank">pdf</a 17284 ><br /> 17285 </div> 17286 <div class="slot-entry"> 17287 <a name="inv102" tabindex="-1"></a> 17288 <div class="slot-title-line"> 17289 <span class="slot-title" 17290 >Understanding Stakeholder Requirements for Digital 17291 Twins in Manufacturing Maintenance</span 17292 > 17293 </div> 17294 <div class="slot-authors"> 17295 Siyuan Chen (Chalmers University of Technology); Paulo 17296 Victor Lopes (Aeronautics Institute of Technology, 17297 Federal University of Sao Paulo); and Juan Pablo 17298 González Sánchez, Ebru Turanoglu Bekar, Jon 17299 Bokrantz, and Anders Skoogh (Chalmers University of 17300 Technology) 17301 </div> 17302 <div class="slot-abstract"> 17303 <div> 17304 <a 17305 class="clickable no-decoration" 17306 id="vhsjs_view_378_1707793552_096278" 17307 onclick="$('#vhsjs_view_378_1707793552_096278').hide(); 17308 $('#vhsjs_hide_378_1707793552_096278').show(); 17309 $('#377_1707793552_0962694').slideDown(function() { 17310 if (typeof Masonry === 'function') { 17311 $('.use_masonry').masonry(); 17312 }; 17313 17314 });" 17315 ><i class="fa fa-caret-right"></i>
17316 <span class="hover_link">Abstract</span></a 17317 ><a 17318 class="clickable no-decoration" 17319 id="vhsjs_hide_378_1707793552_096278" 17320 onclick="$('#377_1707793552_0962694').hide(function() { 17321 if (typeof Masonry === 'function') { 17322 $('.use_masonry').masonry(); 17323 }; 17324 }); 17325 $('#vhsjs_hide_378_1707793552_096278').hide(); 17326 $('#vhsjs_view_378_1707793552_096278').show();" 17327 style="display: none" 17328 ><i class="fa fa-caret-down"></i> 17329 <span class="hover_link">Abstract</span></a 17330 > 17331 <div 17332 data-display-control="378_1707793552_096278" 17333 id="377_1707793552_0962694" 17334 style="display: none" 17335 > 17336 <div class="arrow-slidedown"> 17337 <blockquote> 17338 Digital twin has emerged as a key technology in 17339 the era of smart manufacturing and holds 17340 significant potential for maintenance. However, 17341 gaps remain in understanding stakeholders' 17342 requirements and how this technology support 17343 maintenance-related decisions. This paper aims 17344 to identify stakeholders' requirements for 17345 digital twin implementation and examine the role 17346 of digital twin in supporting maintenance 17347 actions and decision-making process. 17348 Semi-structured interviews and a workshop 17349 involving manufacturing practitioners and 17350 researchers were conducted to attain these 17351 goals. Furthermore, an in-depth qualitative 17352 analysis of the interview data was carried out. 17353 The results shed light on the current state of 17354 digital twin adoption, implementation 17355 challenges, requirements, supported decisions 17356 and actions, and future demand characteristics. 17357 By integrating the findings from the literature 17358 review and interview analysis, this study 17359 outlines the requirements for the digital twins 17360 as expressed by industry stakeholders that will 17361 be used and tested in the drone factory digital 17362 twin model. 17363 </blockquote> 17364 </div> 17365 </div> 17366 </div> 17367 </div> 17368 <div class="slot-urls"></div> 17369 <a href="/wsc23papers/167.pdf" target="_blank">pdf</a 17370 ><br /> 17371 </div> 17372 </div> 17373 <div class="session-entry"> 17374 <span class="session-event-type">Technical Session</span 17375 ><span class="type-track-spacer"> · </span 17376 ><span class="program-track" 17377 >Manufacturing and Industry 4.0</span 17378 ><br /> 17379 <div class="session-title">Assembly Lines</div> 17380 <div class="session-chair"> 17381 Chair: Deogratias Kibira (National Institute of Standards 17382 and Technology, University of Maryland)<br /> 17383 </div> 17384 <div class="slot-entry"> 17385 <a name="con318" tabindex="-1"></a> 17386 <div class="slot-title-line">
17387 <span class="slot-title" 17388 >A Simulation-Based Approach for Line Balancing under 17389 Demand Uncertainty in Production Environment</span 17390 > 17391 </div> 17392 <div class="slot-authors"> 17393 S. M. Atikur Rahman and Md Fashiar Rahman (The 17394 University of Texas at El Paso), Tamanna Kamal (NC State 17395 University), and Tzu-Liang (Bill) Tseng (The University 17396 of Texas at El Paso) 17397 </div> 17398 <div class="slot-abstract"> 17399 <div> 17400 <a 17401 class="clickable no-decoration" 17402 id="vhsjs_view_380_1707793552_1028218" 17403 onclick="$('#vhsjs_view_380_1707793552_1028218').hide(); 17404 $('#vhsjs_hide_380_1707793552_1028218').show(); 17405 $('#379_1707793552_1028135').slideDown(function() { 17406 if (typeof Masonry === 'function') { 17407 $('.use_masonry').masonry(); 17408 }; 17409 17410 });" 17411 ><i class="fa fa-caret-right"></i> 17412 <span class="hover_link">Abstract</span></a 17413 ><a 17414 class="clickable no-decoration" 17415 id="vhsjs_hide_380_1707793552_1028218" 17416 onclick="$('#379_1707793552_1028135').hide(function() { 17417 if (typeof Masonry === 'function') { 17418 $('.use_masonry').masonry(); 17419 }; 17420 }); 17421 $('#vhsjs_hide_380_1707793552_1028218').hide(); 17422 $('#vhsjs_view_380_1707793552_1028218').show();" 17423 style="display: none" 17424 ><i class="fa fa-caret-down"></i> 17425 <span class="hover_link">Abstract</span></a 17426 > 17427 <div 17428 data-display-control="380_1707793552_1028218" 17429 id="379_1707793552_1028135" 17430 style="display: none" 17431 > 17432 <div class="arrow-slidedown"> 17433 <blockquote> 17434 The management of production line is a 17435 challenging task due to the high level of 17436 uncertainty in demand, which can lead to 17437 unbalanced utilization of resources. This may 17438 result in a potential deterioration of 17439 management satisfaction in terms of 17440 cost-effectiveness. Therefore, it requires 17441 efficient tools to optimize resource 17442 utilization. With such inherent needs, this 17443 paper presents a simulation-based decision 17444 support framework for garments industries. The 17445 Discrete Event Simulation (DES) is used to model 17446 different scenarios for the operational 17447 processes. The procedure focuses on the line 17448 balancing technique, which aims to eliminate 17449 bottlenecks and optimize the production process 17450 by balancing the workload. The results of this 17451 study demonstrate the effectiveness of the line 17452 balancing technique in improving line 17453 efficiency, reducing the idle time of the 17454 operators, and increasing productivity. The 17455 simulation was developed using AnyLogic 17456 simulation software. The outcome of the process 17457 is thoroughly evaluated and justified using a 17458 case study. 17459 </blockquote> 17460 </div> 17461 </div> 17462 </div> 17463 </div> 17464 <div class="slot-urls"></div> 17465 <a href="/wsc23papers/168.pdf" target="_blank">pdf</a 17466 ><br /> 17467 </div> 17468 <div class="slot-entry"> 17469 <a name="cea151" tabindex="-1"></a> 17470 <div class="slot-title-line">
17471 <span class="slot-title" 17472 >Optimization of Flat Block Assembly Line Using 17473 Constraint Programming and Discrete-Event 17474 Simulation</span 17475 > 17476 </div> 17477 <div class="slot-authors"> 17478 Dong Hoon Kwak and Jong Hun Woo (Seoul National 17479 University); Ki Young Cho (Seoul National University, 17480 Department of Naval Architecture and Ocean Engineering); 17481 and Hee Chang Yoon (Seoul National University) 17482 </div> 17483 <div class="slot-abstract"> 17484 <div> 17485 <a 17486 class="clickable no-decoration" 17487 id="vhsjs_view_382_1707793552_1049461" 17488 onclick="$('#vhsjs_view_382_1707793552_1049461').hide(); 17489 $('#vhsjs_hide_382_1707793552_1049461').show(); 17490 $('#381_1707793552_104938').slideDown(function() { 17491 if (typeof Masonry === 'function') { 17492 $('.use_masonry').masonry(); 17493 }; 17494 17495 });" 17496 ><i class="fa fa-caret-right"></i> 17497 <span class="hover_link">Abstract</span></a 17498 ><a 17499 class="clickable no-decoration" 17500 id="vhsjs_hide_382_1707793552_1049461" 17501 onclick="$('#381_1707793552_104938').hide(function() { 17502 if (typeof Masonry === 'function') { 17503 $('.use_masonry').masonry(); 17504 }; 17505 }); 17506 $('#vhsjs_hide_382_1707793552_1049461').hide(); 17507 $('#vhsjs_view_382_1707793552_1049461').show();" 17508 style="display: none" 17509 ><i class="fa fa-caret-down"></i> 17510 <span class="hover_link">Abstract</span></a 17511 > 17512 <div 17513 data-display-control="382_1707793552_1049461" 17514 id="381_1707793552_104938" 17515 style="display: none" 17516 > 17517 <div class="arrow-slidedown"> 17518 <blockquote> 17519 Scheduling of flat block assembly in a shipyard 17520 is crucial for productivity performance due to 17521 the high level of workload. This problem is 17522 commonly known as the permutation flowshop 17523 scheduling problem (PFSP) in operation research, 17524 which has been extensively studied in various 17525 papers since the 1950s. However, existing 17526 solutions often involve simplifying real-world 17527 problems with certain assumptions, limiting 17528 their practical applicability. In recent times, 17529 constraint programming (CP) has emerged as a 17530 strong alternative to exact algorithms and has 17531 been successfully applied to various PFSP, 17532 addressing the limitations of exact algorithms. 17533 In light of this, our study proposes a two-step 17534 optimization process to overcome the existing 17535 limitations composed of a CP and discrete-event 17536 simulation(DES). 17537 </blockquote> 17538 </div> 17539 </div> 17540 </div> 17541 </div> 17542 <div class="slot-urls"></div> 17543 <a href="/wsc23papers/cea151.pdf" target="_blank">pdf</a 17544 ><br /> 17545 </div> 17546 <div class="slot-entry"> 17547 <a name="con151" tabindex="-1"></a> 17548 <div class="slot-title-line">
17549 <span class="slot-title" 17550 >Digital Twin Architecture for a Flow Shop Assembly 17551 System</span 17552 > 17553 </div> 17554 <div class="slot-authors"> 17555 Gihan Lee and Seunghwan Chang (Ajou University), Onyu Yu 17556 and Jungik Yoon (LG Production and Research Institute), 17557 and Sangchul Park (Ajou University) 17558 </div> 17559 <div class="slot-abstract"> 17560 <div> 17561 <a 17562 class="clickable no-decoration" 17563 id="vhsjs_view_384_1707793552_1074636" 17564 onclick="$('#vhsjs_view_384_1707793552_1074636').hide(); 17565 $('#vhsjs_hide_384_1707793552_1074636').show(); 17566 $('#383_1707793552_1074555').slideDown(function() { 17567 if (typeof Masonry === 'function') { 17568 $('.use_masonry').masonry(); 17569 }; 17570 17571 });" 17572 ><i class="fa fa-caret-right"></i> 17573 <span class="hover_link">Abstract</span></a 17574 ><a 17575 class="clickable no-decoration" 17576 id="vhsjs_hide_384_1707793552_1074636" 17577 onclick="$('#383_1707793552_1074555').hide(function() { 17578 if (typeof Masonry === 'function') { 17579 $('.use_masonry').masonry(); 17580 }; 17581 }); 17582 $('#vhsjs_hide_384_1707793552_1074636').hide(); 17583 $('#vhsjs_view_384_1707793552_1074636').show();" 17584 style="display: none" 17585 ><i class="fa fa-caret-down"></i> 17586 <span class="hover_link">Abstract</span></a 17587 > 17588 <div 17589 data-display-control="384_1707793552_1074636" 17590 id="383_1707793552_1074555" 17591 style="display: none" 17592 > 17593 <div class="arrow-slidedown"> 17594 <blockquote> 17595 This paper proposes a digital twin architecture 17596 for a flow shop assembly line to maximize 17597 productivity and reduce quality costs. The 17598 proposed digital twin architecture consists of 17599 five major modules; Synchronization module to 17600 synchronize a real factory and the digital twin, 17601 Monitoring module to provide intuitive 17602 information visualization, Event calendar 17603 initialization module to initialize the factory 17604 state at any given time to the starting point of 17605 the CPS (Cyber-Physical System) simulation, CPS 17606 simulation module to identify potential 17607 production losses, and Decision-making module to 17608 take proactive actions to avoid anticipated 17609 production losses. The proposed digital twin 17610 architecture has been implemented for a home 17611 appliance factory of LG Electronics Co., Ltd. In 17612 South Korea, and shows significant improvements 17613 in terms of productivity, quality cost, and 17614 energy efficiency. 17615 </blockquote> 17616 </div> 17617 </div> 17618 </div> 17619 </div> 17620 <div class="slot-urls"></div> 17621 <a href="/wsc23papers/169.pdf" target="_blank">pdf</a 17622 ><br /> 17623 </div> 17624 </div> 17625 <div class="session-entry">
17626 <span class="session-event-type">Technical Session</span 17627 ><span class="type-track-spacer"> · </span 17628 ><span class="program-track" 17629 >Manufacturing and Industry 4.0</span 17630 ><br /> 17631 <div class="session-title">Simulation Approaches</div> 17632 <div class="session-chair"> 17633 Chair: Guodong Shao (National Institute of Standards and 17634 Technology)<br /> 17635 </div> 17636 <div class="slot-entry"> 17637 <a name="inv101" tabindex="-1"></a> 17638 <div class="slot-title-line"> 17639 <span class="slot-title" 17640 >Reverse Engineering the Future – An Automated 17641 Backward Simulation Approach to On-Time Production in 17642 the Semiconductor Industry</span 17643 > 17644 </div> 17645 <div class="slot-authors"> 17646 Madlene Leißau and Christoph Laroque (University of 17647 Applied Sciences Zwickau) 17648 </div> 17649 <div class="slot-abstract"> 17650 <div> 17651 <a 17652 class="clickable no-decoration" 17653 id="vhsjs_view_386_1707793552_1132822" 17654 onclick="$('#vhsjs_view_386_1707793552_1132822').hide(); 17655 $('#vhsjs_hide_386_1707793552_1132822').show(); 17656 $('#385_1707793552_1132739').slideDown(function() { 17657 if (typeof Masonry === 'function') { 17658 $('.use_masonry').masonry(); 17659 }; 17660 17661 });" 17662 ><i class="fa fa-caret-right"></i> 17663 <span class="hover_link">Abstract</span></a 17664 ><a 17665 class="clickable no-decoration" 17666 id="vhsjs_hide_386_1707793552_1132822" 17667 onclick="$('#385_1707793552_1132739').hide(function() { 17668 if (typeof Masonry === 'function') { 17669 $('.use_masonry').masonry(); 17670 }; 17671 }); 17672 $('#vhsjs_hide_386_1707793552_1132822').hide(); 17673 $('#vhsjs_view_386_1707793552_1132822').show();" 17674 style="display: none" 17675 ><i class="fa fa-caret-down"></i> 17676 <span class="hover_link">Abstract</span></a 17677 > 17678 <div 17679 data-display-control="386_1707793552_1132822" 17680 id="385_1707793552_1132739" 17681 style="display: none" 17682 > 17683 <div class="arrow-slidedown"> 17684 <blockquote> 17685 Researchers are investigating innovative 17686 techniques and tools to improve operational 17687 production planning, as manufacturing processes 17688 are increasingly influenced by new product 17689 demands, innovation, and cost-effectiveness. 17690 Backward-oriented discrete event simulation 17691 (SimBack) is one such tool that has shown great 17692 promise in this area. However, conducting 17693 multiple simulation runs for backward simulation 17694 can be time and resource-intensive, hampering 17695 its efficiency. To address this issue, this 17696 paper proposes an automated approach for 17697 executing and evaluating simulation experiments 17698 within the framework of backward-oriented 17699 discrete event simulation for scheduling and 17700 capacity planning. The authors illustrate their 17701 approach by applying it to a simulation model of 17702 the Semiconductor Manufacturing Testbed 2020 17703 (SMT2020). 17704 </blockquote> 17705 </div> 17706 </div> 17707 </div> 17708 </div> 17709 <div class="slot-urls"></div> 17710 <a href="/wsc23papers/170.pdf" target="_blank">pdf</a 17711 ><br /> 17712 </div> 17713 <div class="slot-entry"> 17714 <a name="con229" tabindex="-1"></a> 17715 <div class="slot-title-line">
17716 <span class="slot-title" 17717 >Using Kubernetes to Improve Data Farming 17718 Capabilities</span 17719 > 17720 </div> 17721 <div class="slot-authors"> 17722 Falk Stefan Pappert, Daniel Seufferth, Heiderose Stein, 17723 and Oliver Rose (University of the Bundeswehr Munich) 17724 </div> 17725 <div class="slot-abstract"> 17726 <div> 17727 <a 17728 class="clickable no-decoration" 17729 id="vhsjs_view_388_1707793552_1155794" 17730 onclick="$('#vhsjs_view_388_1707793552_1155794').hide(); 17731 $('#vhsjs_hide_388_1707793552_1155794').show(); 17732 $('#387_1707793552_1155713').slideDown(function() { 17733 if (typeof Masonry === 'function') { 17734 $('.use_masonry').masonry(); 17735 }; 17736 17737 });" 17738 ><i class="fa fa-caret-right"></i> 17739 <span class="hover_link">Abstract</span></a 17740 ><a 17741 class="clickable no-decoration" 17742 id="vhsjs_hide_388_1707793552_1155794" 17743 onclick="$('#387_1707793552_1155713').hide(function() { 17744 if (typeof Masonry === 'function') { 17745 $('.use_masonry').masonry(); 17746 }; 17747 }); 17748 $('#vhsjs_hide_388_1707793552_1155794').hide(); 17749 $('#vhsjs_view_388_1707793552_1155794').show();" 17750 style="display: none" 17751 ><i class="fa fa-caret-down"></i> 17752 <span class="hover_link">Abstract</span></a 17753 > 17754 <div 17755 data-display-control="388_1707793552_1155794" 17756 id="387_1707793552_1155713" 17757 style="display: none" 17758 > 17759 <div class="arrow-slidedown"> 17760 <blockquote> 17761 Simulation can reach computational limits, 17762 especially when running large-scale experiments. 17763 One possibility to counter this issue is 17764 distributed simulation. Recent developments in 17765 containerization and container orchestration 17766 technologies, such as Kubernetes, provide a 17767 stable and scalable infrastructure, that can 17768 serve distributed simulation. Although these 17769 solutions exist, applications within the 17770 simulation community remain scarce. Thus, in 17771 this paper, we present the general setup of such 17772 an infrastructure and discuss the application of 17773 an example case. Adding to the existing 17774 literature, we present our path forward and 17775 insights with different versions, as well as the 17776 efforts needed to construct similar 17777 implementations. As a result, we showcase the 17778 speed-up of simulation experimentation. We aim 17779 to provide a helpful foundation for others in 17780 our community to weigh the effort and benefit of 17781 such a system for their own projects. 17782 </blockquote> 17783 </div> 17784 </div> 17785 </div> 17786 </div> 17787 <div class="slot-urls"></div> 17788 <a href="/wsc23papers/171.pdf" target="_blank">pdf</a 17789 ><br /> 17790 </div> 17791 <div class="slot-entry"> 17792 <a name="cea155" tabindex="-1"></a> 17793 <div class="slot-title-line">
17794 <span class="slot-title" 17795 >Optimizing Production System Configurations across a 17796 Broad Design Space: A Case Study</span 17797 > 17798 </div> 17799 <div class="slot-authors"> 17800 Scott Nill and Larissa Nietner (LineLab, MIT) 17801 </div> 17802 <div class="slot-abstract"> 17803 <div> 17804 <a 17805 class="clickable no-decoration" 17806 id="vhsjs_view_390_1707793552_117693" 17807 onclick="$('#vhsjs_view_390_1707793552_117693').hide(); 17808 $('#vhsjs_hide_390_1707793552_117693').show(); 17809 $('#389_1707793552_1176848').slideDown(function() { 17810 if (typeof Masonry === 'function') { 17811 $('.use_masonry').masonry(); 17812 }; 17813 17814 });" 17815 ><i class="fa fa-caret-right"></i> 17816 <span class="hover_link">Abstract</span></a 17817 ><a 17818 class="clickable no-decoration" 17819 id="vhsjs_hide_390_1707793552_117693" 17820 onclick="$('#389_1707793552_1176848').hide(function() { 17821 if (typeof Masonry === 'function') { 17822 $('.use_masonry').masonry(); 17823 }; 17824 }); 17825 $('#vhsjs_hide_390_1707793552_117693').hide(); 17826 $('#vhsjs_view_390_1707793552_117693').show();" 17827 style="display: none" 17828 ><i class="fa fa-caret-down"></i> 17829 <span class="hover_link">Abstract</span></a 17830 > 17831 <div 17832 data-display-control="390_1707793552_117693" 17833 id="389_1707793552_1176848" 17834 style="display: none" 17835 > 17836 <div class="arrow-slidedown"> 17837 <blockquote> 17838 This paper presents a case study demonstrating 17839 the application of LineLab, a mathematical 17840 production system modeling tool, to optimize 17841 production system configurations and the ramp-up 17842 trajectory for novel mass timber building 17843 modules. The modeling tool can efficiently 17844 co-optimize a large number of variables, such as 17845 machine count, work-in-progress (WIP) count, 17846 average wait times, and throughput, thus helping 17847 to narrow down a broad design space. Sidewalk 17848 Labs, a Google company, faced unique challenges 17849 related to new product development, high-mix 17850 production, and phased ramp-up. This case study 17851 highlights the use of this mathematical 17852 optimization tool, and its integration with 17853 other simulation methodologies, resulting in an 17854 optimized digital pipeline for modeling the 17855 production scale-up for mass timber buildings. 17856 The insights provided contribute to the 17857 advancement of production optimization 17858 techniques and their applications across various 17859 industries. 17860 </blockquote> 17861 </div> 17862 </div> 17863 </div> 17864 </div> 17865 <div class="slot-urls"></div> 17866 <a href="/wsc23papers/cea155.pdf" target="_blank">pdf</a 17867 ><br /> 17868 </div> 17869 </div> 17870 <div class="session-entry">
17871 <span class="session-event-type">Technical Session</span 17872 ><span class="type-track-spacer"> · </span 17873 ><span class="program-track" 17874 >Manufacturing and Industry 4.0</span 17875 ><br /> 17876 <div class="session-title"> 17877 Manufacturing and Supply Chains 17878 </div> 17879 <div class="session-chair"> 17880 Chair: Thomas Felberbauer (St. Pölten University of 17881 Applied Sciences)<br /> 17882 </div> 17883 <div class="slot-entry"> 17884 <a name="inv165" tabindex="-1"></a> 17885 <div class="slot-title-line"> 17886 <span class="slot-title" 17887 >Modeling Risk Prioritization of a Manufacturing 17888 Supply Chain using Discrete Event Simulation</span 17889 > 17890 </div> 17891 <div class="slot-authors"> 17892 Arpita Chari and Silvan Marti (Chalmers University of 17893 Technology); Paulo Victor Lopes (Aeronautics Institute 17894 of Technology (ITA), Chalmers University of Technology); 17895 and Björn Johansson, Mélanie Despeisse, and 17896 Johan Stahre (Chalmers University of Technology) 17897 </div> 17898 <div class="slot-abstract"> 17899 <div> 17900 <a 17901 class="clickable no-decoration" 17902 id="vhsjs_view_392_1707793552_1228356" 17903 onclick="$('#vhsjs_view_392_1707793552_1228356').hide(); 17904 $('#vhsjs_hide_392_1707793552_1228356').show(); 17905 $('#391_1707793552_1228273').slideDown(function() { 17906 if (typeof Masonry === 'function') { 17907 $('.use_masonry').masonry(); 17908 }; 17909 17910 });" 17911 ><i class="fa fa-caret-right"></i> 17912 <span class="hover_link">Abstract</span></a 17913 ><a 17914 class="clickable no-decoration" 17915 id="vhsjs_hide_392_1707793552_1228356" 17916 onclick="$('#391_1707793552_1228273').hide(function() { 17917 if (typeof Masonry === 'function') { 17918 $('.use_masonry').masonry(); 17919 }; 17920 }); 17921 $('#vhsjs_hide_392_1707793552_1228356').hide(); 17922 $('#vhsjs_view_392_1707793552_1228356').show();" 17923 style="display: none" 17924 ><i class="fa fa-caret-down"></i> 17925 <span class="hover_link">Abstract</span></a 17926 > 17927 <div 17928 data-display-control="392_1707793552_1228356" 17929 id="391_1707793552_1228273" 17930 style="display: none" 17931 > 17932 <div class="arrow-slidedown"> 17933 <blockquote> 17934 Supply chains face a myriad of adverse risks 17935 that impact their daily operations and make them 17936 vulnerable. In addition, supply chains continue 17937 to grow in size and complexity which further 17938 sophisticates the problem. Lack of a structured 17939 approach and limitations in existing risk 17940 management methods contribute towards effective 17941 mitigation strategies not being properly 17942 developed. In this paper, we develop a discrete 17943 event simulation modelling approach to quantify 17944 the performance and risk assessment of a 17945 manufacturing supply chain in Swede
17945n which is 17946 under the impact of risks. This approach could 17947 support decision makers by prioritizing risks 17948 according to their performance impact and 17949 facilitating the development of mitigation 17950 strategies to enhance the resilience of the 17951 supply chain. The conceptual digital model can 17952 also be used to generate synthetic data to build 17953 an artificial intelligence-enhanced predictive 17954 demonstrator model to showcase capabilities for 17955 building data-driven resilience of the supply 17956 chain. 17957 </blockquote> 17958 </div> 17959 </div> 17960 </div> 17961 </div> 17962 <div class="slot-urls"></div> 17963 <a href="/wsc23papers/172.pdf" target="_blank">pdf</a 17964 ><br /> 17965 </div> 17966 <div class="slot-entry"> 17967 <a name="cea121" tabindex="-1"></a> 17968 <div class="slot-title-line"> 17969 <span class="slot-title" 17970 >A Simulation-Based Approach for Evaluating Different 17971 Model Mixes for Production Planning of a Contract 17972 Manufacturer in the Automotive Industry</span 17973 > 17974 </div> 17975 <div class="slot-authors"> 17976 Simon Gruber, Clemens Gutschi, Nikolaus Furian, and 17977 Siegfried Vössner (Graz University of Technology, 17978 Institute of Engineering- and Business Informatics) 17979 </div> 17980 <div class="slot-abstract"> 17981 <div> 17982 <a 17983 class="clickable no-decoration" 17984 id="vhsjs_view_394_1707793552_1249418" 17985 onclick="$('#vhsjs_view_394_1707793552_1249418').hide(); 17986 $('#vhsjs_hide_394_1707793552_1249418').show(); 17987 $('#393_1707793552_124934').slideDown(function() { 17988 if (typeof Masonry === 'function') { 17989 $('.use_masonry').masonry(); 17990 }; 17991 17992 });" 17993 ><i class="fa fa-caret-right"></i> 17994 <span class="hover_link">Abstract</span></a 17995 ><a 17996 class="clickable no-decoration" 17997 id="vhsjs_hide_394_1707793552_1249418" 17998 onclick="$('#393_1707793552_124934').hide(function() { 17999 if (typeof Masonry === 'function') { 18000 $('.use_masonry').masonry(); 18001 }; 18002 }); 18003 $('#vhsjs_hide_394_1707793552_1249418').hide(); 18004 $('#vhsjs_view_394_1707793552_1249418').show();" 18005 style="display: none" 18006 ><i class="fa fa-caret-down"></i> 18007 <span class="hover_link">Abstract</span></a 18008 > 18009 <div 18010 data-display-control="394_1707793552_1249418" 18011 id="393_1707793552_124934" 18012 style="display: none" 18013 > 18014 <div class="arrow-slidedown"> 18015 <blockquote> 18016 Contract manufacturers face challenges with 18017 short-term orders, cost pressures, and diverse 18018 customer requirements. Customer trends in the 18019 automotive industry intensify these challenges 18020 with reduced batch sizes and individual 18021 customization. Traditional analytic planning 18022 methods are insufficient for handling the 18023 complexity of modern manufacturing processes. 18024 Computational power alone cannot overcome this 18025 obstacle, careful modeling of production 18026 processes and resources is essential. Simulative 18027 approaches have been developed to address 18028 similar problems. In this use case, we aim to 18029 adapt and implement these approaches for a 18030 leading automotive contract manufacturer. A 18031 comprehensive assessment will then verify the 18032 adapted approach’s viability and 18033 potential. 18034 </blockquote> 18035 </div> 18036 </div> 18037 </div> 18038 </div> 18039 <div class="slot-urls"></div> 18040 <a href="/wsc23papers/cea121.pdf" target="_blank">pdf</a 18041 ><br /> 18042 </div> 18043 <div class="slot-entry"> 18044 <a name="inv174" tabindex="-1"></a> 18045 <div class="slot-title-line">
18046 <span class="slot-title" 18047 >Digital Twins for Supply Chains: Main Functions, 18048 Existing Applications, and Research 18049 Opportunities</span 18050 > 18051 </div> 18052 <div class="slot-authors"> 18053 Giovanni Lugaresi (KU Leuven); Zied Jemai 18054 (CentraleSupelec, Ecole Nationale d'Ingénieurs de 18055 Tunis); and Evren Sahin (CentraleSupelec) 18056 </div> 18057 <div class="slot-abstract"> 18058 <div> 18059 <a 18060 class="clickable no-decoration" 18061 id="vhsjs_view_396_1707793552_127147" 18062 onclick="$('#vhsjs_view_396_1707793552_127147').hide(); 18063 $('#vhsjs_hide_396_1707793552_127147').show(); 18064 $('#395_1707793552_1271389').slideDown(function() { 18065 if (typeof Masonry === 'function') { 18066 $('.use_masonry').masonry(); 18067 }; 18068 18069 });" 18070 ><i class="fa fa-caret-right"></i> 18071 <span class="hover_link">Abstract</span></a 18072 ><a 18073 class="clickable no-decoration" 18074 id="vhsjs_hide_396_1707793552_127147" 18075 onclick="$('#395_1707793552_1271389').hide(function() { 18076 if (typeof Masonry === 'function') { 18077 $('.use_masonry').masonry(); 18078 }; 18079 }); 18080 $('#vhsjs_hide_396_1707793552_127147').hide(); 18081 $('#vhsjs_view_396_1707793552_127147').show();" 18082 style="display: none" 18083 ><i class="fa fa-caret-down"></i> 18084 <span class="hover_link">Abstract</span></a 18085 > 18086 <div 18087 data-display-control="396_1707793552_127147" 18088 id="395_1707793552_1271389" 18089 style="display: none" 18090 > 18091 <div class="arrow-slidedown"> 18092 <blockquote> 18093 In recent times, manufacturing industries and 18094 their related supply chains have faced growing 18095 internal and external pressures. Due to the 18096 complex nature of global supply chain networks 18097 and the increased frequency of disruptive 18098 events, there is a pressing need to implement 18099 digital tools to support these industries. 18100 Digital twins have gained significant interest 18101 from industry and research communities due to 18102 their ability to provide valuable services in 18103 the short term. While there have been many 18104 contributions on digital twin-based 18105 methodologies for system design and production 18106 planning and control, the use of digital twins 18107 in supply chain management still needs to be 18108 improved. This paper presents an overview of the 18109 existing contributions on digital twins for 18110 supply chains. Starting from a preliminary 18111 literature review on the topic, relevant works 18112 are selected and used to identify insights on 18113 the current development level and future 18114 research opportunities. 18115 </blockquote> 18116 </div> 18117 </div> 18118 </div> 18119 </div> 18120 <div class="slot-urls"></div> 18121 <a href="/wsc23papers/173.pdf" target="_blank">pdf</a 18122 ><br /> 18123 </div> 18124 </div> 18125 <div class="session-entry">
18126 <span class="session-event-type">Technical Session</span 18127 ><span class="type-track-spacer"> · </span 18128 ><span class="program-track" 18129 >Manufacturing and Industry 4.0</span 18130 ><br /> 18131 <div class="session-title">Production Planning</div> 18132 <div class="session-chair"> 18133 Chair: Geert van Kollenburg (Eindhoven University of 18134 Technology)<br /> 18135 </div> 18136 <div class="slot-entry"> 18137 <a name="con274" tabindex="-1"></a> 18138 <div class="slot-title-line"> 18139 <span class="slot-title" 18140 >Investigating Production Yield Effect on Inventory 18141 Control Through a Hybrid Simulation Approach</span 18142 > 18143 </div> 18144 <div class="slot-authors"> 18145 Marina Materikina, Atefeh Shoomal, Linh Ho Manh, and 18146 Yuan Zhou (University of Texas Arlington) 18147 </div> 18148 <div class="slot-abstract"> 18149 <div> 18150 <a 18151 class="clickable no-decoration" 18152 id="vhsjs_view_398_1707793552_1352503" 18153 onclick="$('#vhsjs_view_398_1707793552_1352503').hide(); 18154 $('#vhsjs_hide_398_1707793552_1352503').show(); 18155 $('#397_1707793552_1352417').slideDown(function() { 18156 if (typeof Masonry === 'function') { 18157 $('.use_masonry').masonry(); 18158 }; 18159 18160 });" 18161 ><i class="fa fa-caret-right"></i> 18162 <span class="hover_link">Abstract</span></a 18163 ><a 18164 class="clickable no-decoration" 18165 id="vhsjs_hide_398_1707793552_1352503" 18166 onclick="$('#397_1707793552_1352417').hide(function() { 18167 if (typeof Masonry === 'function') { 18168 $('.use_masonry').masonry(); 18169 }; 18170 }); 18171 $('#vhsjs_hide_398_1707793552_1352503').hide(); 18172 $('#vhsjs_view_398_1707793552_1352503').show();" 18173 style="display: none" 18174 ><i class="fa fa-caret-down"></i> 18175 <span class="hover_link">Abstract</span></a 18176 > 18177 <div 18178 data-display-control="398_1707793552_1352503" 18179 id="397_1707793552_1352417" 18180 style="display: none" 18181 > 18182 <div class="arrow-slidedown"> 18183 <blockquote> 18184 Production Planning and Control (PPC) plays a 18185 key role in stabilizing and improving 18186 manufacturing processes under external and 18187 internal uncertainties by providing transparency 18188 in the whole system. This study focuses on PPC 18189 with internal uncertainties such as losses of 18190 work-in-process products during a contact lens 18191 manufacturing process. Although such losses are 18192 expected, the yield rates are uncertain and vary 18193 at different production stages. A hybrid 18194 agent-based simulation (ABS) and discrete-event 18195 simulation (DES) approach was utilized to 18196 resemble the underlying dynamics of the 18197 manufacturing system with uncertain yield rates. 18198 The results of the simulation experiments 18199 demonstrated that a simple average yield 18200 approach for production planning would cause 18201 potential backlogs and extra holding costs for 18202 the excess inventory. The proposed hybrid 18203 simulation could be used to support the 18204 decision-making process on a weekly basis to 18205 help a production planning team make a schedule 18206 that would improve efficiency and customer 18207 satisfaction. 18208 </blockquote> 18209 </div> 18210 </div> 18211 </div> 18212 </div> 18213 <div class="slot-urls"></div> 18214 <a href="/wsc23papers/174.pdf" target="_blank">pdf</a 18215 ><br /> 18216 </div> 18217 <div class="slot-entry"> 18218 <a name="con282" tabindex="-1"></a> 18219 <div class="slot-title-line">
18220 <span class="slot-title" 18221 >Stick to the Plan or Adjust Dynamically? Combining 18222 Order Release and Overtime Planning for Varying Demand 18223 and Process Uncertainty</span 18224 > 18225 </div> 18226 <div class="slot-authors"> 18227 Julian Fodor and Stefan Haeussler (University of 18228 Innsbruck) 18229 </div> 18230 <div class="slot-abstract"> 18231 <div> 18232 <a 18233 class="clickable no-decoration" 18234 id="vhsjs_view_400_1707793552_137571" 18235 onclick="$('#vhsjs_view_400_1707793552_137571').hide(); 18236 $('#vhsjs_hide_400_1707793552_137571').show(); 18237 $('#399_1707793552_137563').slideDown(function() { 18238 if (typeof Masonry === 'function') { 18239 $('.use_masonry').masonry(); 18240 }; 18241 18242 });" 18243 ><i class="fa fa-caret-right"></i> 18244 <span class="hover_link">Abstract</span></a 18245 ><a 18246 class="clickable no-decoration" 18247 id="vhsjs_hide_400_1707793552_137571" 18248 onclick="$('#399_1707793552_137563').hide(function() { 18249 if (typeof Masonry === 'function') { 18250 $('.use_masonry').masonry(); 18251 }; 18252 }); 18253 $('#vhsjs_hide_400_1707793552_137571').hide(); 18254 $('#vhsjs_view_400_1707793552_137571').show();" 18255 style="display: none" 18256 ><i class="fa fa-caret-down"></i> 18257 <span class="hover_link">Abstract</span></a 18258 > 18259 <div 18260 data-display-control="400_1707793552_137571" 18261 id="399_1707793552_137563" 18262 style="display: none" 18263 > 18264 <div class="arrow-slidedown"> 18265 <blockquote> 18266 Within the area of manufacturing planning and 18267 control there is a long ongoing debate on when 18268 and if decisions should be integrated to a 18269 centralized model or split to separate planning 18270 levels. While a centralized monolithic model is 18271 capable of solving separate decisions 18272 simultaneously, a hierarchical approach offers 18273 more degrees of freedom since a local planner 18274 always has more accurate information. The focus 18275 of this paper is on the design and mathematical 18276 assumptions of optimization models for overtime 18277 and order release decisions in order to cope 18278 with different degree of demand and process 18279 uncertainty. We execute the optimal decisions 18280 within a simulation model of a multi-stage, 18281 multi-product stylized flow shop. Our results 18282 show that a fully centralized is outperformed by 18283 a hierarchical design and that planning order 18284 release quantities centrally in combination with 18285 flexible overtime planning yields the lowest 18286 costs for high process uncertainty on the shop 18287 floor. 18288 </blockquote> 18289 </div> 18290 </div> 18291 </div> 18292 </div> 18293 <div class="slot-urls"></div> 18294 <a href="/wsc23papers/175.pdf" target="_blank">pdf</a 18295 ><br /> 18296 </div> 18297 <div class="slot-entry"> 18298 <a name="cea119" tabindex="-1"></a> 18299 <div class="slot-title-line">
18300 <span class="slot-title" 18301 >An MDP Model-Based Reinforcement Learning Approach 18302 for the Nesting Problem: A Case Study in Ship 18303 Design</span 18304 > 18305 </div> 18306 <div class="slot-authors"> 18307 SookYoung Son (Seoul National University, HD KSOE); 18308 YounHyun Kim and KiSun Kim (HD KSOE); and JongHun Woo 18309 (Seoul National University, Research Institute of Marine 18310 Systems Engineering) 18311 </div> 18312 <div class="slot-abstract"> 18313 <div> 18314 <a 18315 class="clickable no-decoration" 18316 id="vhsjs_view_402_1707793552_1397958" 18317 onclick="$('#vhsjs_view_402_1707793552_1397958').hide(); 18318 $('#vhsjs_hide_402_1707793552_1397958').show(); 18319 $('#401_1707793552_1397874').slideDown(function() { 18320 if (typeof Masonry === 'function') { 18321 $('.use_masonry').masonry(); 18322 }; 18323 18324 });" 18325 ><i class="fa fa-caret-right"></i> 18326 <span class="hover_link">Abstract</span></a 18327 ><a 18328 class="clickable no-decoration" 18329 id="vhsjs_hide_402_1707793552_1397958" 18330 onclick="$('#401_1707793552_1397874').hide(function() { 18331 if (typeof Masonry === 'function') { 18332 $('.use_masonry').masonry(); 18333 }; 18334 }); 18335 $('#vhsjs_hide_402_1707793552_1397958').hide(); 18336 $('#vhsjs_view_402_1707793552_1397958').show();" 18337 style="display: none" 18338 ><i class="fa fa-caret-down"></i> 18339 <span class="hover_link">Abstract</span></a 18340 > 18341 <div 18342 data-display-control="402_1707793552_1397958" 18343 id="401_1707793552_1397874" 18344 style="display: none" 18345 > 18346 <div class="arrow-slidedown"> 18347 <blockquote> 18348 The nesting problem in the shipbuilding industry 18349 calls for an increase in the utilization rates 18350 of plates and a decrease in the scrap ratio. To 18351 improve the efficiency of part nesting in ship 18352 design, this paper proposes an approach that 18353 uses a reinforcement learning algorithm to 18354 determine an efficient arrangement of parts. We 18355 frame the ship nesting problem as a Markov 18356 Decision Process (MDP) to apply the Proximal 18357 Policy Optimization (PPO) model, a reinforcement 18358 learning algorithm. A case study on a real-life 18359 nesting design is provided to validate and 18360 compare the proposed approach. 18361 </blockquote> 18362 </div> 18363 </div> 18364 </div> 18365 </div> 18366 <div class="slot-urls"></div> 18367 <a href="/wsc23papers/cea119.pdf" target="_blank">pdf</a 18368 ><br /> 18369 </div> 18370 </div> 18371 <div class="session-entry"> 18372 <span class="session-event-type">Technical Session</span 18373 ><span class="type-track-spacer"> · </span 18374 ><span class="program-track" 18375 >Manufacturing and Industry 4.0</span 18376 ><br /> 18377 <div class="session-title"> 18378 Case Studies in Manufacturing II 18379 </div> 18380 <div class="session-chair"> 18381 Chair: Molly Arthur (Simio)<br /> 18382 </div> 18383 <div class="slot-entry"> 18384 <a name="cea120" tabindex="-1"></a> 18385 <div class="slot-title-line">
18386 <span class="slot-title" 18387 >A Logistics Simulation Model Repository to Accelerate 18388 Simulation Modeling in the Aerospace Industry</span 18389 > 18390 </div> 18391 <div class="slot-authors"> 18392 Bjoern Goedecke (Airbus Operations), Philipp Braun 18393 (Hamburg University of Technology), Tobias Kuhrt (Airbus 18394 Aerostructures), Nadhir Mechai and Arne Anhalt 18395 (Accenture Industry X), Klaus Fischer and Helge Fromm 18396 (Airbus Operations), and Yannik Dreischhoff (Accenture 18397 Industry X) 18398 </div> 18399 <div class="slot-abstract"> 18400 <div> 18401 <a 18402 class="clickable no-decoration" 18403 id="vhsjs_view_404_1707793552_1443563" 18404 onclick="$('#vhsjs_view_404_1707793552_1443563').hide(); 18405 $('#vhsjs_hide_404_1707793552_1443563').show(); 18406 $('#403_1707793552_1443484').slideDown(function() { 18407 if (typeof Masonry === 'function') { 18408 $('.use_masonry').masonry(); 18409 }; 18410 18411 });" 18412 ><i class="fa fa-caret-right"></i> 18413 <span class="hover_link">Abstract</span></a 18414 ><a 18415 class="clickable no-decoration" 18416 id="vhsjs_hide_404_1707793552_1443563" 18417 onclick="$('#403_1707793552_1443484').hide(function() { 18418 if (typeof Masonry === 'function') { 18419 $('.use_masonry').masonry(); 18420 }; 18421 }); 18422 $('#vhsjs_hide_404_1707793552_1443563').hide(); 18423 $('#vhsjs_view_404_1707793552_1443563').show();" 18424 style="display: none" 18425 ><i class="fa fa-caret-down"></i> 18426 <span class="hover_link">Abstract</span></a 18427 > 18428 <div 18429 data-display-control="404_1707793552_1443563" 18430 id="403_1707793552_1443484" 18431 style="display: none" 18432 > 18433 <div class="arrow-slidedown"> 18434 <blockquote> 18435 Airbus established a digitalization strategy to 18436 enhance logistics and production processes using 18437 model-based systems engineering, including 18438 material flow simulation. To store and reuse 18439 simulation model, holdup quality standards and 18440 support logistics planning, novel to the 18441 aerospace industry, a logistics simulation 18442 repository is being developed. This is supported 18443 by presenting ongoing simulation studies. 18444 </blockquote> 18445 </div> 18446 </div> 18447 </div> 18448 </div> 18449 <div class="slot-urls"></div> 18450 <a href="/wsc23papers/cea120.pdf" target="_blank">pdf</a 18451 ><br /> 18452 </div> 18453 <div class="slot-entry"> 18454 <a name="cea141" tabindex="-1"></a> 18455 <div class="slot-title-line"> 18456 <span class="slot-title" 18457 >Specification, Simulation and Analysis of 18458 Alternatives for On-line Scheduling of Independent 18459 Jobs in Different Servers</span 18460 > 18461 </div> 18462 <div class="slot-authors"> 18463 Jaume Figueras Jové and Pau Fonseca Casas 18464 (Universitat Politècnica de Catalunya) 18465 </div> 18466 <div class="slot-abstract"> 18467 <div> 18468 <a 18469 class="clickable no-decoration" 18470 id="vhsjs_view_406_1707793552_1465776" 18471 onclick="$('#vhsjs_view_406_1707793552_1465776').hide(); 18472 $('#vhsjs_hide_406_1707793552_1465776').show(); 18473 $('#405_1707793552_1465693').slideDown(function() { 18474 if (typeof Masonry === 'function') { 18475 $('.use_masonry').masonry(); 18476 }; 18477 18478 });" 18479 ><i class="fa fa-caret-right"></i>
18480 <span class="hover_link">Abstract</span></a 18481 ><a 18482 class="clickable no-decoration" 18483 id="vhsjs_hide_406_1707793552_1465776" 18484 onclick="$('#405_1707793552_1465693').hide(function() { 18485 if (typeof Masonry === 'function') { 18486 $('.use_masonry').masonry(); 18487 }; 18488 }); 18489 $('#vhsjs_hide_406_1707793552_1465776').hide(); 18490 $('#vhsjs_view_406_1707793552_1465776').show();" 18491 style="display: none" 18492 ><i class="fa fa-caret-down"></i> 18493 <span class="hover_link">Abstract</span></a 18494 > 18495 <div 18496 data-display-control="406_1707793552_1465776" 18497 id="405_1707793552_1465693" 18498 style="display: none" 18499 > 18500 <div class="arrow-slidedown"> 18501 <blockquote> 18502 Service companies have the challenge to analyze 18503 a large number of documents in order to extract 18504 relevant information for decision making. Such 18505 analysis can be made automatically reducing 18506 drastically the time amount and human effort 18507 needed. However, the computer system must ensure 18508 that the analysis of each document will be 18509 completed within a specified period of time 18510 which depends on the type of the document. A 18511 real case study is presented in this paper where 18512 the objective is to propose a new scheduling 18513 model for a computer system with 6 servers with 18514 a total of 384 logical cores. The arrival of 18515 documents is aperiodic and the processing time 18516 stochastic though processing time estimation can 18517 be done based on the number of pages and the 18518 type of the document. A simulation model has 18519 been developed to analyze the quality of each 18520 algorithm. A delay maximum time (DMT) algorithm 18521 is also proposed. 18522 </blockquote> 18523 </div> 18524 </div> 18525 </div> 18526 </div> 18527 <div class="slot-urls"></div> 18528 <a href="/wsc23papers/cea141.pdf" target="_blank">pdf</a 18529 ><br /> 18530 </div> 18531 <div class="slot-entry"> 18532 <a name="con208" tabindex="-1"></a> 18533 <div class="slot-title-line"> 18534 <span class="slot-title" 18535 >Simulation-Based Analyses and Improvements of the 18536 Smart Line Management System in Canned Beverage 18537 Industry: A Case Study in Europe</span 18538 > 18539 </div> 18540 <div class="slot-authors"> 18541 Ahmad Attar, Yuqing Jin, Martino Luis, Shuya Zhong, and 18542 Voicu Ion Sucala (University of Exeter) 18543 </div> 18544 <div class="slot-abstract"> 18545 <div> 18546 <a 18547 class="clickable no-decoration" 18548 id="vhsjs_view_408_1707793552_1489408" 18549 onclick="$('#vhsjs_view_408_1707793552_1489408').hide(); 18550 $('#vhsjs_hide_408_1707793552_1489408').show(); 18551 $('#407_1707793552_1489327').slideDown(function() { 18552 if (typeof Masonry === 'function') { 18553 $('.use_masonry').masonry(); 18554 }; 18555 18556 });" 18557 ><i class="fa fa-caret-right"></i>
18558 <span class="hover_link">Abstract</span></a 18559 ><a 18560 class="clickable no-decoration" 18561 id="vhsjs_hide_408_1707793552_1489408" 18562 onclick="$('#407_1707793552_1489327').hide(function() { 18563 if (typeof Masonry === 'function') { 18564 $('.use_masonry').masonry(); 18565 }; 18566 }); 18567 $('#vhsjs_hide_408_1707793552_1489408').hide(); 18568 $('#vhsjs_view_408_1707793552_1489408').show();" 18569 style="display: none" 18570 ><i class="fa fa-caret-down"></i> 18571 <span class="hover_link">Abstract</span></a 18572 > 18573 <div 18574 data-display-control="408_1707793552_1489408" 18575 id="407_1707793552_1489327" 18576 style="display: none" 18577 > 18578 <div class="arrow-slidedown"> 18579 <blockquote> 18580 Canned water is one of the thriving markets in 18581 the food and beverage industry. Given the tight 18582 competition in this market, realistic analysis 18583 in such production lines has become even more 18584 attractive for all participating parties. In 18585 this paper, we apply a KPI-driven 18586 simulation-based approach to a smart production 18587 plant of a key player in the European beverage 18588 market. The project covers realistic 18589 discrete-event modeling and analysis of the 18590 system together with the suggested 18591 scenario-based optimization for performance 18592 improvement. Here, the smart line management 18593 system is modeled and re-coded while considering 18594 machine characteristics, failures, and their 18595 overall influence on the production process. Our 18596 proposed optimized scenario demonstrates 18597 noticeably better results in all performance 18598 indicators when compared to the existing state 18599 of the system. The total increment of the 18600 production speed reaches up to 45 percent, 18601 resource utilization is evenly optimal, and the 18602 overall work-in-progress inventory is reduced 18603 significantly. 18604 </blockquote> 18605 </div> 18606 </div> 18607 </div> 18608 </div> 18609 <div class="slot-urls"></div> 18610 <a href="/wsc23papers/177.pdf" target="_blank">pdf</a 18611 ><br /> 18612 </div> 18613 </div> 18614 <div class="session-entry"> 18615 <span class="session-event-type">Technical Session</span 18616 ><span class="type-track-spacer"> · </span 18617 ><span class="program-track" 18618 >Manufacturing and Industry 4.0</span 18619 ><br /> 18620 <div class="session-title">Assembly Lines II</div> 18621 <div class="session-chair"> 18622 Chair: Ali Ahmad Malik (Oakland University)<br /> 18623 </div> 18624 <div class="slot-entry"> 18625 <a name="con233" tabindex="-1"></a> 18626 <div class="slot-title-line"> 18627 <span class="slot-title" 18628 >Integrating Scheduling of Logistic Support Processes 18629 in Agent-Based Industry 4.0 Assembly Simulation</span 18630 > 18631 </div> 18632 <div class="slot-authors"> 18633 Adrian Freiter (Fraunhofer Institute for Software and 18634 Systems Engineering ISST) and Christian Schwede 18635 (University of Applied Sciences and Arts Bielefeld) 18636 </div> 18637 <div class="slot-abstract"> 18638 <div> 18639 <a 18640 class="clickable no-decoration" 18641 id="vhsjs_view_410_1707793552_1540573" 18642 onclick="$('#vhsjs_view_410_1707793552_1540573').hide(); 18643 $('#vhsjs_hide_410_1707793552_1540573').show(); 18644 $('#409_1707793552_154049').slideDown(function() { 18645 if (typeof Masonry === 'function') { 18646 $('.use_masonry').masonry(); 18647 }; 18648 18649 });" 18650 ><i class="fa fa-caret-right"></i>
18651 <span class="hover_link">Abstract</span></a 18652 ><a 18653 class="clickable no-decoration" 18654 id="vhsjs_hide_410_1707793552_1540573" 18655 onclick="$('#409_1707793552_154049').hide(function() { 18656 if (typeof Masonry === 'function') { 18657 $('.use_masonry').masonry(); 18658 }; 18659 }); 18660 $('#vhsjs_hide_410_1707793552_1540573').hide(); 18661 $('#vhsjs_view_410_1707793552_1540573').show();" 18662 style="display: none" 18663 ><i class="fa fa-caret-down"></i> 18664 <span class="hover_link">Abstract</span></a 18665 > 18666 <div 18667 data-display-control="410_1707793552_1540573" 18668 id="409_1707793552_154049" 18669 style="display: none" 18670 > 18671 <div class="arrow-slidedown"> 18672 <blockquote> 18673 The upcoming decentralized production systems 18674 seem to be promising in Industry 4.0 assembly to 18675 handle the challenges of highly individual 18676 products. Matrix production characterized by 18677 freely linked workstations and an advanced 18678 automation level are highly flexible. That is 18679 why many efforts have already been made to 18680 explore the advantages compared to existing flow 18681 shop production systems, but also the additional 18682 challenges arising from this new paradigm. One 18683 of these challenges is the synchronization of 18684 main product and supply part flow at the 18685 individual workstations during order scheduling. 18686 This paper presents a new approach of 18687 integrating logistics support processes into the 18688 scheduling of the main product flow to consider 18689 the part supply in the decisions taken during 18690 scheduling avoiding waiting times. We compare 18691 our integrated approach with the existing 18692 decoupled scheduling approach, based on a 18693 “bicycle assembly” scenario. The 18694 results are promising particularly when part 18695 supply is a bottleneck. 18696 </blockquote> 18697 </div> 18698 </div> 18699 </div> 18700 </div> 18701 <div class="slot-urls"></div> 18702 <a href="/wsc23papers/176.pdf" target="_blank">pdf</a 18703 ><br /> 18704 </div> 18705 </div> 18706 </div> 18707 <div class="centered"> 18708 <div class="top-link"><a href="#top">Return to Top</a></div> 18709 </div> 18710 <hr /> 18711 </div> 18712 <div class="area-section"> 18713 <div class="centered"> 18714 <a name="ptrack124" tabindex="-1"></a> 18715 <div class="section-title"> 18716 MASM: Semiconductor Manufacturing 18717 </div> 18718 </div> 18719 <div class="centered track-chair"> 18720 <span class="track-chair-role" 18721 >Track Coordinator - MASM: Semiconductor Manufacturing: </span 18722 ><span class="track-chair-names" 18723 >John Fowler (Arizona State University), Young Jae Jang 18724 (Korea Advanced Institute of Science and Technology, Daim 18725 Research), Lars Moench (University of Hagen)</span 18726 > 18727 </div> 18728 <div class="section-entry"> 18729 <div class="session-entry">
18730 <span class="session-event-type">Technical Session</span 18731 ><span class="type-track-spacer"> · </span 18732 ><span class="program-track" 18733 >MASM: Semiconductor Manufacturing</span 18734 ><br /> 18735 <div class="session-title">Scheduling I</div> 18736 <div class="session-chair"> 18737 Chair: Reha Uzsoy (North Carolina State University)<br /> 18738 </div> 18739 <div class="slot-entry"> 18740 <a name="inv111" tabindex="-1"></a> 18741 <div class="slot-title-line"> 18742 <span class="slot-title" 18743 >A Reinforcement Learning Approach for Improved 18744 Photolithography Schedules</span 18745 > 18746 </div> 18747 <div class="slot-authors"> 18748 Tao Zhang (Universität der Bundeswehr 18749 München), Kamil Erkan Kabak (Izmir University of 18750 Economics), Cathal Heavey (University of Limerick), and 18751 Oliver Rose (Universität der Bundeswehr 18752 München) 18753 </div> 18754 <div class="slot-abstract"> 18755 <div> 18756 <a 18757 class="clickable no-decoration" 18758 id="vhsjs_view_412_1707793552_1609101" 18759 onclick="$('#vhsjs_view_412_1707793552_1609101').hide(); 18760 $('#vhsjs_hide_412_1707793552_1609101').show(); 18761 $('#411_1707793552_1609015').slideDown(function() { 18762 if (typeof Masonry === 'function') { 18763 $('.use_masonry').masonry(); 18764 }; 18765 18766 });" 18767 ><i class="fa fa-caret-right"></i> 18768 <span class="hover_link">Abstract</span></a 18769 ><a 18770 class="clickable no-decoration" 18771 id="vhsjs_hide_412_1707793552_1609101" 18772 onclick="$('#411_1707793552_1609015').hide(function() { 18773 if (typeof Masonry === 'function') { 18774 $('.use_masonry').masonry(); 18775 }; 18776 }); 18777 $('#vhsjs_hide_412_1707793552_1609101').hide(); 18778 $('#vhsjs_view_412_1707793552_1609101').show();" 18779 style="display: none" 18780 ><i class="fa fa-caret-down"></i> 18781 <span class="hover_link">Abstract</span></a 18782 > 18783 <div 18784 data-display-control="412_1707793552_1609101" 18785 id="411_1707793552_1609015" 18786 style="display: none" 18787 > 18788 <div class="arrow-slidedown"> 18789 <blockquote> 18790 A Reinforcement Learning (RL) model is applied 18791 for photolithography schedules with direct 18792 consideration of reentrant visits. The 18793 photolithography process is mainly regarded as a 18794 bottleneck process in semiconductor 18795 manufacturing, and improving its schedules would 18796 result in better performances. Most RL-based 18797 research do not consider revisits directly or 18798 guarantee convergence. A simplified discrete 18799 event simulation model of a fabrication facility 18800 is built, and a tabular Q-learning agent is 18801 embedded into the model to learn through 18802 scheduling. The learning environment c
18802onsiders 18803 states and actions consisting of information on 18804 reentrant flows. The agent dynamically chooses 18805 one rule from a pre-defined rule set to dispatch 18806 lots. The set includes the earliest stage first, 18807 the latest stage first, and 8 more composite 18808 rules. Finally, the proposed RL approach is 18809 compared with 7 single and 8 hybrid rules. The 18810 method presents a validated approach in terms of 18811 overall average cycle times. 18812 </blockquote> 18813 </div> 18814 </div> 18815 </div> 18816 </div> 18817 <div class="slot-urls"></div> 18818 <a href="/wsc23papers/178.pdf" target="_blank">pdf</a 18819 ><br /> 18820 </div> 18821 <div class="slot-entry"> 18822 <a name="cea137" tabindex="-1"></a> 18823 <div class="slot-title-line"> 18824 <span class="slot-title" 18825 >Deploying an Advanced AI Diffusion Scheduler at a 18826 Renesas Fab</span 18827 > 18828 </div> 18829 <div class="slot-authors"> 18830 James Adamson and Lio Weinstock (Flexciton Ltd), Jay 18831 Maguire (Renesas), Lara Nichols (FabTime), and Dionysios 18832 Xenos (Flexciton Ltd) 18833 </div> 18834 <div class="slot-abstract"> 18835 <div> 18836 <a 18837 class="clickable no-decoration" 18838 id="vhsjs_view_414_1707793552_1631408" 18839 onclick="$('#vhsjs_view_414_1707793552_1631408').hide(); 18840 $('#vhsjs_hide_414_1707793552_1631408').show(); 18841 $('#413_1707793552_1631327').slideDown(function() { 18842 if (typeof Masonry === 'function') { 18843 $('.use_masonry').masonry(); 18844 }; 18845 18846 });" 18847 ><i class="fa fa-caret-right"></i> 18848 <span class="hover_link">Abstract</span></a 18849 ><a 18850 class="clickable no-decoration" 18851 id="vhsjs_hide_414_1707793552_1631408" 18852 onclick="$('#413_1707793552_1631327').hide(function() { 18853 if (typeof Masonry === 'function') { 18854 $('.use_masonry').masonry(); 18855 }; 18856 }); 18857 $('#vhsjs_hide_414_1707793552_1631408').hide(); 18858 $('#vhsjs_view_414_1707793552_1631408').show();" 18859 style="display: none" 18860 ><i class="fa fa-caret-down"></i> 18861 <span class="hover_link">Abstract</span></a 18862 > 18863 <div 18864 data-display-control="414_1707793552_1631408" 18865 id="413_1707793552_1631327" 18866 style="display: none" 18867 > 18868 <div class="arrow-slidedown"> 18869 <blockquote> 18870 Scheduling the diffusion area in a front-end 18871 wafer fab poses challenges. This industrial
18871case 18872 focuses on scheduling diffusion at 18873 Renesas’ Palm Bay Fab, which is always 18874 seeking scheduling system improvements. 18875 Transitioning to an advanced system, considering 18876 fab-wide impacts on diffusion batching, enhances 18877 Key Performance Indicators (KPIs). Our A.I. 18878 scheduler utilizes optimization, heuristics, and 18879 live data updates every five minutes. 18880 Collaborating with FabTime integrates the 18881 scheduler with the fab’s MES, ensuring 18882 frequent updates. It optimizes batching, tool 18883 allocation, and launch times, aligning with 18884 Renesas’ objective to balance competing 18885 goals. Initial results show a 36% and 13% 18886 increase in diffusion batch sizes at clean and 18887 expensive furnace toolsets. The minor impact on 18888 cycle time reflects the scheduler’s focus 18889 on batching efficiency. This approach improves 18890 efficiency and meets Renesas’ goals, 18891 marking a positive step in optimizing their 18892 wafer fab operations. 18893 </blockquote> 18894 </div> 18895 </div> 18896 </div> 18897 </div> 18898 <div class="slot-urls"></div> 18899 <a href="/wsc23papers/cea137.pdf" target="_blank">pdf</a 18900 ><br /> 18901 </div> 18902 <div class="slot-entry"> 18903 <a name="inv188" tabindex="-1"></a> 18904 <div class="slot-title-line"> 18905 <span class="slot-title" 18906 >Deep Learning Enabling Digital Twin Applications in 18907 Production Scheduling: Case of Flexible Job Shop 18908 Manufacturing Environment</span 18909 > 18910 </div> 18911 <div class="slot-authors"> 18912 Amir Ghasemi (Amsterdam University of Applied Sciences, 18913 Amsterdam School of International Business); Yavar 18914 Taheri Yeganeh and Andrea Matta (Politecnico di Milano); 18915 Kamil Erkan Kabak (Izmir University of Economics); and 18916 Cathal Heavey (University of Limerick) 18917 </div> 18918 <div class="slot-abstract"> 18919 <div> 18920 <a 18921 class="clickable no-decoration" 18922 id="vhsjs_view_416_1707793552_1655293" 18923 onclick="$('#vhsjs_view_416_1707793552_1655293').hide(); 18924 $('#vhsjs_hide_416_1707793552_1655293').show(); 18925 $('#415_1707793552_1655214').slideDown(function() { 18926 if (typeof Masonry === 'function') { 18927 $('.use_masonry').masonry(); 18928 }; 18929 18930 });" 18931 ><i class="fa fa-caret-right"></i> 18932 <span class="hover_link">Abstract</span></a 18933 ><a 18934 class="clickable no-decoration" 18935 id="vhsjs_hide_416_1707793552_1655293" 18936 onclick="$('#415_1707793552_1655214').hide(function() { 18937 if (typeof Masonry === 'function') { 18938 $('.use_masonry').masonry(); 18939 }; 18940 }); 18941 $('#vhsjs_hide_416_1707793552_1655293').hide(); 18942 $('#vhsjs_view_416_1707793552_1655293').show();" 18943 style="display: none" 18944 ><i class="fa fa-caret-down"></i>
18945 <span class="hover_link">Abstract</span></a 18946 > 18947 <div 18948 data-display-control="416_1707793552_1655293" 18949 id="415_1707793552_1655214" 18950 style="display: none" 18951 > 18952 <div class="arrow-slidedown"> 18953 <blockquote> 18954 Digital twin-based Production Scheduling (DTPS) 18955 is a process in which a digital model replicates 18956 a manufacturing system, known as a 18957 “Digital Twin (DT)”. DT is 18958 essentially a virtual representation of physical 18959 equipment and processes that are connected to 18960 the physical environment using an online 18961 data-sharing infrastructure within the 18962 Manufacturing Execution System (MES). In the 18963 case of reactive scheduling, DT is used to 18964 detect fluctuations in the scheduling plan and 18965 execute rescheduling plans. In proactive 18966 scheduling, it is used to simulate different 18967 production scenarios and optimize future states 18968 of production operations. Replicating detailed 18969 simulation models in most PS cases is highly 18970 computationally intensive, which negates against 18971 the main goal of DT (online decision making). 18972 Thus, this research aims to examine the 18973 possibility of using data-driven models within 18974 the DT of a Flexible Job Shop (FJS) production 18975 environment aiming to provide online estimations 18976 of PS metrics enabling DT-based 18977 reactive/proactive scheduling. 18978 </blockquote> 18979 </div> 18980 </div> 18981 </div> 18982 </div> 18983 <div class="slot-urls"></div> 18984 <a href="/wsc23papers/179.pdf" target="_blank">pdf</a 18985 ><br /> 18986 </div> 18987 </div> 18988 <div class="session-entry"> 18989 <span class="session-event-type">Technical Session</span 18990 ><span class="type-track-spacer"> · </span 18991 ><span class="program-track" 18992 >MASM: Semiconductor Manufacturing</span 18993 ><br /> 18994 <div class="session-title">Time Issues in Wafer Fabs</div> 18995 <div class="session-chair"> 18996 Chair: Young Jae Jang (KAIST)<br /> 18997 </div> 18998 <div class="slot-entry"> 18999 <a name="con108" tabindex="-1"></a> 19000 <div class="slot-title-line"> 19001 <span class="slot-title" 19002 >Optimization of Timelinks in Semiconductor 19003 Manufacturing</span 19004 > 19005 </div> 19006 <div class="slot-authors"> 19007 Nina Dybowski, Maria Sander, and Ralf Sprenger (Infineon 19008 Technologies Dresden GmbH) 19009 </div> 19010 <div class="slot-abstract"> 19011 <div> 19012 <a 19013 class="clickable no-decoration" 19014 id="vhsjs_view_418_1707793552_170387" 19015 onclick="$('#vhsjs_view_418_1707793552_170387').hide(); 19016 $('#vhsjs_hide_418_1707793552_170387').show(); 19017 $('#417_1707793552_1703784').slideDown(function() { 19018 if (typeof Masonry === 'function') { 19019 $('.use_masonry').masonry(); 19020 }; 19021 19022 });" 19023 ><i class="fa fa-caret-right"></i>
19024 <span class="hover_link">Abstract</span></a 19025 ><a 19026 class="clickable no-decoration" 19027 id="vhsjs_hide_418_1707793552_170387" 19028 onclick="$('#417_1707793552_1703784').hide(function() { 19029 if (typeof Masonry === 'function') { 19030 $('.use_masonry').masonry(); 19031 }; 19032 }); 19033 $('#vhsjs_hide_418_1707793552_170387').hide(); 19034 $('#vhsjs_view_418_1707793552_170387').show();" 19035 style="display: none" 19036 ><i class="fa fa-caret-down"></i> 19037 <span class="hover_link">Abstract</span></a 19038 > 19039 <div 19040 data-display-control="418_1707793552_170387" 19041 id="417_1707793552_1703784" 19042 style="display: none" 19043 > 19044 <div class="arrow-slidedown"> 19045 <blockquote> 19046 Impact of timelinks to semiconductor 19047 manufacturing has risen due to shrinking 19048 technology sizes. Their operational control 19049 defines on the one hand how good the time 19050 restrictions are met and on the other the impact 19051 to fab capacity. This paper discusses both 19052 aspects and the influencing factors like uptime 19053 stability, length of the timelink etc. A control 19054 approach is proposed, evaluated, and discussed. 19055 Furthermore, a monitoring system is introduced 19056 that enables for fast decision making and 19057 optimization of the control parameters. Finally, 19058 a simulation study is done for evaluating 19059 different parameters and impact of influencing 19060 factors. 19061 </blockquote> 19062 </div> 19063 </div> 19064 </div> 19065 </div> 19066 <div class="slot-urls"></div> 19067 <a href="/wsc23papers/180.pdf" target="_blank">pdf</a 19068 ><br /> 19069 </div> 19070 <div class="slot-entry"> 19071 <a name="con215" tabindex="-1"></a> 19072 <div class="slot-title-line"> 19073 <span class="slot-title" 19074 >Queue Time Prediction Methodology in Semiconductor 19075 Fab</span 19076 > 19077 </div> 19078 <div class="slot-authors"> 19079 Donguk Kim, Byeongseon Lee, and Sangchul Park (Ajou 19080 University) 19081 </div> 19082 <div class="slot-abstract"> 19083 <div> 19084 <a 19085 class="clickable no-decoration" 19086 id="vhsjs_view_420_1707793552_1725938" 19087 onclick="$('#vhsjs_view_420_1707793552_1725938').hide(); 19088 $('#vhsjs_hide_420_1707793552_1725938').show(); 19089 $('#419_1707793552_1725857').slideDown(function() { 19090 if (typeof Masonry === 'function') { 19091 $('.use_masonry').masonry(); 19092 }; 19093 19094 });" 19095 ><i class="fa fa-caret-right"></i> 19096 <span class="hover_link">Abstract</span></a 19097 ><a 19098 class="clickable no-decoration" 19099 id="vhsjs_hide_420_1707793552_1725938" 19100 onclick="$('#419_1707793552_1725857').hide(function() { 19101 if (typeof Masonry === 'function') { 19102 $('.use_masonry').masonry(); 19103 }; 19104 }); 19105 $('#vhsjs_hide_420_1707793552_1725938').hide(); 19106 $('#vhsjs_view_420_1707793552_1725938').show();" 19107 style="display: none" 19108 ><i class="fa fa-caret-down"></i>
19109 <span class="hover_link">Abstract</span></a 19110 > 19111 <div 19112 data-display-control="420_1707793552_1725938" 19113 id="419_1707793552_1725857" 19114 style="display: none" 19115 > 19116 <div class="arrow-slidedown"> 19117 <blockquote> 19118 This paper presents a methodology for predicting 19119 queue times in semiconductor fabrication, where 19120 numerous complex and costly pieces of equipment 19121 are utilized. Queue time, occurring between 19122 continuous single or multi-processes, is a 19123 crucial factor affecting the quality of wafers, 19124 which can significantly impact costs. While most 19125 semiconductor fabrications use queue time limits 19126 as a key dispatching factor, some wafers may 19127 still be scrapped or reworked. By predicting 19128 queue times, we can reduce unnecessary waste by 19129 blocking or re-dispatching wafers. Two 19130 approximations are proposed and compared based 19131 on accuracy and prediction time: a machine 19132 learning model trained using experimental 19133 results and a multi-resolution simulation model 19134 with varying fidelity levels. The simulation 19135 model is validated using the SMAT2022 data set. 19136 </blockquote> 19137 </div> 19138 </div> 19139 </div> 19140 </div> 19141 <div class="slot-urls"></div> 19142 <a href="/wsc23papers/181.pdf" target="_blank">pdf</a 19143 ><br /> 19144 </div> 19145 <div class="slot-entry"> 19146 <a name="cea133" tabindex="-1"></a> 19147 <div class="slot-title-line"> 19148 <span class="slot-title" 19149 >Processing Time and Machine Availability Prediction 19150 in Semiconductor Manufacturing Using Neural 19151 Networks</span 19152 > 19153 </div> 19154 <div class="slot-authors"> 19155 Taki Eddine Korabi, Gerard Goossen, Abhinav Kaushik, 19156 Tijmen Tieleman, Jasper Van Heugten, and Jeroen 19157 Bédorf (Minds.ai) and Shiladitya Chakravorty, 19158 Detlef Pabst, and John Thomas (Globalfoundries) 19159 </div> 19160 <div class="slot-abstract"> 19161 <div> 19162 <a 19163 class="clickable no-decoration" 19164 id="vhsjs_view_422_1707793552_1750288" 19165 onclick="$('#vhsjs_view_422_1707793552_1750288').hide(); 19166 $('#vhsjs_hide_422_1707793552_1750288').show(); 19167 $('#421_1707793552_1750207').slideDown(function() { 19168 if (typeof Masonry === 'function') { 19169 $('.use_masonry').masonry(); 19170 }; 19171 19172 });" 19173 ><i class="fa fa-caret-right"></i> 19174 <span class="hover_link">Abstract</span></a 19175 ><a 19176 class="clickable no-decoration" 19177 id="vhsjs_hide_422_1707793552_1750288" 19178 onclick="$('#421_1707793552_1750207').hide(function() { 19179 if (typeof Masonry === 'function') { 19180 $('.use_masonry').masonry(); 19181 }; 19182 }); 19183 $('#vhsjs_hide_422_1707793552_1750288').hide(); 19184 $('#vhsjs_view_422_1707793552_1750288').show();" 19185 style="display: none" 19186 ><i class="fa fa-caret-down"></i>
19187 <span class="hover_link">Abstract</span></a 19188 > 19189 <div 19190 data-display-control="422_1707793552_1750288" 19191 id="421_1707793552_1750207" 19192 style="display: none" 19193 > 19194 <div class="arrow-slidedown"> 19195 <blockquote> 19196 In partnership with GlobalFoundries we have 19197 significantly advanced Processing Time (PT) and 19198 machine availability prediction in fabrication 19199 plants, utilizing an attention based neural 19200 network. This model is integrated into an MLOps 19201 pipeline consisting of data collection, 19202 preprocessing, training and deployment. The data 19203 is augmented with features such as chamber usage 19204 and process sequences. Compared to the current 19205 model, which calculates average processing times 19206 over a predefined context, our approach has 19207 reduced the Mean Absolute Error (MAE) of PT 19208 predictions by 43% to 80% across the crucial 19209 areas: Etch, Diffusion, and Deposition. The 19210 model also produces high quality predictions for 19211 the remaining tools. The model is in the process 19212 of being implemented in the FAB to improve 19213 scheduling, dispatching, and improve crucial Key 19214 Performance Indicators (KPIs) such as cycle time 19215 and throughput. 19216 </blockquote> 19217 </div> 19218 </div> 19219 </div> 19220 </div> 19221 <div class="slot-urls"></div> 19222 <a href="/wsc23papers/cea133.pdf" target="_blank">pdf</a 19223 ><br /> 19224 </div> 19225 </div> 19226 <div class="session-entry"> 19227 <span class="session-event-type">Technical Session</span 19228 ><span class="type-track-spacer"> · </span 19229 ><span class="program-track" 19230 >MASM: Semiconductor Manufacturing</span 19231 ><br /> 19232 <div class="session-title">Supply Chain Management I</div> 19233 <div class="session-chair"> 19234 Chair: Douniel Lamghari-Idrissi (ASML, Eindhoven 19235 University of Technology)<br /> 19236 </div> 19237 <div class="slot-entry"> 19238 <a name="cea138" tabindex="-1"></a> 19239 <div class="slot-title-line"> 19240 <span class="slot-title" 19241 >Data-driven Warehouse Planning and Control under 19242 Stochastic Demand and Labor Supply in Semi-conductor 19243 Capital Equipment Manufacturing</span 19244 > 19245 </div> 19246 <div class="slot-authors"> 19247 Douglas Morrice, Yanyue (Lilian) Ding, and Jonathan Bard 19248 (The University of Texas at Austin) 19249 </div> 19250 <div class="slot-abstract"> 19251 <div> 19252 <a 19253 class="clickable no-decoration" 19254 id="vhsjs_view_424_1707793552_1793797" 19255 onclick="$('#vhsjs_view_424_1707793552_1793797').hide(); 19256 $('#vhsjs_hide_424_1707793552_1793797').show(); 19257 $('#423_1707793552_1793716').slideDown(function() { 19258 if (typeof Masonry === 'function') { 19259 $('.use_masonry').masonry(); 19260 }; 19261 19262 });" 19263 ><i class="fa fa-caret-right"></i>
19264 <span class="hover_link">Abstract</span></a 19265 ><a 19266 class="clickable no-decoration" 19267 id="vhsjs_hide_424_1707793552_1793797" 19268 onclick="$('#423_1707793552_1793716').hide(function() { 19269 if (typeof Masonry === 'function') { 19270 $('.use_masonry').masonry(); 19271 }; 19272 }); 19273 $('#vhsjs_hide_424_1707793552_1793797').hide(); 19274 $('#vhsjs_view_424_1707793552_1793797').show();" 19275 style="display: none" 19276 ><i class="fa fa-caret-down"></i> 19277 <span class="hover_link">Abstract</span></a 19278 > 19279 <div 19280 data-display-control="424_1707793552_1793797" 19281 id="423_1707793552_1793716" 19282 style="display: none" 19283 > 19284 <div class="arrow-slidedown"> 19285 <blockquote> 19286 Access to more information and sophisticated 19287 analytics enables warehouse management to make 19288 better data-driven decisions. In our study, we 19289 develop a simulation-regression metamodel to 19290 help warehouse managers plan workforce, space, 19291 and equipment requirements for a leading 19292 semiconductor capital equipment company. More 19293 specifically, we use historical inbound and 19294 outbound demand records and performance metrics 19295 (such as workers’ hourly productivity and 19296 moving rates) to predict the space, workforce, 19297 and equipment required for different operation 19298 stages in the warehouse facility. We implement 19299 the simulation model in Python. Simulation 19300 experiments provide insights on resource 19301 planning under different demand scenarios and 19302 supply constraints. 19303 </blockquote> 19304 </div> 19305 </div> 19306 </div> 19307 </div> 19308 <div class="slot-urls"></div> 19309 <a href="/wsc23papers/cea138.pdf" target="_blank">pdf</a 19310 ><br /> 19311 </div> 19312 <div class="slot-entry"> 19313 <a name="con224" tabindex="-1"></a> 19314 <div class="slot-title-line"> 19315 <span class="slot-title" 19316 >Assessing Delivery Commitments in Supply Chains: A 19317 Matrix-Based Framework</span 19318 > 19319 </div> 19320 <div class="slot-authors"> 19321 Madhurima Vangeepuram (Hochschule Neu-Ulm), Hans Ehm and 19322 Marco Ratusny (Infineon Technologies AG), Stefan 19323 Faußer (Hochschule Neu-Ulm), and Stefan Heilmayer 19324 and Tobias Leander Welling (Infineon Technologies AG) 19325 </div> 19326 <div class="slot-abstract"> 19327 <div> 19328 <a 19329 class="clickable no-decoration" 19330 id="vhsjs_view_426_1707793552_1818955" 19331 onclick="$('#vhsjs_view_426_1707793552_1818955').hide(); 19332 $('#vhsjs_hide_426_1707793552_1818955').show(); 19333 $('#425_1707793552_1818871').slideDown(function() { 19334 if (typeof Masonry === 'function') { 19335 $('.use_masonry').masonry(); 19336 }; 19337 19338 });" 19339 ><i class="fa fa-caret-right"></i>
19340 <span class="hover_link">Abstract</span></a 19341 ><a 19342 class="clickable no-decoration" 19343 id="vhsjs_hide_426_1707793552_1818955" 19344 onclick="$('#425_1707793552_1818871').hide(function() { 19345 if (typeof Masonry === 'function') { 19346 $('.use_masonry').masonry(); 19347 }; 19348 }); 19349 $('#vhsjs_hide_426_1707793552_1818955').hide(); 19350 $('#vhsjs_view_426_1707793552_1818955').show();" 19351 style="display: none" 19352 ><i class="fa fa-caret-down"></i> 19353 <span class="hover_link">Abstract</span></a 19354 > 19355 <div 19356 data-display-control="426_1707793552_1818955" 19357 id="425_1707793552_1818871" 19358 style="display: none" 19359 > 19360 <div class="arrow-slidedown"> 19361 <blockquote> 19362 Ensuring reliable and timely customer deliveries 19363 is crucial to supply chain management. The 19364 ability to meet delivery commitments is 19365 essential for maintaining customer satisfaction. 19366 Despite the importance of delivery commitments, 19367 there is a lack of standard measurement 19368 techniques for evaluating their quality. 19369 Therefore, this paper introduces the term 19370 Commitment Quality (CQ) and develops a CQ matrix 19371 that can be used to measure the quality of 19372 delivery commitments. The CQ matrix provides a 19373 comprehensive set of quantitative measures to 19374 evaluate different aspects of delivery 19375 commitments. Finally, a numerical example based 19376 on an order data sample from a semiconductor 19377 manufacturer is presented and discussed. The 19378 proposed framework aims to standardize the CQ, 19379 enhancing transparency in delivery commitments. 19380 </blockquote> 19381 </div> 19382 </div> 19383 </div> 19384 </div> 19385 <div class="slot-urls"></div> 19386 <a href="/wsc23papers/182.pdf" target="_blank">pdf</a 19387 ><br /> 19388 </div> 19389 <div class="slot-entry"> 19390 <a name="con132" tabindex="-1"></a> 19391 <div class="slot-title-line"> 19392 <span class="slot-title" 19393 >The Bullwhip Effect in End-to-end Supply Chains: The 19394 Impact of Reach-based Replenishment Policies with a 19395 Long Cycle Time Supplier</span 19396 > 19397 </div> 19398 <div class="slot-authors"> 19399 Hans Ehm, Chun Hei Chung, Sanchari Kar Chowdhury, Marco 19400 Ratusny, and Abdelgafar Ismail (Infineon Technologies 19401 AG) 19402 </div> 19403 <div class="slot-abstract"> 19404 <div> 19405 <a 19406 class="clickable no-decoration" 19407 id="vhsjs_view_428_1707793552_1842945" 19408 onclick="$('#vhsjs_view_428_1707793552_1842945').hide(); 19409 $('#vhsjs_hide_428_1707793552_1842945').show(); 19410 $('#427_1707793552_1842859').slideDown(function() { 19411 if (typeof Masonry === 'function') { 19412 $('.use_masonry').masonry(); 19413 }; 19414 19415 });" 19416 ><i class="fa fa-caret-right"></i>
19417 <span class="hover_link">Abstract</span></a 19418 ><a 19419 class="clickable no-decoration" 19420 id="vhsjs_hide_428_1707793552_1842945" 19421 onclick="$('#427_1707793552_1842859').hide(function() { 19422 if (typeof Masonry === 'function') { 19423 $('.use_masonry').masonry(); 19424 }; 19425 }); 19426 $('#vhsjs_hide_428_1707793552_1842945').hide(); 19427 $('#vhsjs_view_428_1707793552_1842945').show();" 19428 style="display: none" 19429 ><i class="fa fa-caret-down"></i> 19430 <span class="hover_link">Abstract</span></a 19431 > 19432 <div 19433 data-display-control="428_1707793552_1842945" 19434 id="427_1707793552_1842859" 19435 style="display: none" 19436 > 19437 <div class="arrow-slidedown"> 19438 <blockquote> 19439 The bullwhip effect (BWE), a well-known 19440 phenomenon in supply chain management since it 19441 was first identified in 1958, is causing 19442 significant economic damage after disruptions. 19443 While the role of human factors in BWE has been 19444 widely recognized, the impact of different 19445 replenishment policies on BWE mitigation has not 19446 been thoroughly investigated. This paper 19447 presents a study on the impact of reach-based 19448 Kanban systems on the BWE in supply chains 19449 containing suppliers with intrinsically 19450 non-reducible long cycle times, such as those in 19451 the semiconductor industry. Our findings suggest 19452 that a reach-based replenishment system acts as 19453 a BWE accelerator after significant disruptions, 19454 which can end up in line-downs downstream. We 19455 propose a change to absolute stock targets for 19456 replenishment policies during disruption to 19457 mitigate this aspect of the BWE root cause for 19458 supply chain with long cycle time suppliers to 19459 reduce the risk of line downs. 19460 </blockquote> 19461 </div> 19462 </div> 19463 </div> 19464 </div> 19465 <div class="slot-urls"></div> 19466 <a href="/wsc23papers/183.pdf" target="_blank">pdf</a 19467 ><br /> 19468 </div> 19469 </div> 19470 <div class="session-entry"> 19471 <span class="session-event-type">Technical Session</span 19472 ><span class="type-track-spacer"> · </span 19473 ><span class="program-track" 19474 >MASM: Semiconductor Manufacturing</span 19475 ><br /> 19476 <div class="session-title">Planning</div> 19477 <div class="session-chair"> 19478 Chair: Tobias Voelker (University of Hagen)<br /> 19479 </div> 19480 <div class="slot-entry"> 19481 <a name="inv142" tabindex="-1"></a> 19482 <div class="slot-title-line"> 19483 <span class="slot-title" 19484 >Decentralized Decision-making Framework for Managing 19485 Product Rollovers in the Semiconductor 19486 Manufacturing</span 19487 > 19488 </div> 19489 <div class="slot-authors"> 19490 Carlos Leca (North Carolina State University), Karl 19491 Kempf (Intel Corporation), and Reha Uzsoy (North 19492 Carolina State University) 19493 </div> 19494 <div class="slot-abstract"> 19495 <div> 19496 <a 19497 class="clickable no-decoration" 19498 id="vhsjs_view_430_1707793552_189684" 19499 onclick="$('#vhsjs_view_430_1707793552_189684').hide(); 19500 $('#vhsjs_hide_430_1707793552_189684').show(); 19501 $('#429_1707793552_1896758').slideDown(function() { 19502 if (typeof Masonry === 'function') { 19503 $('.use_masonry').masonry(); 19504 }; 19505 19506 });" 19507 ><i class="fa fa-caret-right"></i>
19508 <span class="hover_link">Abstract</span></a 19509 ><a 19510 class="clickable no-decoration" 19511 id="vhsjs_hide_430_1707793552_189684" 19512 onclick="$('#429_1707793552_1896758').hide(function() { 19513 if (typeof Masonry === 'function') { 19514 $('.use_masonry').masonry(); 19515 }; 19516 }); 19517 $('#vhsjs_hide_430_1707793552_189684').hide(); 19518 $('#vhsjs_view_430_1707793552_189684').show();" 19519 style="display: none" 19520 ><i class="fa fa-caret-down"></i> 19521 <span class="hover_link">Abstract</span></a 19522 > 19523 <div 19524 data-display-control="430_1707793552_189684" 19525 id="429_1707793552_1896758" 19526 style="display: none" 19527 > 19528 <div class="arrow-slidedown"> 19529 <blockquote> 19530 Competitiveness in the semiconductor industry 19531 requires continuous management of product 19532 rollovers, the process of introducing new 19533 products and retiring older ones to maintain 19534 market share. This paper presents a 19535 decentralized decision-making framework to 19536 coordinate product rollover decisions using 19537 Lagrangian decomposition of a centralized model 19538 using quadratic coordination errors in the 19539 subproblem objectives, and a decentralized 19540 heuristic that recovers the feasible solutions 19541 from the relaxed ones obtained from the 19542 Lagrangian procedure. Experimental results show 19543 that this decentralized framework delivers 19544 promising results, obtaining near-optimal 19545 solutions in modest CPU times. 19546 </blockquote> 19547 </div> 19548 </div> 19549 </div> 19550 </div> 19551 <div class="slot-urls"></div> 19552 <a href="/wsc23papers/184.pdf" target="_blank">pdf</a 19553 ><br /> 19554 </div> 19555 <div class="slot-entry"> 19556 <a name="inv160" tabindex="-1"></a> 19557 <div class="slot-title-line"> 19558 <span class="slot-title" 19559 >Data-driven Production Planning Formulations with 19560 Inventory Considerations</span 19561 > 19562 </div> 19563 <div class="slot-authors"> 19564 Tobias Voelker and Lars Moench (University of Hagen) 19565 </div> 19566 <div class="slot-abstract"> 19567 <div> 19568 <a 19569 class="clickable no-decoration" 19570 id="vhsjs_view_432_1707793552_1918068" 19571 onclick="$('#vhsjs_view_432_1707793552_1918068').hide(); 19572 $('#vhsjs_hide_432_1707793552_1918068').show(); 19573 $('#431_1707793552_1917984').slideDown(function() { 19574 if (typeof Masonry === 'function') { 19575 $('.use_masonry').masonry(); 19576 }; 19577 19578 });" 19579 ><i class="fa fa-caret-right"></i> 19580 <span class="hover_link">Abstract</span></a 19581 ><a 19582 class="clickable no-decoration" 19583 id="vhsjs_hide_432_1707793552_1918068" 19584 onclick="$('#431_1707793552_1917984').hide(function() { 19585 if (typeof Masonry === 'function') { 19586 $('.use_masonry').masonry(); 19587 }; 19588 }); 19589 $('#vhsjs_hide_432_1707793552_1918068').hide(); 19590 $('#vhsjs_view_432_1707793552_1918068').show();" 19591 style="display: none" 19592 ><i class="fa fa-caret-down"></i>
19593 <span class="hover_link">Abstract</span></a 19594 > 19595 <div 19596 data-display-control="432_1707793552_1918068" 19597 id="431_1707793552_1917984" 19598 style="display: none" 19599 > 19600 <div class="arrow-slidedown"> 19601 <blockquote> 19602 Data-driven (DD) production planning 19603 formulations for semiconductor wafer fabrication 19604 facilities (wafer fabs) are studied in this 19605 paper. These formulations are based on a set of 19606 system states representing the congestion 19607 behavior of the wafer fab with work in process 19608 and resulting output levels. We establish two DD 19609 formulations with inventory considerations. The 19610 first variant is a shortfall-based 19611 chance-constrained formulation that considers 19612 safety stocks at the finished goods inventory 19613 level. The second variant is a simple 19614 scenario-based stochastic program where the 19615 objective function reflects the expected 19616 inventory holding and backlog cost under 19617 uncertainty. The two variants are compared with 19618 the conventional DD formulation in a rolling 19619 horizon environment using a simulation model of 19620 a large-scaled wafer fab. The simulation 19621 experiments demonstrate that the stochastic 19622 program achieves the largest profit under all 19623 experimental conditions. 19624 </blockquote> 19625 </div> 19626 </div> 19627 </div> 19628 </div> 19629 <div class="slot-urls"></div> 19630 <a href="/wsc23papers/185.pdf" target="_blank">pdf</a 19631 ><br /> 19632 </div> 19633 <div class="slot-entry"> 19634 <a name="inv159" tabindex="-1"></a> 19635 <div class="slot-title-line"> 19636 <span class="slot-title" 19637 >Agent-based Decision Support in Borderless Fab 19638 Scenarios in Semiconductor Manufacturing</span 19639 > 19640 </div> 19641 <div class="slot-authors"> 19642 Raphael Herding (Forschungsinstitut für 19643 Telekommunikation und Kooperation, Westfälische 19644 Hochschule) and Lars Moench (Forschungsinstitut für 19645 Telekommunikation und Kooperation, University of Hagen) 19646 </div> 19647 <div class="slot-abstract"> 19648 <div> 19649 <a 19650 class="clickable no-decoration" 19651 id="vhsjs_view_434_1707793552_1940637" 19652 onclick="$('#vhsjs_view_434_1707793552_1940637').hide(); 19653 $('#vhsjs_hide_434_1707793552_1940637').show(); 19654 $('#433_1707793552_1940553').slideDown(function() { 19655 if (typeof Masonry === 'function') { 19656 $('.use_masonry').masonry(); 19657 }; 19658 19659 });" 19660 ><i class="fa fa-caret-right"></i> 19661 <span class="hover_link">Abstract</span></a 19662 ><a 19663 class="clickable no-decoration" 19664 id="vhsjs_hide_434_1707793552_1940637" 19665 onclick="$('#433_1707793552_1940553').hide(function() { 19666 if (typeof Masonry === 'function') { 19667 $('.use_masonry').masonry(); 19668 }; 19669 }); 19670 $('#vhsjs_hide_434_1707793552_1940637').hide(); 19671 $('#vhsjs_view_434_1707793552_1940637').show();" 19672 style="display: none" 19673 ><i class="fa fa-caret-down"></i>
19674 <span class="hover_link">Abstract</span></a 19675 > 19676 <div 19677 data-display-control="434_1707793552_1940637" 19678 id="433_1707793552_1940553" 19679 style="display: none" 19680 > 19681 <div class="arrow-slidedown"> 19682 <blockquote> 19683 The design and the implementation of a 19684 multi-agent system (MAS) for a borderless fab 19685 scenario is described. In such a scenario, lots 19686 are transferred from one wafer fab to a nearby 19687 one to perform process steps of the transferred 19688 lots. Production planning is carried out 19689 individually for each of the wafer fabs. The 19690 modeling of the available and requested capacity 19691 in the production planning models of the 19692 participating wafer fabs is affected by the lot 19693 transfer. The transfer of the route information 19694 from one wafer fab to another to automatically 19695 generate the linear programming models is 19696 described. Production planning is carried out in 19697 a rolling horizon setting using a cloud-based 19698 infrastructure. We show by simulation 19699 experiments with the MAS with a correct modeling 19700 of the capacity in production planning results 19701 in improved profit compared to a setting where 19702 the lot transfer is not taken into account in 19703 the planning formulations. 19704 </blockquote> 19705 </div> 19706 </div> 19707 </div> 19708 </div> 19709 <div class="slot-urls"></div> 19710 <a href="/wsc23papers/186.pdf" target="_blank">pdf</a 19711 ><br /> 19712 </div> 19713 </div> 19714 <div class="session-entry"> 19715 <span class="session-event-type">Technical Session</span 19716 ><span class="type-track-spacer"> · </span 19717 ><span class="program-track" 19718 >MASM: Semiconductor Manufacturing</span 19719 ><br /> 19720 <div class="session-title">Supply Chain Management II</div> 19721 <div class="session-chair"> 19722 Chair: Hans Ehm (Infineon Technologies AG)<br /> 19723 </div> 19724 <div class="slot-entry"> 19725 <a name="con165" tabindex="-1"></a> 19726 <div class="slot-title-line"> 19727 <span class="slot-title" 19728 >Component Redesigns and the Impact of their 19729 Implementation Policy</span 19730 > 19731 </div> 19732 <div> 19733 <span class="BAP award" 19734 >Best Contributed Applied Paper - Finalist</span 19735 > 19736 </div> 19737 <div class="slot-authors"> 19738 Steffi Neefs and Douniel Lamghari-Idrissi (ASML 19739 Netherlands B.V., Eindhoven University of Technology) 19740 and Rob Basten and Geert-Jan van Houtum (Eindhoven 19741 University of Technology) 19742 </div> 19743 <div class="slot-abstract"> 19744 <div> 19745 <a 19746 class="clickable no-decoration" 19747 id="vhsjs_view_436_1707793552_1996708" 19748 onclick="$('#vhsjs_view_436_1707793552_1996708').hide(); 19749 $('#vhsjs_hide_436_1707793552_1996708').show(); 19750 $('#435_1707793552_1996627').slideDown(function() { 19751 if (typeof Masonry === 'function') { 19752 $('.use_masonry').masonry(); 19753 }; 19754 19755 });" 19756 ><i class="fa fa-caret-right"></i>
19757 <span class="hover_link">Abstract</span></a 19758 ><a 19759 class="clickable no-decoration" 19760 id="vhsjs_hide_436_1707793552_1996708" 19761 onclick="$('#435_1707793552_1996627').hide(function() { 19762 if (typeof Masonry === 'function') { 19763 $('.use_masonry').masonry(); 19764 }; 19765 }); 19766 $('#vhsjs_hide_436_1707793552_1996708').hide(); 19767 $('#vhsjs_view_436_1707793552_1996708').show();" 19768 style="display: none" 19769 ><i class="fa fa-caret-down"></i> 19770 <span class="hover_link">Abstract</span></a 19771 > 19772 <div 19773 data-display-control="436_1707793552_1996708" 19774 id="435_1707793552_1996627" 19775 style="display: none" 19776 > 19777 <div class="arrow-slidedown"> 19778 <blockquote> 19779 An OEM who maintains a fleet of complex systems 19780 strives for high system availability for its 19781 customers. Frequently failing components lead to 19782 system unavailability and high maintenance 19783 costs. Consequently, the OEM might decide to 19784 upgrade components. We develop a model that 19785 quantifies the impact of the introduction of an 19786 upgraded component on the OEM's costs and number 19787 of failures to define the best implementation 19788 strategy. Using a Markov process, we evaluate 19789 four policies differing in the roll-out strategy 19790 of new parts, either immediate or corrective, 19791 and the phase-out strategy of old parts, either 19792 rework or salvage. The model is used in a case 19793 study at ASML. We conclude that, in the case 19794 study, reworking is preferred over salvaging as 19795 the phase-out strategy and corrective 19796 replacements are generally preferred over 19797 immediate replacements for the roll-out 19798 strategy. 19799 </blockquote> 19800 </div> 19801 </div> 19802 </div> 19803 </div> 19804 <div class="slot-urls"></div> 19805 <a href="/wsc23papers/187.pdf" target="_blank">pdf</a 19806 ><br /> 19807 </div> 19808 <div class="slot-entry"> 19809 <a name="inv162" tabindex="-1"></a> 19810 <div class="slot-title-line"> 19811 <span class="slot-title" 19812 >Exact and Heuristic Algorithms for a Bi-criteria 19813 Order-lot Pegging Problem in a Multi-Fab Setting</span 19814 > 19815 </div> 19816 <div class="slot-authors"> 19817 Andreas Haspecker and Lars Moench (University of Hagen) 19818 </div> 19819 <div class="slot-abstract"> 19820 <div> 19821 <a 19822 class="clickable no-decoration" 19823 id="vhsjs_view_438_1707793552_201798" 19824 onclick="$('#vhsjs_view_438_1707793552_201798').hide(); 19825 $('#vhsjs_hide_438_1707793552_201798').show(); 19826 $('#437_1707793552_2017899').slideDown(function() { 19827 if (typeof Masonry === 'function') { 19828 $('.use_masonry').masonry(); 19829 }; 19830 19831 });" 19832 ><i class="fa fa-caret-right"></i>
19833 <span class="hover_link">Abstract</span></a 19834 ><a 19835 class="clickable no-decoration" 19836 id="vhsjs_hide_438_1707793552_201798" 19837 onclick="$('#437_1707793552_2017899').hide(function() { 19838 if (typeof Masonry === 'function') { 19839 $('.use_masonry').masonry(); 19840 }; 19841 }); 19842 $('#vhsjs_hide_438_1707793552_201798').hide(); 19843 $('#vhsjs_view_438_1707793552_201798').show();" 19844 style="display: none" 19845 ><i class="fa fa-caret-down"></i> 19846 <span class="hover_link">Abstract</span></a 19847 > 19848 <div 19849 data-display-control="438_1707793552_201798" 19850 id="437_1707793552_2017899" 19851 style="display: none" 19852 > 19853 <div class="arrow-slidedown"> 19854 <blockquote> 19855 We study an order-lot pegging problem in 19856 semiconductor supply chains. The problem deals 19857 with assigning already released lots to orders 19858 and with planning wafer releases to fulfill 19859 orders if there are not enough lots in the wafer 19860 fabs. The objectives are minimizing the total 19861 tardiness of the orders and minimizing the total 19862 cost. We are interested in computing the set of 19863 Pareto-optimal plans. Based on a mixed-integer 19864 linear formulation, a ϵ-constraint method 19865 is proposed for small-sized problem instances. 19866 Moreover, a non-dominated sorting genetic 19867 algorithm (NSGA)-II algorithm is designed for 19868 tackling larger problem instances within a 19869 reasonable amount of computing time. We perform 19870 computational experiments with the 19871 ε-constraint method for small-sized problem 19872 instances and with the NSGA-II scheme for small- 19873 and medium-sized problem instances. 19874 </blockquote> 19875 </div> 19876 </div> 19877 </div> 19878 </div> 19879 <div class="slot-urls"></div> 19880 <a href="/wsc23papers/188.pdf" target="_blank">pdf</a 19881 ><br /> 19882 </div> 19883 <div class="slot-entry"> 19884 <a name="cea148" tabindex="-1"></a> 19885 <div class="slot-title-line"> 19886 <span class="slot-title" 19887 >A Case Study for Modeling the Economics of Foundry 19888 Operations</span 19889 > 19890 </div> 19891 <div class="slot-authors"> 19892 Larissa Nietner (LineLab, MIT); Parker Gould (InchFab); 19893 and Scott Nill (LineLab, MIT) 19894 </div> 19895 <div class="slot-abstract"> 19896 <div> 19897 <a 19898 class="clickable no-decoration" 19899 id="vhsjs_view_440_1707793552_2039216" 19900 onclick="$('#vhsjs_view_440_1707793552_2039216').hide(); 19901 $('#vhsjs_hide_440_1707793552_2039216').show(); 19902 $('#439_1707793552_2039132').slideDown(function() { 19903 if (typeof Masonry === 'function') { 19904 $('.use_masonry').masonry(); 19905 }; 19906 19907 });" 19908 ><i class="fa fa-caret-right"></i>
19909 <span class="hover_link">Abstract</span></a 19910 ><a 19911 class="clickable no-decoration" 19912 id="vhsjs_hide_440_1707793552_2039216" 19913 onclick="$('#439_1707793552_2039132').hide(function() { 19914 if (typeof Masonry === 'function') { 19915 $('.use_masonry').masonry(); 19916 }; 19917 }); 19918 $('#vhsjs_hide_440_1707793552_2039216').hide(); 19919 $('#vhsjs_view_440_1707793552_2039216').show();" 19920 style="display: none" 19921 ><i class="fa fa-caret-down"></i> 19922 <span class="hover_link">Abstract</span></a 19923 > 19924 <div 19925 data-display-control="440_1707793552_2039216" 19926 id="439_1707793552_2039132" 19927 style="display: none" 19928 > 19929 <div class="arrow-slidedown"> 19930 <blockquote> 19931 This case study presents a novel approach for 19932 modeling a fab, which allows for more rapid 19933 results than traditional simulation, while 19934 optimizing various variables like tool count or 19935 throughput, and capturing equipment sharing 19936 between co-produced devices. This modeling 19937 method was applied at InchFab, a foundry that 19938 uses ultra-small substrate sizes to allow for 19939 more flexibility and lower costs when 19940 fabricating small production quantities. The new 19941 approach was used to find the cost-optimal rate 19942 achievable for a primary product on certain tool 19943 counts - and then the cost-optimal rate of a 19944 secondary product, without any changes to 19945 equipment count. Using novel types of analyses 19946 and sensitivity figures, we demonstrate that it 19947 can be economically sensible to add a product to 19948 a fab that is already producing the cost-optimal 19949 quantity of a base product. This is an important 19950 finding, as some fabs consider offering 19951 additional foundry services on existing 19952 equipment. 19953 </blockquote> 19954 </div> 19955 </div> 19956 </div> 19957 </div> 19958 <div class="slot-urls"></div> 19959 <a href="/wsc23papers/cea148.pdf" target="_blank">pdf</a 19960 ><br /> 19961 </div> 19962 </div> 19963 <div class="session-entry"> 19964 <span class="session-event-type">Technical Session</span 19965 ><span class="type-track-spacer"> · </span 19966 ><span class="program-track" 19967 >MASM: Semiconductor Manufacturing</span 19968 ><br /> 19969 <div class="session-title"> 19970 Digital Twins and Simulation 19971 </div> 19972 <div class="session-chair"> 19973 Chair: Cathal Heavey (University of Limerick)<br /> 19974 </div> 19975 <div class="slot-entry"> 19976 <a name="cea110" tabindex="-1"></a> 19977 <div class="slot-title-line"> 19978 <span class="slot-title" 19979 >Digital Twin for Design and Analysis of Cluster Tool 19980 in Wafer Fabrication</span 19981 > 19982 </div> 19983 <div class="slot-authors">
19984 Joonick Hwang and Sang Do Noh (Sungkyunkwan University) 19985 </div> 19986 <div class="slot-abstract"> 19987 <div> 19988 <a 19989 class="clickable no-decoration" 19990 id="vhsjs_view_442_1707793552_2084289" 19991 onclick="$('#vhsjs_view_442_1707793552_2084289').hide(); 19992 $('#vhsjs_hide_442_1707793552_2084289').show(); 19993 $('#441_1707793552_2084203').slideDown(function() { 19994 if (typeof Masonry === 'function') { 19995 $('.use_masonry').masonry(); 19996 }; 19997 19998 });" 19999 ><i class="fa fa-caret-right"></i> 20000 <span class="hover_link">Abstract</span></a 20001 ><a 20002 class="clickable no-decoration" 20003 id="vhsjs_hide_442_1707793552_2084289" 20004 onclick="$('#441_1707793552_2084203').hide(function() { 20005 if (typeof Masonry === 'function') { 20006 $('.use_masonry').masonry(); 20007 }; 20008 }); 20009 $('#vhsjs_hide_442_1707793552_2084289').hide(); 20010 $('#vhsjs_view_442_1707793552_2084289').show();" 20011 style="display: none" 20012 ><i class="fa fa-caret-down"></i> 20013 <span class="hover_link">Abstract</span></a 20014 > 20015 <div 20016 data-display-control="442_1707793552_2084289" 20017 id="441_1707793552_2084203" 20018 style="display: none" 20019 > 20020 <div class="arrow-slidedown"> 20021 <blockquote> 20022 In the semiconductor industry, many retrofits 20023 are being made to improve the production 20024 efficiency of manufacturing facilities. However, 20025 due to the nature of the data provided by the 20026 cluster tool, which is a semiconductor 20027 manufacturing facility, engineers have some 20028 limitations in utilizing it. To address this 20029 issue, it is necessary to introduce a digital 20030 twin model that can verify the performance of 20031 the semiconductor process cluster tool in a 20032 virtual environment, and to apply optimal mass 20033 production conditions based on this predictive 20034 data in the operational stage. In this study, we 20035 propose a digital twin model that visualize 20036 congestion factors during wafer transfer and 20037 evaluate the productivity of cluster tools. 20038 </blockquote> 20039 </div> 20040 </div> 20041 </div> 20042 </div> 20043 <div class="slot-urls"></div> 20044 <a href="/wsc23papers/cea110.pdf" target="_blank">pdf</a 20045 ><br /> 20046 </div> 20047 <div class="slot-entry"> 20048 <a name="inv195" tabindex="-1"></a> 20049 <div class="slot-title-line"> 20050 <span class="slot-title" 20051 >A Study on the Impact of Lot Priorities Mix on Cycle 20052 Times in Semiconductor Manufacturing</span 20053 > 20054 </div> 20055 <div class="slot-authors"> 20056 Adrien Wartelle, Stéphane 20057 Dauzère-Pérès, and Claude Yugma (Ecole 20058 des Mines de Saint-Etienne) and Quentin Christ and 20059 Renaud Roussel (STMicroelectronics) 20060 </div> 20061 <div class="slot-abstract"> 20062 <div> 20063 <a 20064 class="clickable no-decoration" 20065 id="vhsjs_view_444_1707793552_210738" 20066 onclick="$('#vhsjs_view_444_1707793552_210738').hide(); 20067 $('#vhsjs_hide_444_1707793552_210738').show(); 20068 $('#443_1707793552_2107294').slideDown(function() { 20069 if (typeof Masonry === 'function') { 20070 $('.use_masonry').masonry(); 20071 }; 20072 20073 });" 20074 ><i class="fa fa-caret-right"></i>
20075 <span class="hover_link">Abstract</span></a 20076 ><a 20077 class="clickable no-decoration" 20078 id="vhsjs_hide_444_1707793552_210738" 20079 onclick="$('#443_1707793552_2107294').hide(function() { 20080 if (typeof Masonry === 'function') { 20081 $('.use_masonry').masonry(); 20082 }; 20083 }); 20084 $('#vhsjs_hide_444_1707793552_210738').hide(); 20085 $('#vhsjs_view_444_1707793552_210738').show();" 20086 style="display: none" 20087 ><i class="fa fa-caret-down"></i> 20088 <span class="hover_link">Abstract</span></a 20089 > 20090 <div 20091 data-display-control="444_1707793552_210738" 20092 id="443_1707793552_2107294" 20093 style="display: none" 20094 > 20095 <div class="arrow-slidedown"> 20096 <blockquote> 20097 This paper presents a study on the priority mix 20098 planning problem in semiconductor fabrication 20099 using simulation. The objective of the study is 20100 to analyze the impact of the mix of different 20101 lot types associated with their priority on the 20102 cycle time of the Implantation workshop. We have 20103 specifically analyzed the waiting time lots and 20104 the associated speed up or speed down on a 20105 work-center. The tests were conducted using 20106 Anylogic 8 on industrial instances from 20107 STMicroelectronics Crolles. Results shows that a 20108 speedup of more than 300% for high priority lots 20109 and speed down of less than 10% is possible if 20110 the proportion high priority lots is kept under 20111 10%. This study initiates a first step toward a 20112 better priority mix management which has a 20113 strategic central place of in the semiconductor 20114 industry. 20115 </blockquote> 20116 </div> 20117 </div> 20118 </div> 20119 </div> 20120 <div class="slot-urls"></div> 20121 <a href="/wsc23papers/189.pdf" target="_blank">pdf</a 20122 ><br /> 20123 </div> 20124 <div class="slot-entry"> 20125 <a name="inv152" tabindex="-1"></a> 20126 <div class="slot-title-line"> 20127 <span class="slot-title" 20128 >Backward Simulation: A Customer-Focused 20129 Diversification of Fab Simulation Applications in a 20130 Highly Automated Semiconductor Production Line</span 20131 > 20132 </div> 20133 <div class="slot-authors"> 20134 Wolfgang Scholl and Patrick Preuß (Infineon 20135 Technologies Dresden GmbH) and Christoph Laroque and 20136 Madlene Leissau (University of Applied Sciences Zwickau) 20137 </div> 20138 <div class="slot-abstract"> 20139 <div> 20140 <a 20141 class="clickable no-decoration" 20142 id="vhsjs_view_446_1707793552_2138097" 20143 onclick="$('#vhsjs_view_446_1707793552_2138097').hide(); 20144 $('#vhsjs_hide_446_1707793552_2138097').show(); 20145 $('#445_1707793552_2138014').slideDown(function() { 20146 if (typeof Masonry === 'function') { 20147 $('.use_masonry').masonry(); 20148 }; 20149 20150 });" 20151 ><i class="fa fa-caret-right"></i>
20152 <span class="hover_link">Abstract</span></a 20153 ><a 20154 class="clickable no-decoration" 20155 id="vhsjs_hide_446_1707793552_2138097" 20156 onclick="$('#445_1707793552_2138014').hide(function() { 20157 if (typeof Masonry === 'function') { 20158 $('.use_masonry').masonry(); 20159 }; 20160 }); 20161 $('#vhsjs_hide_446_1707793552_2138097').hide(); 20162 $('#vhsjs_view_446_1707793552_2138097').show();" 20163 style="display: none" 20164 ><i class="fa fa-caret-down"></i> 20165 <span class="hover_link">Abstract</span></a 20166 > 20167 <div 20168 data-display-control="446_1707793552_2138097" 20169 id="445_1707793552_2138014" 20170 style="display: none" 20171 > 20172 <div class="arrow-slidedown"> 20173 <blockquote> 20174 In modern manufacturing environments, the 20175 digital transformation to smart factories cannot 20176 be achieved without data-driven methods like 20177 discrete, event-driven simulation. This paper 20178 provides an overview of existing current 20179 simulation applications at Infineon Dresden in 20180 this area, especially on short-term simulation 20181 for production control and long-term simulations 20182 to forecast process flows in the wafer 20183 fabrication facilities. Furthermore, it 20184 illustrates the current status of research 20185 activities in the area of backward simulation 20186 for operational decision support for order 20187 scheduling by some latest research results. 20188 </blockquote> 20189 </div> 20190 </div> 20191 </div> 20192 </div> 20193 <div class="slot-urls"></div> 20194 <a href="/wsc23papers/190.pdf" target="_blank">pdf</a 20195 ><br /> 20196 </div> 20197 </div> 20198 <div class="session-entry"> 20199 <span class="session-event-type">Technical Session</span 20200 ><span class="type-track-spacer"> · </span 20201 ><span class="program-track" 20202 >MASM: Semiconductor Manufacturing</span 20203 ><br /> 20204 <div class="session-title"> 20205 Modeling Techniques in Semiconductor Manufacturing 20206 </div> 20207 <div class="session-chair"> 20208 Chair: Robert Dodge (Arizona State University)<br /> 20209 </div> 20210 <div class="slot-entry"> 20211 <a name="cea123" tabindex="-1"></a> 20212 <div class="slot-title-line"> 20213 <span class="slot-title" 20214 >Duplicate Reticles Management System</span 20215 > 20216 </div> 20217 <div class="slot-authors"> 20218 Sandar Kyaw, Ronald Taylor, and Jean Fakhoury 20219 (GLOBALFOUNDRIES) 20220 </div> 20221 <div class="slot-abstract"> 20222 <div> 20223 <a 20224 class="clickable no-decoration" 20225 id="vhsjs_view_448_1707793552_21838" 20226 onclick="$('#vhsjs_view_448_1707793552_21838').hide(); 20227 $('#vhsjs_hide_448_1707793552_21838').show(); 20228 $('#447_1707793552_2183716').slideDown(function() { 20229 if (typeof Masonry === 'function') { 20230 $('.use_masonry').masonry(); 20231 }; 20232 20233 });" 20234 ><i class="fa fa-caret-right"></i>
20235 <span class="hover_link">Abstract</span></a 20236 ><a 20237 class="clickable no-decoration" 20238 id="vhsjs_hide_448_1707793552_21838" 20239 onclick="$('#447_1707793552_2183716').hide(function() { 20240 if (typeof Masonry === 'function') { 20241 $('.use_masonry').masonry(); 20242 }; 20243 }); 20244 $('#vhsjs_hide_448_1707793552_21838').hide(); 20245 $('#vhsjs_view_448_1707793552_21838').show();" 20246 style="display: none" 20247 ><i class="fa fa-caret-down"></i> 20248 <span class="hover_link">Abstract</span></a 20249 > 20250 <div 20251 data-display-control="448_1707793552_21838" 20252 id="447_1707793552_2183716" 20253 style="display: none" 20254 > 20255 <div class="arrow-slidedown"> 20256 <blockquote> 20257 Duplicate reticles provide a fab with an 20258 opportunity to mitigate the impact of 20259 catastrophic reticle damage or the need for 20260 offsite repair/cleaning and provide the 20261 necessary capacity for products in a high volume 20262 manufacturing environment. Implementation of a 20263 management system for duplicate reticles helps 20264 to maintain a minimum number of run paths while 20265 ensuring availability of multiple reticles to 20266 process lots simultaneously. Dedicating the 20267 duplicate reticles each to a group of exposure 20268 tools prevents duplicate reticles from ending up 20269 on the same exposure tool, and managing this 20270 dedication by tool/reticle inhibits has proven 20271 to be an effective method of distributing the 20272 WIP between the exposure tools while minimizing 20273 the management of the layer supported by those 20274 duplicate reticles. 20275 </blockquote> 20276 </div> 20277 </div> 20278 </div> 20279 </div> 20280 <div class="slot-urls"></div> 20281 <a href="/wsc23papers/cea123.pdf" target="_blank">pdf</a 20282 ><br /> 20283 </div> 20284 <div class="slot-entry"> 20285 <a name="inv201" tabindex="-1"></a> 20286 <div class="slot-title-line"> 20287 <span class="slot-title" 20288 >A Testing Based Approach for Security Analysis of 20289 Smart Semiconductor Systems</span 20290 > 20291 </div> 20292 <div class="slot-authors"> 20293 Robert Dodge, Giulia Pedrielli, and Petar Jevtić 20294 (Arizona State University) 20295 </div> 20296 <div class="slot-abstract"> 20297 <div> 20298 <a 20299 class="clickable no-decoration" 20300 id="vhsjs_view_450_1707793552_2207055" 20301 onclick="$('#vhsjs_view_450_1707793552_2207055').hide(); 20302 $('#vhsjs_hide_450_1707793552_2207055').show(); 20303 $('#449_1707793552_2206972').slideDown(function() { 20304 if (typeof Masonry === 'function') { 20305 $('.use_masonry').masonry(); 20306 }; 20307 20308 });" 20309 ><i class="fa fa-caret-right"></i>
20310 <span class="hover_link">Abstract</span></a 20311 ><a 20312 class="clickable no-decoration" 20313 id="vhsjs_hide_450_1707793552_2207055" 20314 onclick="$('#449_1707793552_2206972').hide(function() { 20315 if (typeof Masonry === 'function') { 20316 $('.use_masonry').masonry(); 20317 }; 20318 }); 20319 $('#vhsjs_hide_450_1707793552_2207055').hide(); 20320 $('#vhsjs_view_450_1707793552_2207055').show();" 20321 style="display: none" 20322 ><i class="fa fa-caret-down"></i> 20323 <span class="hover_link">Abstract</span></a 20324 > 20325 <div 20326 data-display-control="450_1707793552_2207055" 20327 id="449_1707793552_2206972" 20328 style="display: none" 20329 > 20330 <div class="arrow-slidedown"> 20331 <blockquote> 20332 Digital factories have been recognized as a 20333 paradigm with considerable promise for improving 20334 manufacturing performance. Digital Twins have 20335 emerged as a powerful tool to improve control 20336 performance for large-scale smart manufacturing 20337 systems. We argue that DT-based smart factories 20338 are vulnerable to attacks that use the DT to 20339 damage the system while remaining undetectable, 20340 specifically in high-cost processes, where DT 20341 technologies are more likely to be deployed. As 20342 an instructive example, we consider smart 20343 semiconductor processes with focus on 20344 photolithography. To this end, we formulate a 20345 static optimization problem to maximize the 20346 damage of a cyber-attack against a 20347 photolithography digital twin that minimizes 20348 detectability to the process controller. Results 20349 demonstrate that this problem formulation 20350 provides attack policies that successfully 20351 reduce the throughput of the system at trade off 20352 of increased detectability to a common process 20353 control technique. Results encourage more 20354 research in the domain, especially to face 20355 scalability and policy-like solutions. 20356 </blockquote> 20357 </div> 20358 </div> 20359 </div> 20360 </div> 20361 <div class="slot-urls"></div> 20362 <a href="/wsc23papers/191.pdf" target="_blank">pdf</a 20363 ><br /> 20364 </div> 20365 <div class="slot-entry"> 20366 <a name="con176" tabindex="-1"></a> 20367 <div class="slot-title-line"> 20368 <span class="slot-title" 20369 >Reusable Ontology Generation and Matching from 20370 Simulation Models</span 20371 > 20372 </div> 20373 <div class="slot-authors"> 20374 Ming-Yu Tu, Hans Ehm, Abdelgafar Ismail, and Philipp 20375 Ulrich (Infineon Technologies AG) 20376 </div> 20377 <div class="slot-abstract"> 20378 <div> 20379 <a 20380 class="clickable no-decoration" 20381 id="vhsjs_view_452_1707793552_223002" 20382 onclick="$('#vhsjs_view_452_1707793552_223002').hide(); 20383 $('#vhsjs_hide_452_1707793552_223002').show(); 20384 $('#451_1707793552_2229939').slideDown(function() { 20385 if (typeof Masonry === 'function') { 20386 $('.use_masonry').masonry(); 20387 }; 20388 20389 });" 20390 ><i class="fa fa-caret-right"></i>
20391 <span class="hover_link">Abstract</span></a 20392 ><a 20393 class="clickable no-decoration" 20394 id="vhsjs_hide_452_1707793552_223002" 20395 onclick="$('#451_1707793552_2229939').hide(function() { 20396 if (typeof Masonry === 'function') { 20397 $('.use_masonry').masonry(); 20398 }; 20399 }); 20400 $('#vhsjs_hide_452_1707793552_223002').hide(); 20401 $('#vhsjs_view_452_1707793552_223002').show();" 20402 style="display: none" 20403 ><i class="fa fa-caret-down"></i> 20404 <span class="hover_link">Abstract</span></a 20405 > 20406 <div 20407 data-display-control="452_1707793552_223002" 20408 id="451_1707793552_2229939" 20409 style="display: none" 20410 > 20411 <div class="arrow-slidedown"> 20412 <blockquote> 20413 As simulating semiconductor manufacturing grows 20414 complex, model reuse becomes appealing since it 20415 can reduce the time incurred in developing 20416 future models. Also, considering a large network 20417 of the semiconductor supply chain, knowledge 20418 sharing can enable the efficient development of 20419 simulation models in a collaborative 20420 organization. Such necessity of reusability and 20421 interoperability of simulation models motivates 20422 this paper. We will address these challenges 20423 through ontological modeling and linking of the 20424 simulation components. The first application is 20425 generating reusable ontologies from simulation 20426 models. Another discussed application is 20427 ontology matching for knowledge sharing between 20428 simulation components and a meta-model of the 20429 semiconductor supply chain. The proposed 20430 approach succeeds in automatically transforming 20431 simulation into reusable knowledge and 20432 identifying interconnection in a semiconductor 20433 manufacturing system. 20434 </blockquote> 20435 </div> 20436 </div> 20437 </div> 20438 </div> 20439 <div class="slot-urls"></div> 20440 <a href="/wsc23papers/192.pdf" target="_blank">pdf</a 20441 ><br /> 20442 </div> 20443 </div> 20444 <div class="session-entry"> 20445 <span class="session-event-type">Technical Session</span 20446 ><span class="type-track-spacer"> · </span 20447 ><span class="program-track" 20448 >MASM: Semiconductor Manufacturing</span 20449 ><br /> 20450 <div class="session-title">Scheduling II</div> 20451 <div class="session-chair"> 20452 Chair: Stephane Dauzère-Pérès (École 20453 Nationale Supérieure des Mines de Saint-Étienne, 20454 BI Norwegian Business School)<br /> 20455 </div> 20456 <div class="slot-entry"> 20457 <a name="cea122" tabindex="-1"></a> 20458 <div class="slot-title-line"> 20459 <span class="slot-title" 20460 >Industrial Multi-Objective Optimization of a Large 20461 Complex Job-Shop in Semiconductor Manufacturing</span 20462 > 20463 </div> 20464 <div class="slot-authors"> 20465 Abdel Bitar and Sebastian Knopp (Planimize); Karim 20466 Tamssaouet (Planimize, BI Norwegian School of 20467 Management); Stéphane Dauzère-Pérès 20468 (Ecole des Mines de Saint-Etienne); and Ludovic Delcloy 20469 and Renaud Roussel (STMicroelectronics, Crolles) 20470 </div> 20471 <div class="slot-abstract"> 20472 <div> 20473 <a 20474 class="clickable no-decoration" 20475 id="vhsjs_view_454_1707793552_2284927" 20476 onclick="$('#vhsjs_view_454_1707793552_2284927').hide(); 20477 $('#vhsjs_hide_454_1707793552_2284927').show(); 20478 $('#453_1707793552_2284846').slideDown(function() { 20479 if (typeof Masonry === 'function') { 20480 $('.use_masonry').masonry(); 20481 }; 20482 20483 });" 20484 ><i class="fa fa-caret-right"></i>
20485 <span class="hover_link">Abstract</span></a 20486 ><a 20487 class="clickable no-decoration" 20488 id="vhsjs_hide_454_1707793552_2284927" 20489 onclick="$('#453_1707793552_2284846').hide(function() { 20490 if (typeof Masonry === 'function') { 20491 $('.use_masonry').masonry(); 20492 }; 20493 }); 20494 $('#vhsjs_hide_454_1707793552_2284927').hide(); 20495 $('#vhsjs_view_454_1707793552_2284927').show();" 20496 style="display: none" 20497 ><i class="fa fa-caret-down"></i> 20498 <span class="hover_link">Abstract</span></a 20499 > 20500 <div 20501 data-display-control="454_1707793552_2284927" 20502 id="453_1707793552_2284846" 20503 style="display: none" 20504 > 20505 <div class="arrow-slidedown"> 20506 <blockquote> 20507 This paper surveys the industrialization of an 20508 advanced optimization engine that was developed 20509 by Planimize and put into production in the 20510 cleaning and diffusion work center of the most 20511 advanced factory of a semiconductor 20512 manufacturing company. Hundreds of lots 20513 requiring several thousands operations in the 20514 work center must be scheduled on about 150 20515 machines, while taking complex constraints into 20516 account, in particular hundreds of time 20517 constraints, and optimizing a collection of 20518 criteria. The optimization engine provides 20519 significantly better results, runs significantly 20520 faster, and can handle much larger problem 20521 instances than the previous Constraint 20522 Programming optimization engine used in the 20523 factory. 20524 </blockquote> 20525 </div> 20526 </div> 20527 </div> 20528 </div> 20529 <div class="slot-urls"></div> 20530 <a href="/wsc23papers/cea122.pdf" target="_blank">pdf</a 20531 ><br /> 20532 </div> 20533 <div class="slot-entry"> 20534 <a name="cea112" tabindex="-1"></a> 20535 <div class="slot-title-line"> 20536 <span class="slot-title" 20537 >Minimizing Makespan for a Multiple Orders Per Job 20538 Scheduling Problem in a Two-stage Permutation 20539 Flowshop</span 20540 > 20541 </div> 20542 <div class="slot-authors"> 20543 Rohan Korde and John Fowler (Arizona State University) 20544 and Lars Mönch (FernUniversität in Hagen) 20545 </div> 20546 <div class="slot-abstract"> 20547 <div> 20548 <a 20549 class="clickable no-decoration" 20550 id="vhsjs_view_456_1707793552_230617" 20551 onclick="$('#vhsjs_view_456_1707793552_230617').hide(); 20552 $('#vhsjs_hide_456_1707793552_230617').show(); 20553 $('#455_1707793552_2306092').slideDown(function() { 20554 if (typeof Masonry === 'function') { 20555 $('.use_masonry').masonry(); 20556 }; 20557 20558 });" 20559 ><i class="fa fa-caret-right"></i>
20560 <span class="hover_link">Abstract</span></a 20561 ><a 20562 class="clickable no-decoration" 20563 id="vhsjs_hide_456_1707793552_230617" 20564 onclick="$('#455_1707793552_2306092').hide(function() { 20565 if (typeof Masonry === 'function') { 20566 $('.use_masonry').masonry(); 20567 }; 20568 }); 20569 $('#vhsjs_hide_456_1707793552_230617').hide(); 20570 $('#vhsjs_view_456_1707793552_230617').show();" 20571 style="display: none" 20572 ><i class="fa fa-caret-down"></i> 20573 <span class="hover_link">Abstract</span></a 20574 > 20575 <div 20576 data-display-control="456_1707793552_230617" 20577 id="455_1707793552_2306092" 20578 style="display: none" 20579 > 20580 <div class="arrow-slidedown"> 20581 <blockquote> 20582 The scheduling problem we study in this paper is 20583 known as a multiple orders per job (MOJ) (Mason 20584 et al. 2004) problem which is encountered in a 20585 few different industries including front-end 20586 semiconductor manufacturing. We look at the MOJ 20587 scheduling problem in a two-stage permutation 20588 flowshop with some real-world constraints with 20589 the goal of minimizing the makespan. We use a 20590 MIP solver and various heuristics to solve this 20591 NP-hard scheduling problem for various stage 20592 configurations and bottleneck types. For 20593 moj(ipm-ipm) the makespan was minimized by the 20594 MIP solver regardless of the bottleneck type for 20595 over 90% of the small-sized problem instances. 20596 When the heuristics minimized the makespan, the 20597 Slope heuristic was the fastest NEH heuristic 20598 was the slowest for over 90% of the large-sized 20599 problem instances. 20600 </blockquote> 20601 </div> 20602 </div> 20603 </div> 20604 </div> 20605 <div class="slot-urls"></div> 20606 <a href="/wsc23papers/cea112.pdf" target="_blank">pdf</a 20607 ><br /> 20608 </div> 20609 <div class="slot-entry"> 20610 <a name="con336" tabindex="-1"></a> 20611 <div class="slot-title-line"> 20612 <span class="slot-title" 20613 >Combining Time Series Data and Snapshot Data for 20614 Situation Aware Dispatching in Semiconductor 20615 Manufacturing</span 20616 > 20617 </div> 20618 <div class="slot-authors"> 20619 Chew Wye Chan and Boon Ping Gan (D-SIMLAB Technologies 20620 Pte Ltd) and Wentong Cai (Nanyang Technological 20621 University) 20622 </div> 20623 <div class="slot-abstract"> 20624 <div> 20625 <a 20626 class="clickable no-decoration" 20627 id="vhsjs_view_458_1707793552_2330377" 20628 onclick="$('#vhsjs_view_458_1707793552_2330377').hide(); 20629 $('#vhsjs_hide_458_1707793552_2330377').show(); 20630 $('#457_1707793552_23303').slideDown(function() { 20631 if (typeof Masonry === 'function') { 20632 $('.use_masonry').masonry(); 20633 }; 20634 20635 });" 20636 ><i class="fa fa-caret-right"></i>
20637 <span class="hover_link">Abstract</span></a 20638 ><a 20639 class="clickable no-decoration" 20640 id="vhsjs_hide_458_1707793552_2330377" 20641 onclick="$('#457_1707793552_23303').hide(function() { 20642 if (typeof Masonry === 'function') { 20643 $('.use_masonry').masonry(); 20644 }; 20645 }); 20646 $('#vhsjs_hide_458_1707793552_2330377').hide(); 20647 $('#vhsjs_view_458_1707793552_2330377').show();" 20648 style="display: none" 20649 ><i class="fa fa-caret-down"></i> 20650 <span class="hover_link">Abstract</span></a 20651 > 20652 <div 20653 data-display-control="458_1707793552_2330377" 20654 id="457_1707793552_23303" 20655 style="display: none" 20656 > 20657 <div class="arrow-slidedown"> 20658 <blockquote> 20659 Dispatch rules are commonly used to schedule 20660 lots in the semiconductor industry. Previous 20661 studies have indicated that adapting dispatch 20662 rules can improve overall factory performance. 20663 Machine learning has proven useful in learning 20664 the relationship between manufacturing 20665 situations and dispatch rules. However, using 20666 only snapshot data at a given point in time to 20667 generate features for these models does not 20668 account for trends in the manufacturing 20669 situation, which can be represented as time 20670 series data. To address this issue, the proposed 20671 method generates features from time series data 20672 and combines them with features from snapshot 20673 data to train machine learning models for 20674 dispatch rule prediction. The results 20675 demonstrate the effectiveness of this 20676 methodology, as the combination of features from 20677 both types of data achieves the highest 20678 prediction accuracy. Simulation results show 20679 that this approach can adapt the dispatch rule 20680 according to the manufacturing situation and 20681 achieve a comparable factory performance. 20682 </blockquote> 20683 </div> 20684 </div> 20685 </div> 20686 </div> 20687 <div class="slot-urls"></div> 20688 <a href="/wsc23papers/193.pdf" target="_blank">pdf</a 20689 ><br /> 20690 </div> 20691 </div> 20692 <div class="session-entry"> 20693 <span class="session-event-type">Technical Session</span 20694 ><span class="type-track-spacer"> · </span 20695 ><span class="program-track" 20696 >MASM: Semiconductor Manufacturing</span 20697 ><br /> 20698 <div class="session-title"> 20699 MASM Keynote: Simulation, Optimization and AI for 20700 Semiconductor Manufacturing and Supply Chains: Four 20701 Decades of Progress and a Vision for the Future 20702 </div> 20703 <div class="session-chair"> 20704 Chair: Lars Moench (University of Hagen)<br /> 20705 </div> 20706 <div class="slot-entry"> 20707 <a name="prog105" tabindex="-1"></a> 20708 <div class="slot-title-line">
20709 <span class="slot-title" 20710 >Simulation, Optimization and AI for Semiconductor 20711 Manufacturing and Supply Chains: Four Decades of 20712 Progress and a Vision for the Future</span 20713 > 20714 </div> 20715 <div class="slot-authors"> 20716 Hans Ehm (Infineon Technologies AG) 20717 </div> 20718 <div class="slot-abstract"> 20719 <div> 20720 <a 20721 class="clickable no-decoration" 20722 id="vhsjs_view_460_1707793552_2387755" 20723 onclick="$('#vhsjs_view_460_1707793552_2387755').hide(); 20724 $('#vhsjs_hide_460_1707793552_2387755').show(); 20725 $('#459_1707793552_2387674').slideDown(function() { 20726 if (typeof Masonry === 'function') { 20727 $('.use_masonry').masonry(); 20728 }; 20729 20730 });" 20731 ><i class="fa fa-caret-right"></i> 20732 <span class="hover_link">Abstract</span></a 20733 ><a 20734 class="clickable no-decoration" 20735 id="vhsjs_hide_460_1707793552_2387755" 20736 onclick="$('#459_1707793552_2387674').hide(function() { 20737 if (typeof Masonry === 'function') { 20738 $('.use_masonry').masonry(); 20739 }; 20740 }); 20741 $('#vhsjs_hide_460_1707793552_2387755').hide(); 20742 $('#vhsjs_view_460_1707793552_2387755').show();" 20743 style="display: none" 20744 ><i class="fa fa-caret-down"></i> 20745 <span class="hover_link">Abstract</span></a 20746 > 20747 <div 20748 data-display-control="460_1707793552_2387755" 20749 id="459_1707793552_2387674" 20750 style="display: none" 20751 > 20752 <div class="arrow-slidedown"> 20753 <blockquote> 20754 Semiconductor manufacturing and supply chain 20755 processes are one of the most complex but can be 20756 considered at the same time also as one of the 20757 most rewarding processes in the world. In 20758 thousands of detailed unit chemical and physical 20759 processes in cleanrooms and under statistical 20760 process control chips on wafers emerge and are 20761 assembled and tested to components. The Modeling 20762 and Analysis of Semiconductor Manufacturing 20763 (MASM) conference embedded in the annual Winter 20764 Simulation Conference (WSC) was, is, and will be 20765 key to understand the optimization and 20766 simulation challenges in this domain. 20767 <br /><br />The operating curve management 20768 targeting a low variability value and thus 20769 enabling a low flow factor - thus speed - and 20770 high utilization - thus a good cost position - 20771 at the same time has been an early achievement. 20772 With discrete-event, agent based, and system 20773 dynamic simulations on the four levels (machine, 20774 fab, internal and external supply chain) 20775 solution options for complex interactions could 20776 be proposed based on sophisticated mathematical 20777 models running on simulation testbeds like the 20778 MIMAC models and their successors. Accurate 20779 planning and advanced scheduling, available to 20780 promise (ATP) generation and usage with 20781 traditional or artificial intelligence (AI) / 20782 deep learning (DL) methods requires a huge 20783 amount of real data or qualified synthetic data 20784 (QSD). <br /><br />The semantic web for 20785 semiconductor and supply chain containing 20786 semiconductors bears the potential to enable the 20787 provision of those urgently needed QSD in 20788 volume, (integrated) complexity and accuracy 20789 needed. Quantum bit (qubit) based algorithm 20790 could provide the speed for the next and 20791 over-next generation for optimization and 20792 simulation in our domain. 20793 </blockquote> 20794 </div> 20795 </div> 20796 </div> 20797 </div> 20798 <div class="slot-urls"></div> 20799 <a href="/wsc23papers/prog105.pdf" target="_blank">pdf</a 20800 ><br /> 20801 </div> 20802 </div> 20803 <div class="session-entry">
20804 <span class="session-event-type">Technical Session</span 20805 ><span class="type-track-spacer"> · </span 20806 ><span class="program-track" 20807 >MASM: Semiconductor Manufacturing</span 20808 ><br /> 20809 <div class="session-title"> 20810 Panel: Semiconductor Manufacturing in Times of 20811 Geopolitical Tensions 20812 </div> 20813 <div class="session-chair"> 20814 Chair: Peter Lendermann (D-SIMLAB Technologies Pte Ltd)<br /> 20815 </div> 20816 <div class="slot-entry"> 20817 <a name="prog108" tabindex="-1"></a> 20818 <div class="slot-title-line"> 20819 <span class="slot-title" 20820 >Semiconductor Manufacturing in Times of Geopolitical 20821 Tensions: How MASM Can Help with Making Supply Chains 20822 More Resilient</span 20823 > 20824 </div> 20825 <div class="slot-authors"> 20826 Peter Lendermann (D-SIMLAB Technologies) 20827 </div> 20828 <div class="slot-abstract"> 20829 <div> 20830 <a 20831 class="clickable no-decoration" 20832 id="vhsjs_view_462_1707793552_2492929" 20833 onclick="$('#vhsjs_view_462_1707793552_2492929').hide(); 20834 $('#vhsjs_hide_462_1707793552_2492929').show(); 20835 $('#461_1707793552_2492843').slideDown(function() { 20836 if (typeof Masonry === 'function') { 20837 $('.use_masonry').masonry(); 20838 }; 20839 20840 });" 20841 ><i class="fa fa-caret-right"></i> 20842 <span class="hover_link">Abstract</span></a 20843 ><a 20844 class="clickable no-decoration" 20845 id="vhsjs_hide_462_1707793552_2492929" 20846 onclick="$('#461_1707793552_2492843').hide(function() { 20847 if (typeof Masonry === 'function') { 20848 $('.use_masonry').masonry(); 20849 }; 20850 }); 20851 $('#vhsjs_hide_462_1707793552_2492929').hide(); 20852 $('#vhsjs_view_462_1707793552_2492929').show();" 20853 style="display: none" 20854 ><i class="fa fa-caret-down"></i> 20855 <span class="hover_link">Abstract</span></a 20856 > 20857 <div 20858 data-display-control="462_1707793552_2492929" 20859 id="461_1707793552_2492843" 20860 style="display: none" 20861 > 20862 <div class="arrow-slidedown"> 20863 <blockquote> 20864 This panel assembles a number of prominent 20865 representatives from industry and academia to 20866 discuss how semiconductor supply chains in times 20867 of increasing geopolitical risks can be made 20868 more resilient through Modeling and Analysis of 20869 Semiconductor Manufacturing (MASM) techniques 20870 and enabling software solutions. 20871 </blockquote> 20872 </div> 20873 </div> 20874 </div> 20875 </div> 20876 <div class="slot-urls"></div> 20877 <a href="/wsc23papers/prog108.pdf" target="_blank">pdf</a 20878 ><br /> 20879 </div> 20880 </div> 20881 <div class="session-entry"> 20882 <span class="session-event-type">Technical Session</span 20883 ><span class="type-track-spacer"> · </span 20884 ><span class="program-track" 20885 >MASM: Semiconductor Manufacturing</span 20886 ><br /> 20887 <div class="session-title">Data and Modeling Issues</div> 20888 <div class="session-chair"> 20889 Chair: Oliver Rose (University of the Bundeswehr 20890 Munich)<br /> 20891 </div> 20892 <div class="slot-entry"> 20893 <a name="con159" tabindex="-1"></a> 20894 <div class="slot-title-line">
20895 <span class="slot-title" 20896 >Semiconductor Equipment Health Monitoring with 20897 Multi-View Data</span 20898 > 20899 </div> 20900 <div class="slot-authors"> 20901 Jeongsun Ahn, Hong-Yeon Kim, Sang-Hyun Cho, and 20902 Hyun-Jung Kim (Korea Advanced Institute of Science and 20903 Technology) and Hongyeon Kim, Hyeonjeong Choi, and Dain 20904 Ham (Wonik IPS) 20905 </div> 20906 <div class="slot-abstract"> 20907 <div> 20908 <a 20909 class="clickable no-decoration" 20910 id="vhsjs_view_464_1707793552_2557478" 20911 onclick="$('#vhsjs_view_464_1707793552_2557478').hide(); 20912 $('#vhsjs_hide_464_1707793552_2557478').show(); 20913 $('#463_1707793552_2557395').slideDown(function() { 20914 if (typeof Masonry === 'function') { 20915 $('.use_masonry').masonry(); 20916 }; 20917 20918 });" 20919 ><i class="fa fa-caret-right"></i> 20920 <span class="hover_link">Abstract</span></a 20921 ><a 20922 class="clickable no-decoration" 20923 id="vhsjs_hide_464_1707793552_2557478" 20924 onclick="$('#463_1707793552_2557395').hide(function() { 20925 if (typeof Masonry === 'function') { 20926 $('.use_masonry').masonry(); 20927 }; 20928 }); 20929 $('#vhsjs_hide_464_1707793552_2557478').hide(); 20930 $('#vhsjs_view_464_1707793552_2557478').show();" 20931 style="display: none" 20932 ><i class="fa fa-caret-down"></i> 20933 <span class="hover_link">Abstract</span></a 20934 > 20935 <div 20936 data-display-control="464_1707793552_2557478" 20937 id="463_1707793552_2557395" 20938 style="display: none" 20939 > 20940 <div class="arrow-slidedown"> 20941 <blockquote> 20942 Monitoring the state of semiconductor equipment 20943 is crucial for ensuring optimal performance and 20944 preventing downtime. In previous studies, 20945 researchers have attempted to derive a health 20946 index that represents the overall condition of 20947 the equipment as a single index. However, these 20948 studies have often relied solely on time-series 20949 data from each sensor, neglecting other 20950 important viewpoints engineers consider when 20951 monitoring the equipment. To address this 20952 limitation, we propose a multi-view data set 20953 specifically designed for semiconductor 20954 equipment, which incorporates process, trend, 20955 and spatial data. In addition, we present a 20956 framework for deriving a hierarchical health 20957 index based on a multi-view data set. The 20958 hierarchical structure is derived using a 20959 hierarchical spectral clustering method, and an 20960 autoencoder-based health index is used. We have 20961 verified the effectiveness of our approach with 20962 real data sets, demonstrating its potential as a 20963 valuable tool for monitoring the condition of 20964 semiconductor equipment. 20965 </blockquote> 20966 </div> 20967 </div> 20968 </div> 20969 </div> 20970 <div class="slot-urls"></div> 20971 <a href="/wsc23papers/194.pdf" target="_blank">pdf</a 20972 ><br /> 20973 </div> 20974 <div class="slot-entry"> 20975 <a name="con122" tabindex="-1"></a> 20976 <div class="slot-title-line">
20977 <span class="slot-title" 20978 >Modeling Multivariate Relations in Multiblock 20979 Semiconductor Manufacturing Data Using Process PLS to 20980 Enhance Process Understanding</span 20981 > 20982 </div> 20983 <div class="slot-authors"> 20984 Geert van Kollenburg and Richard Verhoeven (Eindhoven 20985 University of Technology), Daniele Pagano 20986 (STMicroelectronics s.r.l.), and Mike Holenderski and 20987 Nirvana Meratnia (Eindhoven University of Technology) 20988 </div> 20989 <div class="slot-abstract"> 20990 <div> 20991 <a 20992 class="clickable no-decoration" 20993 id="vhsjs_view_466_1707793552_2581372" 20994 onclick="$('#vhsjs_view_466_1707793552_2581372').hide(); 20995 $('#vhsjs_hide_466_1707793552_2581372').show(); 20996 $('#465_1707793552_258129').slideDown(function() { 20997 if (typeof Masonry === 'function') { 20998 $('.use_masonry').masonry(); 20999 }; 21000 21001 });" 21002 ><i class="fa fa-caret-right"></i> 21003 <span class="hover_link">Abstract</span></a 21004 ><a 21005 class="clickable no-decoration" 21006 id="vhsjs_hide_466_1707793552_2581372" 21007 onclick="$('#465_1707793552_258129').hide(function() { 21008 if (typeof Masonry === 'function') { 21009 $('.use_masonry').masonry(); 21010 }; 21011 }); 21012 $('#vhsjs_hide_466_1707793552_2581372').hide(); 21013 $('#vhsjs_view_466_1707793552_2581372').show();" 21014 style="display: none" 21015 ><i class="fa fa-caret-down"></i> 21016 <span class="hover_link">Abstract</span></a 21017 > 21018 <div 21019 data-display-control="466_1707793552_2581372" 21020 id="465_1707793552_258129" 21021 style="display: none" 21022 > 21023 <div class="arrow-slidedown"> 21024 <blockquote> 21025 The complexity of manufacturing process data has 21026 made it more challenging to extract useful 21027 insights. Data-analytic solutions have therefore 21028 become essential for analyzing and optimizing 21029 manufacturing processes. Path modeling, also 21030 known as structural equation modeling, is a 21031 statistical approach that can provide new 21032 insights into complex multivariate relationships 21033 between process variables from different stages 21034 of the manufacturing process. The incorporation 21035 of expert process knowledge and subsequent 21036 interpretation of model results can facilitate 21037 communication between stakeholders, promoting 21038 lean manufacturing and achieving the 21039 sustainability goals of Industry 5.0. This paper 21040 describes the use of a path modeling algorithm 21041 called Process Partial Least Squares (Process 21042 PLS) to gain new insights into the relationships 21043 between equipment data from several machines 21044 within the semiconductor manufacturing process. 21045 The methods used in this study can assist 21046 manufacturers in understanding the relations 21047 between different machines and identify the most 21048 influential variables that may be used to 21049 develop soft-sensors. 21050 </blockquote> 21051 </div> 21052 </div> 21053 </div> 21054 </div> 21055 <div class="slot-urls"></div> 21056 <a href="/wsc23papers/195.pdf" target="_blank">pdf</a 21057 ><br /> 21058 </div> 21059 <div class="slot-entry"> 21060 <a name="con164" tabindex="-1"></a> 21061 <div class="slot-title-line">
21062 <span class="slot-title" 21063 >Multi-Resolution Modeling Method for Automated 21064 Material Handling System Systems in Semiconductor 21065 FABs</span 21066 > 21067 </div> 21068 <div class="slot-authors"> 21069 Kwanwoo Lee, Woosung Jeon, and Sangchul Park (Ajou 21070 University) 21071 </div> 21072 <div class="slot-abstract"> 21073 <div> 21074 <a 21075 class="clickable no-decoration" 21076 id="vhsjs_view_468_1707793552_2604165" 21077 onclick="$('#vhsjs_view_468_1707793552_2604165').hide(); 21078 $('#vhsjs_hide_468_1707793552_2604165').show(); 21079 $('#467_1707793552_2604082').slideDown(function() { 21080 if (typeof Masonry === 'function') { 21081 $('.use_masonry').masonry(); 21082 }; 21083 21084 });" 21085 ><i class="fa fa-caret-right"></i> 21086 <span class="hover_link">Abstract</span></a 21087 ><a 21088 class="clickable no-decoration" 21089 id="vhsjs_hide_468_1707793552_2604165" 21090 onclick="$('#467_1707793552_2604082').hide(function() { 21091 if (typeof Masonry === 'function') { 21092 $('.use_masonry').masonry(); 21093 }; 21094 }); 21095 $('#vhsjs_hide_468_1707793552_2604165').hide(); 21096 $('#vhsjs_view_468_1707793552_2604165').show();" 21097 style="display: none" 21098 ><i class="fa fa-caret-down"></i> 21099 <span class="hover_link">Abstract</span></a 21100 > 21101 <div 21102 data-display-control="468_1707793552_2604165" 21103 id="467_1707793552_2604082" 21104 style="display: none" 21105 > 21106 <div class="arrow-slidedown"> 21107 <blockquote> 21108 This paper presents a novel modeling framework 21109 for semiconductor fabrication facilities (FABs) 21110 that integrates production and material handling 21111 systems. Because the productivity of 21112 semiconductor FABs is significantly influenced 21113 by their material-handling systems, existing 21114 research has focused on optimizing operational 21115 logic considering both aspects. However, the 21116 scale and complexity of modern FABs make 21117 implementation of fully integrated models 21118 challenging, resulting in slow simulation speeds 21119 for long periods. To address this issue, we 21120 propose a multi-resolution modeling framework 21121 that creates material-handling system models at 21122 two distinct resolution levels, enabling fast, 21123 fully integrated FAB models while accounting for 21124 material-handling effects. Experimental results 21125 demonstrated accelerated simulation completion 21126 compared to single-resolution models while 21127 maintaining consistent results. The proposed 21128 method provides a practical approach for 21129 semiconductor FABs to investigate long-term 21130 phenomena and urgent decision-making problems 21131 while considering both production and 21132 material-handling systems. 21133 </blockquote> 21134 </div> 21135 </div> 21136 </div> 21137 </div> 21138 <div class="slot-urls"></div> 21139 <a href="/wsc23papers/196.pdf" target="_blank">pdf</a 21140 ><br /> 21141 </div> 21142 </div> 21143 <div class="session-entry">
21144 <span class="session-event-type">Technical Session</span 21145 ><span class="type-track-spacer"> · </span 21146 ><span class="program-track" 21147 >MASM: Semiconductor Manufacturing</span 21148 ><br /> 21149 <div class="session-title"> 21150 Machine Learning Applications 21151 </div> 21152 <div class="session-chair"> 21153 Chair: John Fowler (Arizona State University)<br /> 21154 </div> 21155 <div class="slot-entry"> 21156 <a name="cea132" tabindex="-1"></a> 21157 <div class="slot-title-line"> 21158 <span class="slot-title" 21159 >A Self-supervised Learning Based Framework for 21160 TFT-LCD Defect Classification</span 21161 > 21162 </div> 21163 <div class="slot-authors"> 21164 Sheng-Xiang Kao (International Intercollegiate Ph.D. 21165 Program, National Tsing Hua University); Yu-Hsun Lin 21166 (Department of Industrial Engineering and Engineering 21167 Management, National Tsing Hua University); and Chen-Fu 21168 Chien (Intelligent Manufacturing and Circular Economy 21169 Research Center, National Tsing Hua University) 21170 </div> 21171 <div class="slot-abstract"> 21172 <div> 21173 <a 21174 class="clickable no-decoration" 21175 id="vhsjs_view_470_1707793552_265371" 21176 onclick="$('#vhsjs_view_470_1707793552_265371').hide(); 21177 $('#vhsjs_hide_470_1707793552_265371').show(); 21178 $('#469_1707793552_265363').slideDown(function() { 21179 if (typeof Masonry === 'function') { 21180 $('.use_masonry').masonry(); 21181 }; 21182 21183 });" 21184 ><i class="fa fa-caret-right"></i> 21185 <span class="hover_link">Abstract</span></a 21186 ><a 21187 class="clickable no-decoration" 21188 id="vhsjs_hide_470_1707793552_265371" 21189 onclick="$('#469_1707793552_265363').hide(function() { 21190 if (typeof Masonry === 'function') { 21191 $('.use_masonry').masonry(); 21192 }; 21193 }); 21194 $('#vhsjs_hide_470_1707793552_265371').hide(); 21195 $('#vhsjs_view_470_1707793552_265371').show();" 21196 style="display: none" 21197 ><i class="fa fa-caret-down"></i> 21198 <span class="hover_link">Abstract</span></a 21199 > 21200 <div 21201 data-display-control="470_1707793552_265371" 21202 id="469_1707793552_265363" 21203 style="display: none" 21204 > 21205 <div class="arrow-slidedown"> 21206 <blockquote> 21207 This study presents a self-supervised learning 21208 based framework for TFT-LCD defect 21209 classification in semiconductor smart 21210 manufacturing. Utilizing the Swapping 21211 Assignments between Views (SwAV) model trained 21212 on 1,000,000 unlabeled TFT-LCD images, the 21213 framework achieves an overall top-1 accuracy of 21214 0.709 and precision of 0.7812 in downstream task 21215 of classifying 13 types of TFT-LCD defects. 21216 Compared to using SwAV pre-trained weighs on 21217 ImageNet, proposed domain-specific
21218 self-supervised learning model significantly 21219 outperforms, emphasizing the importance of 21220 domain-specific training. The framework offers 21221 manufacturers a cost-efficient decision support 21222 system, enhancing TFT-LCD defect classification 21223 quality. 21224 </blockquote> 21225 </div> 21226 </div> 21227 </div> 21228 </div> 21229 <div class="slot-urls"></div> 21230 <a href="/wsc23papers/cea132.pdf" target="_blank">pdf</a 21231 ><br /> 21232 </div> 21233 <div class="slot-entry"> 21234 <a name="cea128" tabindex="-1"></a> 21235 <div class="slot-title-line"> 21236 <span class="slot-title" 21237 >Root Cause Analysis in Supply Chain Planning Using 21238 Explainable Machine Learning</span 21239 > 21240 </div> 21241 <div class="slot-authors"> 21242 Pavle Kecman, Josephine Fang, and Ana Glaser (NXP 21243 Semiconductors) 21244 </div> 21245 <div class="slot-abstract"> 21246 <div> 21247 <a 21248 class="clickable no-decoration" 21249 id="vhsjs_view_472_1707793552_2674596" 21250 onclick="$('#vhsjs_view_472_1707793552_2674596').hide(); 21251 $('#vhsjs_hide_472_1707793552_2674596').show(); 21252 $('#471_1707793552_2674518').slideDown(function() { 21253 if (typeof Masonry === 'function') { 21254 $('.use_masonry').masonry(); 21255 }; 21256 21257 });" 21258 ><i class="fa fa-caret-right"></i> 21259 <span class="hover_link">Abstract</span></a 21260 ><a 21261 class="clickable no-decoration" 21262 id="vhsjs_hide_472_1707793552_2674596" 21263 onclick="$('#471_1707793552_2674518').hide(function() { 21264 if (typeof Masonry === 'function') { 21265 $('.use_masonry').masonry(); 21266 }; 21267 }); 21268 $('#vhsjs_hide_472_1707793552_2674596').hide(); 21269 $('#vhsjs_view_472_1707793552_2674596').show();" 21270 style="display: none" 21271 ><i class="fa fa-caret-down"></i> 21272 <span class="hover_link">Abstract</span></a 21273 > 21274 <div 21275 data-display-control="472_1707793552_2674596" 21276 id="471_1707793552_2674518" 21277 style="display: none" 21278 > 21279 <div class="arrow-slidedown"> 21280 <blockquote> 21281 In the highly dynamic world of semiconductor 21282 manufacturing, planning analysts are asked to 21283 analyze variations between weekly production 21284 plans with the goal of identifying a resolution 21285 in a landscape involving elaborate optimization 21286 models with significant interdependence between 21287 data elements. We propose a solution to 21288 effectively analyze the weekly planning engine 21289 output and identify the data elements with 21290 significant contribution to the outcome. An 21291 explainable Machine Learning model is trained 21292 and deployed to simulate the behavior of the 21293 planning engine. Each model execution can be 21294 explained to identify the features with the most 21295 significant contribution to prediction. The 21296 resulting application contributes to a timely 21297 resolution to the production plan deviation, 21298 while generating significant productivity gains. 21299 </blockquote> 21300 </div> 21301 </div> 21302 </div> 21303 </div> 21304 <div class="slot-urls"></div> 21305 <a href="/wsc23papers/cea128.pdf" target="_blank">pdf</a 21306 ><br /> 21307 </div> 21308 <div class="slot-entry"> 21309 <a name="cea111" tabindex="-1"></a> 21310 <div class="slot-title-line">
21311 <span class="slot-title" 21312 >Scaling Deep Reinforcement Learning for Queue-time 21313 Management in Semiconductor Manufacturing</span 21314 > 21315 </div> 21316 <div class="slot-authors"> 21317 Harel Yedidsion, Prafulla Dawadi, David Norman, and 21318 Emrah Zarifoglu (Applied Materials) 21319 </div> 21320 <div class="slot-abstract"> 21321 <div> 21322 <a 21323 class="clickable no-decoration" 21324 id="vhsjs_view_474_1707793552_2697895" 21325 onclick="$('#vhsjs_view_474_1707793552_2697895').hide(); 21326 $('#vhsjs_hide_474_1707793552_2697895').show(); 21327 $('#473_1707793552_2697814').slideDown(function() { 21328 if (typeof Masonry === 'function') { 21329 $('.use_masonry').masonry(); 21330 }; 21331 21332 });" 21333 ><i class="fa fa-caret-right"></i> 21334 <span class="hover_link">Abstract</span></a 21335 ><a 21336 class="clickable no-decoration" 21337 id="vhsjs_hide_474_1707793552_2697895" 21338 onclick="$('#473_1707793552_2697814').hide(function() { 21339 if (typeof Masonry === 'function') { 21340 $('.use_masonry').masonry(); 21341 }; 21342 }); 21343 $('#vhsjs_hide_474_1707793552_2697895').hide(); 21344 $('#vhsjs_view_474_1707793552_2697895').show();" 21345 style="display: none" 21346 ><i class="fa fa-caret-down"></i> 21347 <span class="hover_link">Abstract</span></a 21348 > 21349 <div 21350 data-display-control="474_1707793552_2697895" 21351 id="473_1707793552_2697814" 21352 style="display: none" 21353 > 21354 <div class="arrow-slidedown"> 21355 <blockquote> 21356 Queue-Time Constraints (QTCs) set a maximum 21357 waiting time for lots between consecutive 21358 process steps. In semiconductor manufacturing, 21359 exceeding these limits results in yield loss, 21360 rework, or scrapping. Managing QTCs is 21361 challenging due to the need for lots to wait 21362 until there is available capacity for the final 21363 step. Specifically, accurately calculating the 21364 capacity is computationally expensive, making it 21365 difficult to handle large instances. Our 21366 research addresses the scalability of QTC 21367 management in real fabs with numerous 21368 constraints. We propose a deep Reinforcement 21369 Learning (RL) solution to handle lot release 21370 into the QTC. We describe the infrastructure 21371 developed for RL training using actual fab data, 21372 assess the performance of our RL approach, and 21373 compare it to three baseline solutions. Our 21374 empirical evaluation demonstrates that the RL 21375 method surpasses the baselines in key 21376 performance metrics including queue-time 21377 violations, while requiring negligible online 21378 compute time. 21379 </blockquote> 21380 </div> 21381 </div> 21382 </div> 21383 </div> 21384 <div class="slot-urls"></div> 21385 <a href="/wsc23papers/cea111.pdf" target="_blank">pdf</a 21386 ><br /> 21387 </div> 21388 </div> 21389 </div> 21390 <div class="centered"> 21391 <div class="top-link"><a href="#top">Return to Top</a></div> 21392 </div> 21393 <hr /> 21394 </div> 21395 <div class="area-section"> 21396 <div class="centered"> 21397 <a name="ptrack108" tabindex="-1"></a> 21398 <div class="section-title"> 21399 Military and National Security Applications 21400 </div> 21401 </div> 21402 <div class="centered track-chair">
21403 <span class="track-chair-role" 21404 >Track Coordinator - Military and National Security 21405 Applications: </span 21406 ><span class="track-chair-names" 21407 >Clay Koschnick (Air Force Institute of Technology), James 21408 Starling (U.S. Military Academy)</span 21409 > 21410 </div> 21411 <div class="section-entry"> 21412 <div class="session-entry"> 21413 <span class="session-event-type">Technical Session</span 21414 ><span class="type-track-spacer"> · </span 21415 ><span class="program-track" 21416 >Military and National Security Applications</span 21417 ><br /> 21418 <div class="session-title"> 21419 Military Keynote: Creating Live Virtual Constructive 21420 Environments to Evaluate Human and System Resilience 21421 </div> 21422 <div class="session-chair"> 21423 Chair: James Starling (U.S. Military Academy)<br /> 21424 </div> 21425 <div class="slot-entry"> 21426 <a name="prog106" tabindex="-1"></a> 21427 <div class="slot-title-line"> 21428 <span class="slot-title" 21429 >Creating Live Virtual Constructive Environments to 21430 Evaluate Human and System Resilience</span 21431 > 21432 </div> 21433 <div class="slot-authors"> 21434 Imre Balogh (Naval Postgraduate School) 21435 </div> 21436 <div class="slot-abstract"> 21437 <div> 21438 <a 21439 class="clickable no-decoration" 21440 id="vhsjs_view_476_1707793552_2762892" 21441 onclick="$('#vhsjs_view_476_1707793552_2762892').hide(); 21442 $('#vhsjs_hide_476_1707793552_2762892').show(); 21443 $('#475_1707793552_2762804').slideDown(function() { 21444 if (typeof Masonry === 'function') { 21445 $('.use_masonry').masonry(); 21446 }; 21447 21448 });" 21449 ><i class="fa fa-caret-right"></i> 21450 <span class="hover_link">Abstract</span></a 21451 ><a 21452 class="clickable no-decoration" 21453 id="vhsjs_hide_476_1707793552_2762892" 21454 onclick="$('#475_1707793552_2762804').hide(function() { 21455 if (typeof Masonry === 'function') { 21456 $('.use_masonry').masonry(); 21457 }; 21458 }); 21459 $('#vhsjs_hide_476_1707793552_2762892').hide(); 21460 $('#vhsjs_view_476_1707793552_2762892').show();" 21461 style="display: none" 21462 ><i class="fa fa-caret-down"></i> 21463 <span class="hover_link">Abstract</span></a 21464 > 21465 <div 21466 data-display-control="476_1707793552_2762892" 21467 id="475_1707793552_2762804" 21468 style="display: none" 21469 > 21470 <div class="arrow-slidedown"> 21471 <blockquote> 21472 Live Virtual Constructive (LVC) exercises are 21473 becoming ubiquitous for training and mission 21474 rehearsal in the military domain because the use 21475 of LVC provides the most realistic environment 21476 available short of actual military operations. 21477 The mixture of live exercises with simulated 21478 components (constructive simulations and virtual 21479 simulators) allows for the creation of a context 21480 for the training or rehearsal that is richer and 21481 more representative of the real world than would 21482 be possible with only live events. This ability 21483 to embed live activity into synthetic 21484 environment to provide realism has attracted the 21485 interest of the Test and Evaluation community 21486 (T&E) and recently there are increasing efforts 21487 to start including LVC in the T&E tool suite. 21488 This talk will discuss some of the work we have 21489 been doing at the Naval Postgraduate School with 21490 LVC and how these environments can be used to 21491 assess and improve system and human resilience 21492 in operational environments. 21493 </blockquote> 21494 </div> 21495 </div> 21496 </div> 21497 </div> 21498 <div class="slot-urls"></div> 21499 <a href="/wsc23papers/prog106.pdf" target="_blank">pdf</a 21500 ><br /> 21501 </div> 21502 </div> 21503 <div class="session-entry">
21504 <span class="session-event-type">Technical Session</span 21505 ><span class="type-track-spacer"> · </span 21506 ><span class="program-track" 21507 >Military and National Security Applications</span 21508 ><br /> 21509 <div class="session-title"> 21510 Enhancing Military Decision-Making: Strategies for Success 21511 </div> 21512 <div class="session-chair"> 21513 Chair: Mehdi Benhassine (Royal Military Academy)<br /> 21514 </div> 21515 <div class="slot-entry"> 21516 <a name="con213" tabindex="-1"></a> 21517 <div class="slot-title-line"> 21518 <span class="slot-title" 21519 >Incorporation of Military Doctrines and Objectives 21520 into an AI Agent via Natural Language and Reward in 21521 Reinforcement Learning</span 21522 > 21523 </div> 21524 <div class="slot-authors"> 21525 Michael Möbius, Daniel Kallfass, and Matthias Flock 21526 (Airbus Defence and Space GmbH) and Thomas Doll and 21527 Dietmar Kunde (German Armed Forces) 21528 </div> 21529 <div class="slot-abstract"> 21530 <div> 21531 <a 21532 class="clickable no-decoration" 21533 id="vhsjs_view_478_1707793552_2809439" 21534 onclick="$('#vhsjs_view_478_1707793552_2809439').hide(); 21535 $('#vhsjs_hide_478_1707793552_2809439').show(); 21536 $('#477_1707793552_280936').slideDown(function() { 21537 if (typeof Masonry === 'function') { 21538 $('.use_masonry').masonry(); 21539 }; 21540 21541 });" 21542 ><i class="fa fa-caret-right"></i> 21543 <span class="hover_link">Abstract</span></a 21544 ><a 21545 class="clickable no-decoration" 21546 id="vhsjs_hide_478_1707793552_2809439" 21547 onclick="$('#477_1707793552_280936').hide(function() { 21548 if (typeof Masonry === 'function') { 21549 $('.use_masonry').masonry(); 21550 }; 21551 }); 21552 $('#vhsjs_hide_478_1707793552_2809439').hide(); 21553 $('#vhsjs_view_478_1707793552_2809439').show();" 21554 style="display: none" 21555 ><i class="fa fa-caret-down"></i> 21556 <span class="hover_link">Abstract</span></a 21557 > 21558 <div 21559 data-display-control="478_1707793552_2809439" 21560 id="477_1707793552_280936" 21561 style="display: none" 21562 > 21563 <div class="arrow-slidedown"> 21564 <blockquote> 21565 This paper emphasizes the integration of sound 21566 tactical behavior in the generation of realistic 21567 military simulations, which includes the 21568 definition of combat tactics, doctrine, rules of 21569 engagement, and concepts of operations. Recent 21570 advances in reinforcement learning (RL) enable 21571 RL agents to generate a wide range of tactical 21572 actions. A multi-agent ground combat scenario is 21573 used in this paper to demonstrate how a machine 21574 learning (ML) application generates strategies 21575 and issues commands while following a given 21576 objective. Natural language is used to issue 21577 doctrines and objectives to improve 21578 communication between the human advisor and the 21579 ML agent. This allows us to embed objectives and 21580 existing doctrines into the reasoning of an 21581 artificial intelligence (AI). The research 21582 demonstrates the successful integration of 21583 natural language to enable an agent to achieve 21584 different objectives. This groundwork will 21585 enhance RL agents' ability in the future to 21586 uphold the doctrines and rules of military 21587 operations. 21588 </blockquote> 21589 </div> 21590 </div> 21591 </div> 21592 </div> 21593 <div class="slot-urls"></div> 21594 <a href="/wsc23papers/197.pdf" target="_blank">pdf</a 21595 ><br /> 21596 </div> 21597 <div class="slot-entry"> 21598 <a name="con243" tabindex="-1"></a> 21599 <div class="slot-title-line">
21600 <span class="slot-title" 21601 >Accounting for Individual Shooting Skills in Combat 21602 Models</span 21603 > 21604 </div> 21605 <div class="slot-authors"> 21606 Vikram Mittal and Paul F. Evangelista (United States 21607 Military Academy) 21608 </div> 21609 <div class="slot-abstract"> 21610 <div> 21611 <a 21612 class="clickable no-decoration" 21613 id="vhsjs_view_480_1707793552_2831216" 21614 onclick="$('#vhsjs_view_480_1707793552_2831216').hide(); 21615 $('#vhsjs_hide_480_1707793552_2831216').show(); 21616 $('#479_1707793552_2831135').slideDown(function() { 21617 if (typeof Masonry === 'function') { 21618 $('.use_masonry').masonry(); 21619 }; 21620 21621 });" 21622 ><i class="fa fa-caret-right"></i> 21623 <span class="hover_link">Abstract</span></a 21624 ><a 21625 class="clickable no-decoration" 21626 id="vhsjs_hide_480_1707793552_2831216" 21627 onclick="$('#479_1707793552_2831135').hide(function() { 21628 if (typeof Masonry === 'function') { 21629 $('.use_masonry').masonry(); 21630 }; 21631 }); 21632 $('#vhsjs_hide_480_1707793552_2831216').hide(); 21633 $('#vhsjs_view_480_1707793552_2831216').show();" 21634 style="display: none" 21635 ><i class="fa fa-caret-down"></i> 21636 <span class="hover_link">Abstract</span></a 21637 > 21638 <div 21639 data-display-control="480_1707793552_2831216" 21640 id="479_1707793552_2831135" 21641 style="display: none" 21642 > 21643 <div class="arrow-slidedown"> 21644 <blockquote> 21645 There is significant variation in shooting 21646 ability among U.S. Army soldiers, which is often 21647 overlooked in combat simulations. This study 21648 introduces a Monte-Carlo model to estimate the 21649 dispersion of a soldier's shot group based on 21650 their marksmanship score. This model is used to 21651 assess the impact of marksmanship on a squad's 21652 performance through two analyses. The first 21653 analysis employs a dueling model to examine 21654 various marksmanship skills between dueling 21655 teams, offering insights into overmatch 21656 requirements. The second analysis uses an 21657 agent-based combat simulation to investigate the 21658 influence of marksmanship on squad performance 21659 in a dueling scenario in addition to tactical 21660 rural and urban missions. The results reveal 21661 that marksmanship becomes increasingly crucial 21662 in enhancing lethality and survivability as the 21663 distance between combatants grows. Notably, 21664 superior marksmanship skills are particularly 21665 vital in offensive, rural operations. These 21666 findings emphasize the significance of 21667 marksmanship and its implications for military 21668 requirements and tactical decision-making. 21669 </blockquote> 21670 </div> 21671 </div> 21672 </div> 21673 </div> 21674 <div class="slot-urls"></div> 21675 <a href="/wsc23papers/199.pdf" target="_blank">pdf</a 21676 ><br /> 21677 </div> 21678 </div> 21679 <div class="session-entry">
21680 <span class="session-event-type">Technical Session</span 21681 ><span class="type-track-spacer"> · </span 21682 ><span class="program-track" 21683 >Military and National Security Applications</span 21684 ><br /> 21685 <div class="session-title"> 21686 Protection: Modeling Mass Casualty Incidents 21687 </div> 21688 <div class="session-chair"> 21689 Chair: David Beskow (United States Military Academy)<br /> 21690 </div> 21691 <div class="slot-entry"> 21692 <a name="con120" tabindex="-1"></a> 21693 <div class="slot-title-line"> 21694 <span class="slot-title" 21695 >Open-Air Artillery Strike in a Rural Area: A 21696 Hypothetical Scenario</span 21697 > 21698 </div> 21699 <div class="slot-authors"> 21700 Mehdi Benhassine (Royal Military Academy); Ruben De 21701 Rouck, Michel Debacker, and Ives Hubloue (Vrije 21702 Universiteit Brussel); Erwin Dhondt (DO Consultancy); 21703 John Quinn (Charles University); and Filip Van 21704 Utterbeeck (Royal Military Academy) 21705 </div> 21706 <div class="slot-abstract"> 21707 <div> 21708 <a 21709 class="clickable no-decoration" 21710 id="vhsjs_view_482_1707793552_2879508" 21711 onclick="$('#vhsjs_view_482_1707793552_2879508').hide(); 21712 $('#vhsjs_hide_482_1707793552_2879508').show(); 21713 $('#481_1707793552_2879424').slideDown(function() { 21714 if (typeof Masonry === 'function') { 21715 $('.use_masonry').masonry(); 21716 }; 21717 21718 });" 21719 ><i class="fa fa-caret-right"></i> 21720 <span class="hover_link">Abstract</span></a 21721 ><a 21722 class="clickable no-decoration" 21723 id="vhsjs_hide_482_1707793552_2879508" 21724 onclick="$('#481_1707793552_2879424').hide(function() { 21725 if (typeof Masonry === 'function') { 21726 $('.use_masonry').masonry(); 21727 }; 21728 }); 21729 $('#vhsjs_hide_482_1707793552_2879508').hide(); 21730 $('#vhsjs_view_482_1707793552_2879508').show();" 21731 style="display: none" 21732 ><i class="fa fa-caret-down"></i> 21733 <span class="hover_link">Abstract</span></a 21734 > 21735 <div 21736 data-display-control="482_1707793552_2879508" 21737 id="481_1707793552_2879424" 21738 style="display: none" 21739 > 21740 <div class="arrow-slidedown"> 21741 <blockquote> 21742 The escalation of the Russian invasion in 21743 Ukraine, characterized by the deployment of 21744 conventional weapon systems, inflicts 21745 significant morbidity and mortality on the 21746 victims. It is imperative to ascertain optimal 21747 medical practices and disaster response 21748 strategies throughout the battlefield to 21749 minimize casualties and safeguard the well-being 21750 of medical and disaster responders. The 21751 challenges posed by large-scale battlefield 21752 threats can rapidly overwhelm healthcare 21753 providers due to the sheer number of victims, 21754 which can result in the depletion of medical 21755 supplies and insufficient training and 21756 resources. To address these issues, we utilized 21757 the SIMEDIS simulator to establish and implement 21758 a battlefield scenario involving an open-air 21759 artillery strike in a field. Mortality rates 21760 were calculated based on the application of 21761 bleeding control measures and the distribution 21762 policy for allocating victims to medical 21763 treatment facilities. Controlling hemorrhage 21764 remains the most crucial factor influencing 21765 mortality outcomes. 21766 </blockquote> 21767 </div> 21768 </div> 21769 </div> 21770 </div> 21771 <div class="slot-urls"></div> 21772 <a href="/wsc23papers/200.pdf" target="_blank">pdf</a 21773 ><br /> 21774 </div> 21775 <div class="slot-entry"> 21776 <a name="con234" tabindex="-1"></a> 21777 <div class="slot-title-line">
21778 <span class="slot-title" 21779 >A Modular Simulation Model for Mass Casualty 21780 Incidents</span 21781 > 21782 </div> 21783 <div class="slot-authors"> 21784 Kai Meisner (Bundeswehr Medical Academy, University of 21785 the Bundeswehr Munich) and Heiderose Stein, Nadiia 21786 Leopold, Tobias Uhlig, and Oliver Rose (University of 21787 the Bundeswehr Munich) 21788 </div> 21789 <div class="slot-abstract"> 21790 <div> 21791 <a 21792 class="clickable no-decoration" 21793 id="vhsjs_view_484_1707793552_290349" 21794 onclick="$('#vhsjs_view_484_1707793552_290349').hide(); 21795 $('#vhsjs_hide_484_1707793552_290349').show(); 21796 $('#483_1707793552_2903407').slideDown(function() { 21797 if (typeof Masonry === 'function') { 21798 $('.use_masonry').masonry(); 21799 }; 21800 21801 });" 21802 ><i class="fa fa-caret-right"></i> 21803 <span class="hover_link">Abstract</span></a 21804 ><a 21805 class="clickable no-decoration" 21806 id="vhsjs_hide_484_1707793552_290349" 21807 onclick="$('#483_1707793552_2903407').hide(function() { 21808 if (typeof Masonry === 'function') { 21809 $('.use_masonry').masonry(); 21810 }; 21811 }); 21812 $('#vhsjs_hide_484_1707793552_290349').hide(); 21813 $('#vhsjs_view_484_1707793552_290349').show();" 21814 style="display: none" 21815 ><i class="fa fa-caret-down"></i> 21816 <span class="hover_link">Abstract</span></a 21817 > 21818 <div 21819 data-display-control="484_1707793552_290349" 21820 id="483_1707793552_2903407" 21821 style="display: none" 21822 > 21823 <div class="arrow-slidedown"> 21824 <blockquote> 21825 During military conflicts, the number of 21826 casualties is likely to exceed medical 21827 capabilities. For best treatment results, the 21828 patients must be distributed according to their 21829 needs to the available resources such as medical 21830 facilities and means of transportation. Computer 21831 simulations are used to verify and optimize 21832 current medical planning. However, recent models 21833 lack the capability of testing a wide range of 21834 decision rules. In this paper, we address this 21835 issue and propose a modular simulation concept 21836 whose components can be adapted and exchanged 21837 independently. Using modular submodels to 21838 control the simulated objects, we enable the 21839 implementation of a wide range of object 21840 behavior. A prototype implementation of the 21841 proposed concept is presented, showing the 21842 effects of applying different dispatching rules 21843 in an evacuation scenario. 21844 </blockquote> 21845 </div> 21846 </div> 21847 </div> 21848 </div> 21849 <div class="slot-urls"></div> 21850 <a href="/wsc23papers/201.pdf" target="_blank">pdf</a 21851 ><br /> 21852 </div> 21853 </div> 21854 <div class="session-entry">
21855 <span class="session-event-type">Technical Session</span 21856 ><span class="type-track-spacer"> · </span 21857 ><span class="program-track" 21858 >Military and National Security Applications</span 21859 ><br /> 21860 <div class="session-title"> 21861 Optimizing Aerial Operations: Advancements in Air Mission 21862 Planning 21863 </div> 21864 <div class="session-chair"> 21865 Chair: Nicholas Shallcross (U.S. Army, University of 21866 Arkansas)<br /> 21867 </div> 21868 <div class="slot-entry"> 21869 <a name="con131" tabindex="-1"></a> 21870 <div class="slot-title-line"> 21871 <span class="slot-title" 21872 >Implementing Efficient Dynamic Threat Avoidance 21873 Routing Based on Dijkstra's Shortest Path Algorithm in 21874 the Advanced Framework for Simulation, Integration, 21875 and Modeling (AFSIM)</span 21876 > 21877 </div> 21878 <div class="slot-authors"> 21879 Dante Reid, Lance Champagne, and Nathan Gaw (Air Force 21880 Institute of Technology) 21881 </div> 21882 <div class="slot-abstract"> 21883 <div> 21884 <a 21885 class="clickable no-decoration" 21886 id="vhsjs_view_486_1707793552_2967057" 21887 onclick="$('#vhsjs_view_486_1707793552_2967057').hide(); 21888 $('#vhsjs_hide_486_1707793552_2967057').show(); 21889 $('#485_1707793552_2966974').slideDown(function() { 21890 if (typeof Masonry === 'function') { 21891 $('.use_masonry').masonry(); 21892 }; 21893 21894 });" 21895 ><i class="fa fa-caret-right"></i> 21896 <span class="hover_link">Abstract</span></a 21897 ><a 21898 class="clickable no-decoration" 21899 id="vhsjs_hide_486_1707793552_2967057" 21900 onclick="$('#485_1707793552_2966974').hide(function() { 21901 if (typeof Masonry === 'function') { 21902 $('.use_masonry').masonry(); 21903 }; 21904 }); 21905 $('#vhsjs_hide_486_1707793552_2967057').hide(); 21906 $('#vhsjs_view_486_1707793552_2967057').show();" 21907 style="display: none" 21908 ><i class="fa fa-caret-down"></i> 21909 <span class="hover_link">Abstract</span></a 21910 > 21911 <div 21912 data-display-control="486_1707793552_2967057" 21913 id="485_1707793552_2966974" 21914 style="display: none" 21915 > 21916 <div class="arrow-slidedown"> 21917 <blockquote> 21918 Simulating pre-planned routes and dynamic threat 21919 avoidance routing represents a significant 21920 problem for operations analysts. Without methods 21921 to create operationally valid routes through 21922 automation, the analyst is generally faced with 21923 hard coding individual routes for multiple 21924 aircraft over the entirety of the mission set. 21925 This research developed, implemented, and 21926 analyzed threat avoidance routing based on 21927 Dijkstra's algorithm for aircraft attempting to 21928 operate in an anti-access area denial (A2AD) 21929 environment capable of dynamically updating the 21930 mission route as new threat information is 21931 learned. A designed experiment was conducted to 21932 determine the impact of grid parameters on 21933 operational effectiveness metrics and 21934 computational costs. Statistical analysis 21935 results show that the proposed algorithm 21936 produced the best operational performance with 21937 grid spacing set to 50% of the smallest surface 21938 to air missile (SAM) threat radius without 21939 incurring prohibitive computational costs. 21940 </blockquote> 21941 </div> 21942 </div> 21943 </div> 21944 </div> 21945 <div class="slot-urls"></div> 21946 <a href="/wsc23papers/202.pdf" target="_blank">pdf</a 21947 ><br /> 21948 </div> 21949 <div class="slot-entry"> 21950 <a name="con149" tabindex="-1"></a> 21951 <div class="slot-title-line">
21952 <span class="slot-title" 21953 >Simulation-Based Optimization of Air Force Mission 21954 Planning</span 21955 > 21956 </div> 21957 <div> 21958 <span class="BAP award" 21959 >Best Contributed Applied Paper - Finalist</span 21960 > 21961 </div> 21962 <div class="slot-authors"> 21963 Mihaela Lechner and Alexander Roman (University of the 21964 Bundeswehr Munich), Thomas Mayer (ESG Elektroniksystem- 21965 und Logistik-GmbH), and Tobias Uhlig and Oliver Rose 21966 (University of the Bundeswehr Munich) 21967 </div> 21968 <div class="slot-abstract"> 21969 <div> 21970 <a 21971 class="clickable no-decoration" 21972 id="vhsjs_view_488_1707793552_2991424" 21973 onclick="$('#vhsjs_view_488_1707793552_2991424').hide(); 21974 $('#vhsjs_hide_488_1707793552_2991424').show(); 21975 $('#487_1707793552_2991343').slideDown(function() { 21976 if (typeof Masonry === 'function') { 21977 $('.use_masonry').masonry(); 21978 }; 21979 21980 });" 21981 ><i class="fa fa-caret-right"></i> 21982 <span class="hover_link">Abstract</span></a 21983 ><a 21984 class="clickable no-decoration" 21985 id="vhsjs_hide_488_1707793552_2991424" 21986 onclick="$('#487_1707793552_2991343').hide(function() { 21987 if (typeof Masonry === 'function') { 21988 $('.use_masonry').masonry(); 21989 }; 21990 }); 21991 $('#vhsjs_hide_488_1707793552_2991424').hide(); 21992 $('#vhsjs_view_488_1707793552_2991424').show();" 21993 style="display: none" 21994 ><i class="fa fa-caret-down"></i> 21995 <span class="hover_link">Abstract</span></a 21996 > 21997 <div 21998 data-display-control="488_1707793552_2991424" 21999 id="487_1707793552_2991343" 22000 style="display: none" 22001 > 22002 <div class="arrow-slidedown"> 22003 <blockquote> 22004 Military planning operations deal with highly 22005 dynamic environments and a variety of complex 22006 optimization challenges. In order to support 22007 decision-makers in this process, innovative 22008 concepts are required that can automatically 22009 generate applicable solutions for certain 22010 aspects of mission planning. Such instruments 22011 can simplify the planning process, reduce risks, 22012 and lower operating costs. This paper presents a 22013 simulation-based optimization framework that 22014 addresses three problems in the context of 22015 aerial warfare planning: task assignment, 22016 scheduling, and route planning. These problems 22017 are tackled with interconnected heuristics based 22018 on either greedy approaches or genetic 22019 algorithms. Additionally, hierarchical task 22020 networks are employed to incorporate domain 22021 knowledge in form of tactical doctrines into the 22022 solution. Our simulation results confirm the 22023 viability of the proposed approach for small to 22024 medium-sized scenarios. However, further 22025 investigation with regard to the evaluation 22026 function and the simulation environment is 22027 required. 22028 </blockquote> 22029 </div> 22030 </div> 22031 </div> 22032 </div> 22033 <div class="slot-urls"></div> 22034 <a href="/wsc23papers/203.pdf" target="_blank">pdf</a 22035 ><br /> 22036 </div> 22037 <div class="slot-entry"> 22038 <a name="con214" tabindex="-1"></a> 22039 <div class="slot-title-line">
22040 <span class="slot-title" 22041 >Discrete Event Simulation of Aircraft Sortie 22042 Generation on an Aircraft Carrier</span 22043 > 22044 </div> 22045 <div class="slot-authors"> 22046 Hee Chang Yoon and Seung Heon Oh (Seoul National 22047 University); Jung-Hoon Chung, Hyuk Lee, and Sun-Ah Jung 22048 (Korea Institute of Machinery & Materials); and Jong Hun 22049 Woo (Seoul National University) 22050 </div> 22051 <div class="slot-abstract"> 22052 <div> 22053 <a 22054 class="clickable no-decoration" 22055 id="vhsjs_view_490_1707793552_3017967" 22056 onclick="$('#vhsjs_view_490_1707793552_3017967').hide(); 22057 $('#vhsjs_hide_490_1707793552_3017967').show(); 22058 $('#489_1707793552_3017883').slideDown(function() { 22059 if (typeof Masonry === 'function') { 22060 $('.use_masonry').masonry(); 22061 }; 22062 22063 });" 22064 ><i class="fa fa-caret-right"></i> 22065 <span class="hover_link">Abstract</span></a 22066 ><a 22067 class="clickable no-decoration" 22068 id="vhsjs_hide_490_1707793552_3017967" 22069 onclick="$('#489_1707793552_3017883').hide(function() { 22070 if (typeof Masonry === 'function') { 22071 $('.use_masonry').masonry(); 22072 }; 22073 }); 22074 $('#vhsjs_hide_490_1707793552_3017967').hide(); 22075 $('#vhsjs_view_490_1707793552_3017967').show();" 22076 style="display: none" 22077 ><i class="fa fa-caret-down"></i> 22078 <span class="hover_link">Abstract</span></a 22079 > 22080 <div 22081 data-display-control="490_1707793552_3017967" 22082 id="489_1707793552_3017883" 22083 style="display: none" 22084 > 22085 <div class="arrow-slidedown"> 22086 <blockquote> 22087 The Sortie Generation Rate (SGR) which refers to 22088 the number of sorties that can be generated per 22089 unit time, is a key indicator for evaluating the 22090 ability of an airbase. However, an aircraft 22091 carrier has many constraints compared to a 22092 land-based airbase, such as spatial and 22093 environmental constraints, making it difficult 22094 to apply existing land-based research to analyze 22095 aircraft carrier operations. On the other hand, 22096 the Sortie Generation Process (SGP) on an 22097 aircraft carrier is similar to a 22098 logistics/production system in that sorties are 22099 generated through aircraft. Therefore, this 22100 study proposes a framework for analyzing the SGP 22101 on an aircraft carrier using discrete event 22102 simulation and defines the classes that make up 22103 the simulation. In addition, SGP analysis 22104 simulations were implemented using the proposed 22105 framework and several experiments were performed 22106 to demonstrate the feasibility of applying the 22107 proposed framework in practice. 22108 </blockquote> 22109 </div> 22110 </div> 22111 </div> 22112 </div> 22113 <div class="slot-urls"></div> 22114 <a href="/wsc23papers/204.pdf" target="_blank">pdf</a 22115 ><br /> 22116 </div> 22117 </div> 22118 <div class="session-entry">
22119 <span class="session-event-type">Technical Session</span 22120 ><span class="type-track-spacer"> · </span 22121 ><span class="program-track" 22122 >Military and National Security Applications</span 22123 ><br /> 22124 <div class="session-title"> 22125 Improving Cyber and Information Warfare Operations 22126 </div> 22127 <div class="session-chair"> 22128 Chair: Josiah Steckenrider (United States Military 22129 Academy)<br /> 22130 </div> 22131 <div class="slot-entry"> 22132 <a name="con142" tabindex="-1"></a> 22133 <div class="slot-title-line"> 22134 <span class="slot-title" 22135 >The Holistic Prioritized SATCOM Throughput 22136 Requirements (HPSTR) Stochastic Model</span 22137 > 22138 </div> 22139 <div class="slot-authors"> 22140 Matthew Wesloh, Noelle Douglas, Brianne White, and 22141 Nicholas Shallcross (United States Army, The Research 22142 and Analysis Center) 22143 </div> 22144 <div class="slot-abstract"> 22145 <div> 22146 <a 22147 class="clickable no-decoration" 22148 id="vhsjs_view_492_1707793552_3073788" 22149 onclick="$('#vhsjs_view_492_1707793552_3073788').hide(); 22150 $('#vhsjs_hide_492_1707793552_3073788').show(); 22151 $('#491_1707793552_3073702').slideDown(function() { 22152 if (typeof Masonry === 'function') { 22153 $('.use_masonry').masonry(); 22154 }; 22155 22156 });" 22157 ><i class="fa fa-caret-right"></i> 22158 <span class="hover_link">Abstract</span></a 22159 ><a 22160 class="clickable no-decoration" 22161 id="vhsjs_hide_492_1707793552_3073788" 22162 onclick="$('#491_1707793552_3073702').hide(function() { 22163 if (typeof Masonry === 'function') { 22164 $('.use_masonry').masonry(); 22165 }; 22166 }); 22167 $('#vhsjs_hide_492_1707793552_3073788').hide(); 22168 $('#vhsjs_view_492_1707793552_3073788').show();" 22169 style="display: none" 22170 ><i class="fa fa-caret-down"></i> 22171 <span class="hover_link">Abstract</span></a 22172 > 22173 <div 22174 data-display-control="492_1707793552_3073788" 22175 id="491_1707793552_3073702" 22176 style="display: none" 22177 > 22178 <div class="arrow-slidedown"> 22179 <blockquote> 22180 The U.S. Army's command and control 22181 modernization efforts rely upon an 22182 expeditionary, mobile, hardened, and resilient 22183 network. Dispersed network access and data 22184 availability are central to increasing the 22185 operational speed required for effective command 22186 and control. The Army must define its satellite 22187 communication (SATCOM) requirements to support 22188 network modernization. This paper proposes the 22189 Holistic Prioritized SATCOM Throughput 22190 Requirements (HPSTR) simulation that prioritizes 22191 and adjudicates SATCOM throughput requirements 22192 for operational military units. Additionally, 22193 the simulation evaluates the impact of a 22194 contested, degraded, and operationally limited 22195 (CDO) communication environment on force 22196 effectiveness. HPSTR addresses knowledge gaps 22197 concerning U.S. Army SATCOM activities in a 22198 large-scale combat operation (LSCO) to inform 22199 modernization decisions. 22200 </blockquote> 22201 </div> 22202 </div> 22203 </div> 22204 </div> 22205 <div class="slot-urls"></div> 22206 <a href="/wsc23papers/205.pdf" target="_blank">pdf</a 22207 ><br /> 22208 </div> 22209 <div class="slot-entry"> 22210 <a name="con293" tabindex="-1"></a> 22211 <div class="slot-title-line">
22212 <span class="slot-title" 22213 >Using Simulated Narratives to Understand Attribution 22214 in the Information Dimension</span 22215 > 22216 </div> 22217 <div class="slot-authors"> 22218 Elijah Bellamy and David Beskow (United States Military 22219 Academy) 22220 </div> 22221 <div class="slot-abstract"> 22222 <div> 22223 <a 22224 class="clickable no-decoration" 22225 id="vhsjs_view_494_1707793552_309553" 22226 onclick="$('#vhsjs_view_494_1707793552_309553').hide(); 22227 $('#vhsjs_hide_494_1707793552_309553').show(); 22228 $('#493_1707793552_3095446').slideDown(function() { 22229 if (typeof Masonry === 'function') { 22230 $('.use_masonry').masonry(); 22231 }; 22232 22233 });" 22234 ><i class="fa fa-caret-right"></i> 22235 <span class="hover_link">Abstract</span></a 22236 ><a 22237 class="clickable no-decoration" 22238 id="vhsjs_hide_494_1707793552_309553" 22239 onclick="$('#493_1707793552_3095446').hide(function() { 22240 if (typeof Masonry === 'function') { 22241 $('.use_masonry').masonry(); 22242 }; 22243 }); 22244 $('#vhsjs_hide_494_1707793552_309553').hide(); 22245 $('#vhsjs_view_494_1707793552_309553').show();" 22246 style="display: none" 22247 ><i class="fa fa-caret-down"></i> 22248 <span class="hover_link">Abstract</span></a 22249 > 22250 <div 22251 data-display-control="494_1707793552_309553" 22252 id="493_1707793552_3095446" 22253 style="display: none" 22254 > 22255 <div class="arrow-slidedown"> 22256 <blockquote> 22257 Conducting a measured response to cyber or 22258 information attack is predicated on attribution. 22259 When these operations are conducted covertly or 22260 through proxies, uncertainty in attribution 22261 limits response options. To increase attribution 22262 certainty in the information dimension, the 22263 authors have developed a suite of supervised 22264 machine learning models that attribute an 22265 emerging narrative to historical narratives from 22266 known actors. These models were first developed 22267 on simulated narratives produced with a Large 22268 Language Model. Once the supervised 22269 classification models were developed and tested 22270 on the simulated narratives, they are evaluated 22271 on known actor social media narratives from 22272 three known actors. The attribution models are 22273 language agnostic and offer one-vs-rest and 22274 multi-class options. All models performed at 22275 relatively high accuracy and can provide 22276 decision support for cyber response decisions. 22277 </blockquote> 22278 </div> 22279 </div> 22280 </div> 22281 </div> 22282 <div class="slot-urls"></div> 22283 <a href="/wsc23papers/206.pdf" target="_blank">pdf</a 22284 ><br /> 22285 </div> 22286 <div class="slot-entry"> 22287 <a name="inv103" tabindex="-1"></a> 22288 <div class="slot-title-line">
22289 <span class="slot-title" 22290 >Uncertainty-Quantified, Robust Deep Learning for 22291 Network Intrusion Detection</span 22292 > 22293 </div> 22294 <div class="slot-authors"> 22295 Joshua Wong, Alexander Berenbeim, David Bierbrauer, and 22296 Nathaniel Bastian (United States Military Academy) 22297 </div> 22298 <div class="slot-abstract"> 22299 <div> 22300 <a 22301 class="clickable no-decoration" 22302 id="vhsjs_view_496_1707793552_311826" 22303 onclick="$('#vhsjs_view_496_1707793552_311826').hide(); 22304 $('#vhsjs_hide_496_1707793552_311826').show(); 22305 $('#495_1707793552_311818').slideDown(function() { 22306 if (typeof Masonry === 'function') { 22307 $('.use_masonry').masonry(); 22308 }; 22309 22310 });" 22311 ><i class="fa fa-caret-right"></i> 22312 <span class="hover_link">Abstract</span></a 22313 ><a 22314 class="clickable no-decoration" 22315 id="vhsjs_hide_496_1707793552_311826" 22316 onclick="$('#495_1707793552_311818').hide(function() { 22317 if (typeof Masonry === 'function') { 22318 $('.use_masonry').masonry(); 22319 }; 22320 }); 22321 $('#vhsjs_hide_496_1707793552_311826').hide(); 22322 $('#vhsjs_view_496_1707793552_311826').show();" 22323 style="display: none" 22324 ><i class="fa fa-caret-down"></i> 22325 <span class="hover_link">Abstract</span></a 22326 > 22327 <div 22328 data-display-control="496_1707793552_311826" 22329 id="495_1707793552_311818" 22330 style="display: none" 22331 > 22332 <div class="arrow-slidedown"> 22333 <blockquote> 22334 Cyber threats are moving beyond human 22335 comprehension and reaction capability in a 22336 rapidly evolving world. Deep learning models for 22337 network intrusion detection are becoming 22338 evermore crucial in processing network traffic 22339 to filter benign content from malicious 22340 activity. However, novel attacks such as 22341 zero-days are becoming more frequent, 22342 demonstrating the need for robust deep learning 22343 models to flag attacks while providing 22344 predictive certainty guarantees. Therefore, 22345 detecting out-of-distribution (OOD) inputs at 22346 inference time is crucial to address the rapidly 22347 changing environment while keeping up with 22348 evolving cyber threats. We develop multi-class 22349 deep learning models for network intrusion 22350 detection, comparing deterministic with Bayesian 22351 neural networks estimated using Hamiltonian 22352 Monte Carlo. We also propose new uncertainty 22353 quantification scoring measures for performance 22354 evaluation to evaluate certainty in predictions. 22355 During our experimentation, our best performing 22356 proposed Bayesian deep learning model detected 22357 89.1% and 86.9% of the OOD packets at the 5% and 22358 0.1% significance levels, respectively. 22359 </blockquote> 22360 </div> 22361 </div> 22362 </div> 22363 </div> 22364 <div class="slot-urls"></div> 22365 <a href="/wsc23papers/207.pdf" target="_blank">pdf</a 22366 ><br /> 22367 </div> 22368 </div> 22369 <div class="session-entry">
22370 <span class="session-event-type">Technical Session</span 22371 ><span class="type-track-spacer"> · </span 22372 ><span class="program-track" 22373 >Military and National Security Applications</span 22374 ><br /> 22375 <div class="session-title"> 22376 Simulating Search and Naval Operations 22377 </div> 22378 <div class="session-chair"> 22379 Chair: Lance Champagne (AFIT)<br /> 22380 </div> 22381 <div class="slot-entry"> 22382 <a name="con257" tabindex="-1"></a> 22383 <div class="slot-title-line"> 22384 <span class="slot-title" 22385 >A Comparison of Lissajous Curves to Traditional 22386 Patterns in Aerial Search Simulations</span 22387 > 22388 </div> 22389 <div class="slot-authors"> 22390 Mitchell J. Miller, Victor E. Trujillo, James E. Bluman, 22391 and J. Josiah Steckenrider (United States Military 22392 Academy) 22393 </div> 22394 <div class="slot-abstract"> 22395 <div> 22396 <a 22397 class="clickable no-decoration" 22398 id="vhsjs_view_498_1707793552_3165302" 22399 onclick="$('#vhsjs_view_498_1707793552_3165302').hide(); 22400 $('#vhsjs_hide_498_1707793552_3165302').show(); 22401 $('#497_1707793552_316522').slideDown(function() { 22402 if (typeof Masonry === 'function') { 22403 $('.use_masonry').masonry(); 22404 }; 22405 22406 });" 22407 ><i class="fa fa-caret-right"></i> 22408 <span class="hover_link">Abstract</span></a 22409 ><a 22410 class="clickable no-decoration" 22411 id="vhsjs_hide_498_1707793552_3165302" 22412 onclick="$('#497_1707793552_316522').hide(function() { 22413 if (typeof Masonry === 'function') { 22414 $('.use_masonry').masonry(); 22415 }; 22416 }); 22417 $('#vhsjs_hide_498_1707793552_3165302').hide(); 22418 $('#vhsjs_view_498_1707793552_3165302').show();" 22419 style="display: none" 22420 ><i class="fa fa-caret-down"></i> 22421 <span class="hover_link">Abstract</span></a 22422 > 22423 <div 22424 data-display-control="498_1707793552_3165302" 22425 id="497_1707793552_316522" 22426 style="display: none" 22427 > 22428 <div class="arrow-slidedown"> 22429 <blockquote> 22430 Technological advancements have made autonomous 22431 aerial search using unmanned systems a promising 22432 approach to search and rescue, targeting, and 22433 other mission sets. A handful of standard flight 22434 paths are traditionally used for aerial search, 22435 but this research presents the Lissajous pattern 22436 as an alternative to these traditional paths 22437 that could potentially locate targets more 22438 quickly. This research considers a searching 22439 agent with imperfect detection capability and 22440 leverages Monte Carlo simulations to generate 22441 data for various flight paths. Each flight path 22442 is evaluated by cumulative density functions 22443 representing the time it takes an unmanned 22444 aircraft system (UAS) to reach some desired 22445 percent certainty of locating a randomly 22446 generated target in a search area. Results show 22447 that Lissajous curves are viable search paths 22448 for superior aerial target detection, 22449 particularly for evasive targets in a Reciprocal 22450 Gaussian sampling distribution. 22451 </blockquote> 22452 </div> 22453 </div> 22454 </div> 22455 </div> 22456 <div class="slot-urls"></div> 22457 <a href="/wsc23papers/208.pdf" target="_blank">pdf</a 22458 ><br /> 22459 </div> 22460 <div class="slot-entry"> 22461 <a name="cea130" tabindex="-1"></a> 22462 <div class="slot-title-line">
22463 <span class="slot-title" 22464 >Naval Combat Wargame Simulation for Susceptibility 22465 Analysis</span 22466 > 22467 </div> 22468 <div class="slot-authors"> 22469 Gun-Woong Byun and Seung-Heon Oh (Seoul National 22470 University, Department of Naval Architecture and Ocean 22471 Engineering); Jong-Ho Nam (Korea Maritime & Ocean 22472 University, Division of Naval Architecture and Ocean 22473 Systems Engineering); and Jong Hun Woo (Seoul National 22474 University, Department of Naval Architecture and Ocean 22475 Engineering) 22476 </div> 22477 <div class="slot-abstract"> 22478 <div> 22479 <a 22480 class="clickable no-decoration" 22481 id="vhsjs_view_500_1707793552_318719" 22482 onclick="$('#vhsjs_view_500_1707793552_318719').hide(); 22483 $('#vhsjs_hide_500_1707793552_318719').show(); 22484 $('#499_1707793552_3187106').slideDown(function() { 22485 if (typeof Masonry === 'function') { 22486 $('.use_masonry').masonry(); 22487 }; 22488 22489 });" 22490 ><i class="fa fa-caret-right"></i> 22491 <span class="hover_link">Abstract</span></a 22492 ><a 22493 class="clickable no-decoration" 22494 id="vhsjs_hide_500_1707793552_318719" 22495 onclick="$('#499_1707793552_3187106').hide(function() { 22496 if (typeof Masonry === 'function') { 22497 $('.use_masonry').masonry(); 22498 }; 22499 }); 22500 $('#vhsjs_hide_500_1707793552_318719').hide(); 22501 $('#vhsjs_view_500_1707793552_318719').show();" 22502 style="display: none" 22503 ><i class="fa fa-caret-down"></i> 22504 <span class="hover_link">Abstract</span></a 22505 > 22506 <div 22507 data-display-control="500_1707793552_318719" 22508 id="499_1707793552_3187106" 22509 style="display: none" 22510 > 22511 <div class="arrow-slidedown"> 22512 <blockquote> 22513 An engagement between naval ships is defined as 22514 a multi-agent system with multiple ships 22515 interacting. Because of the limitations of 22516 conducting and analyzing engagement, it is 22517 common to use modeling and simulation or wargame 22518 simulations. Most of the existing wargame 22519 simulation studies focus on simulation 22520 frameworks rather than real-world applications 22521 and tend to focus on the evaluation of single 22522 entities that comprise a wargame. Thus, this 22523 study improves the reality of the simulation by 22524 modeling objects that constitute a complex 22525 engagement situation based on the simulation 22526 framework. In addition, developed analytical 22527 tools to automate and accelerate Monte Carlo 22528 simulations of engagement-level wargames that 22529 require large numbers of human and time 22530 resources. The developed simulations enable the 22531 application of various engagement scenarios to 22532 evaluate strategies and tactics. Furthermore, 22533 experiments are possible while altering the 22534 design parameters of the naval ship, which 22535 allows for the evaluation of the ship's 22536 performance in combat. 22537 </blockquote> 22538 </div> 22539 </div> 22540 </div> 22541 </div> 22542 <div class="slot-urls"></div> 22543 <a href="/wsc23papers/cea130.pdf" target="_blank">pdf</a 22544 ><br /> 22545 </div> 22546 </div> 22547 </div> 22548 <div class="centered"> 22549 <div class="top-link"><a href="#top">Return to Top</a></div> 22550 </div> 22551 <hr /> 22552 </div> 22553 <div class="area-section"> 22554 <div class="centered"> 22555 <a name="ptrack117" tabindex="-1"></a> 22556 <div class="section-title">Modeling Methodology</div> 22557 </div> 22558 <div class="centered track-chair">
22559 <span class="track-chair-role" 22560 >Track Coordinator - Modeling Methodology: </span 22561 ><span class="track-chair-names" 22562 >Rodrigo Castro (ICC-CONICET, Universidad de Buenos Aires), 22563 Andrea D'Ambrogio (University of Roma TorVergata), Gerd 22564 Wagner (Brandenburg University of Technology), Gabriel 22565 Wainer (Carleton University)</span 22566 > 22567 </div> 22568 <div class="section-entry"> 22569 <div class="session-entry"> 22570 <span class="session-event-type">Technical Session</span 22571 ><span class="type-track-spacer"> · </span 22572 ><span class="program-track">Modeling Methodology</span 22573 ><br /> 22574 <div class="session-title">Complex Systems</div> 22575 <div class="session-chair"> 22576 Chair: Margaret Loper (Georgia Tech Research Institute)<br /> 22577 </div> 22578 <div class="slot-entry"> 22579 <a name="con126" tabindex="-1"></a> 22580 <div class="slot-title-line"> 22581 <span class="slot-title" 22582 >Towards an Automatic Construction of Simulation 22583 Scenarios: A Systematic Review</span 22584 > 22585 </div> 22586 <div class="slot-authors"> 22587 Christopher W.H. Davis (Microsoft), Antonie J. Jetter 22588 (Portland State University), and Philippe J. Giabbanelli 22589 (Miami University) 22590 </div> 22591 <div class="slot-abstract"> 22592 <div> 22593 <a 22594 class="clickable no-decoration" 22595 id="vhsjs_view_502_1707793552_3265965" 22596 onclick="$('#vhsjs_view_502_1707793552_3265965').hide(); 22597 $('#vhsjs_hide_502_1707793552_3265965').show(); 22598 $('#501_1707793552_3265882').slideDown(function() { 22599 if (typeof Masonry === 'function') { 22600 $('.use_masonry').masonry(); 22601 }; 22602 22603 });" 22604 ><i class="fa fa-caret-right"></i> 22605 <span class="hover_link">Abstract</span></a 22606 ><a 22607 class="clickable no-decoration" 22608 id="vhsjs_hide_502_1707793552_3265965" 22609 onclick="$('#501_1707793552_3265882').hide(function() { 22610 if (typeof Masonry === 'function') { 22611 $('.use_masonry').masonry(); 22612 }; 22613 }); 22614 $('#vhsjs_hide_502_1707793552_3265965').hide(); 22615 $('#vhsjs_view_502_1707793552_3265965').show();" 22616 style="display: none" 22617 ><i class="fa fa-caret-down"></i> 22618 <span class="hover_link">Abstract</span></a 22619 > 22620 <div 22621 data-display-control="502_1707793552_3265965" 22622 id="501_1707793552_3265882" 22623 style="display: none" 22624 > 22625 <div class="arrow-slidedown"> 22626 <blockquote> 22627 A predictive simulation is built on a conceptual 22628 model (e.g., to identify relevant constructs and 22629 relationships) and serves to estimate the 22630 potential effects of `what-if' scenarios. 22631 Developing the conceptual model and plausible 22632 scenarios has long been a time-consuming 22633 activity, often involving the manual processes 22634 of identifying and engaging with experts, then 22635 performing desk research, and finally crafting a 22636 compelling narrative about the potential futures 22637 captured as scenarios. Automation could speed-up 22638 these activities, particularly through text 22639 mining. We performed the first review on 22640 automation for simulation scenario building. 22641 Starting with 420 articles published between 22642 1995 and 2022, we reduced them to 11 relevant 22643 works. We examined them through four research 22644 questions concerning data collection, extraction 22645 of individual elements, connecting elements of 22646 insight and (degree of automation of) scenario 22647 generation. Our review identifies opportunities 22648 to guide this growing research area by 22649 emphasizing consistency and transparency in the 22650 choice of datasets or methods. 22651 </blockquote> 22652 </div> 22653 </div> 22654 </div> 22655 </div> 22656 <div class="slot-urls"></div> 22657 <a href="/wsc23papers/209.pdf" target="_blank">pdf</a 22658 ><br /> 22659 </div> 22660 <div class="slot-entry"> 22661 <a name="inv161" tabindex="-1"></a> 22662 <div class="slot-title-line">
22663 <span class="slot-title" 22664 >Evolving LVC to Include Evaluation of Human-AI 22665 Teaming Dynamics</span 22666 > 22667 </div> 22668 <div class="slot-authors"> 22669 Margaret Loper and Valerie Sitterle (GTRI) 22670 </div> 22671 <div class="slot-abstract"> 22672 <div> 22673 <a 22674 class="clickable no-decoration" 22675 id="vhsjs_view_504_1707793552_328824" 22676 onclick="$('#vhsjs_view_504_1707793552_328824').hide(); 22677 $('#vhsjs_hide_504_1707793552_328824').show(); 22678 $('#503_1707793552_328816').slideDown(function() { 22679 if (typeof Masonry === 'function') { 22680 $('.use_masonry').masonry(); 22681 }; 22682 22683 });" 22684 ><i class="fa fa-caret-right"></i> 22685 <span class="hover_link">Abstract</span></a 22686 ><a 22687 class="clickable no-decoration" 22688 id="vhsjs_hide_504_1707793552_328824" 22689 onclick="$('#503_1707793552_328816').hide(function() { 22690 if (typeof Masonry === 'function') { 22691 $('.use_masonry').masonry(); 22692 }; 22693 }); 22694 $('#vhsjs_hide_504_1707793552_328824').hide(); 22695 $('#vhsjs_view_504_1707793552_328824').show();" 22696 style="display: none" 22697 ><i class="fa fa-caret-down"></i> 22698 <span class="hover_link">Abstract</span></a 22699 > 22700 <div 22701 data-display-control="504_1707793552_328824" 22702 id="503_1707793552_328816" 22703 style="display: none" 22704 > 22705 <div class="arrow-slidedown"> 22706 <blockquote> 22707 There are significant differences between using 22708 systems as human-controlled tools to accomplish 22709 a specific task and using systems designed to 22710 “cooperate and partner” with humans 22711 to achieve capabilities beyond either side 22712 acting alone. The live, virtual, constructive 22713 (LVC) paradigm increasingly emphasized by the 22714 DoD has wide acceptance and is congruent with 22715 how the military thinks about training, 22716 evaluation, and mission rehearsal. Consequently, 22717 it may help address these challenges. This paper 22718 aims to overview the current LVC construct, 22719 challenges associated with human-AI teaming and 22720 intentional design of these dynamics to achieve 22721 new capabilities, and the resulting need to 22722 evolve the LVC construct to improve our pursuit 22723 of understanding and evaluation that leads to 22724 effective fielding. 22725 </blockquote> 22726 </div> 22727 </div> 22728 </div> 22729 </div> 22730 <div class="slot-urls"></div> 22731 <a href="/wsc23papers/210.pdf" target="_blank">pdf</a 22732 ><br /> 22733 </div> 22734 <div class="slot-entry"> 22735 <a name="con265" tabindex="-1"></a> 22736 <div class="slot-title-line">
22737 <span class="slot-title" 22738 >How to Combine Models? Principles and Mechanisms to 22739 Aggregate Fuzzy Cognitive Maps</span 22740 > 22741 </div> 22742 <div class="slot-authors"> 22743 Ryan Schuerkamp and Philippe J. Giabbanelli (Miami 22744 University) and Umberto Grandi and Sylvie Doutre 22745 (Université Toulouse Capitole) 22746 </div> 22747 <div class="slot-abstract"> 22748 <div> 22749 <a 22750 class="clickable no-decoration" 22751 id="vhsjs_view_506_1707793552_3311102" 22752 onclick="$('#vhsjs_view_506_1707793552_3311102').hide(); 22753 $('#vhsjs_hide_506_1707793552_3311102').show(); 22754 $('#505_1707793552_3311017').slideDown(function() { 22755 if (typeof Masonry === 'function') { 22756 $('.use_masonry').masonry(); 22757 }; 22758 22759 });" 22760 ><i class="fa fa-caret-right"></i> 22761 <span class="hover_link">Abstract</span></a 22762 ><a 22763 class="clickable no-decoration" 22764 id="vhsjs_hide_506_1707793552_3311102" 22765 onclick="$('#505_1707793552_3311017').hide(function() { 22766 if (typeof Masonry === 'function') { 22767 $('.use_masonry').masonry(); 22768 }; 22769 }); 22770 $('#vhsjs_hide_506_1707793552_3311102').hide(); 22771 $('#vhsjs_view_506_1707793552_3311102').show();" 22772 style="display: none" 22773 ><i class="fa fa-caret-down"></i> 22774 <span class="hover_link">Abstract</span></a 22775 > 22776 <div 22777 data-display-control="506_1707793552_3311102" 22778 id="505_1707793552_3311017" 22779 style="display: none" 22780 > 22781 <div class="arrow-slidedown"> 22782 <blockquote> 22783 Fuzzy Cognitive Maps (FCMs) are graph-based 22784 simulation models commonly used to model complex 22785 systems. They are often built by participants 22786 and aggregated to compare the viewpoints of 22787 homogenous groups (e.g., anglers and ecologists) 22788 and increase the reliability of the FCM. 22789 However, the default approach for aggregation 22790 may propagate the errors of an individual 22791 participant, producing an aggregate FCM whose 22792 structure and simulation outcomes do not align 22793 with the system of interest. Alternative 22794 aggregation methods exist; however, there are no 22795 criteria to assess the quality of aggregation 22796 methods. We define nine desirable criteria for 22797 FCM aggregation algorithms and demonstrate how 22798 three existing aggregation procedures from 22799 social choice theory can aggregate FCMs and 22800 fulfill desirable criteria, enabling the 22801 assessment and comparison of FCM aggregation 22802 procedures to support modelers in selecting an 22803 aggregation algorithm. Moreover, we classify 22804 existing aggregation algorithms to provide 22805 structure to the growing body of aggregation 22806 approaches. 22807 </blockquote> 22808 </div> 22809 </div> 22810 </div> 22811 </div> 22812 <div class="slot-urls"></div> 22813 <a href="/wsc23papers/211.pdf" target="_blank">pdf</a 22814 ><br /> 22815 </div> 22816 </div> 22817 <div class="session-entry">
22818 <span class="session-event-type">Technical Session</span 22819 ><span class="type-track-spacer"> · </span 22820 ><span class="program-track">Modeling Methodology</span 22821 ><br /> 22822 <div class="session-title">Modeling Methods</div> 22823 <div class="session-chair"> 22824 Chair: Gabriel Wainer (Carleton University)<br /> 22825 </div> 22826 <div class="slot-entry"> 22827 <a name="con246" tabindex="-1"></a> 22828 <div class="slot-title-line"> 22829 <span class="slot-title" 22830 >A Low-Code Approach for Simulation-based Analysis of 22831 Process Collaborations</span 22832 > 22833 </div> 22834 <div class="slot-authors"> 22835 Paolo Bocciarelli and Andrea D'Ambrogio (University of 22836 Rome Tor Vergata) 22837 </div> 22838 <div class="slot-abstract"> 22839 <div> 22840 <a 22841 class="clickable no-decoration" 22842 id="vhsjs_view_508_1707793552_3358648" 22843 onclick="$('#vhsjs_view_508_1707793552_3358648').hide(); 22844 $('#vhsjs_hide_508_1707793552_3358648').show(); 22845 $('#507_1707793552_3358562').slideDown(function() { 22846 if (typeof Masonry === 'function') { 22847 $('.use_masonry').masonry(); 22848 }; 22849 22850 });" 22851 ><i class="fa fa-caret-right"></i> 22852 <span class="hover_link">Abstract</span></a 22853 ><a 22854 class="clickable no-decoration" 22855 id="vhsjs_hide_508_1707793552_3358648" 22856 onclick="$('#507_1707793552_3358562').hide(function() { 22857 if (typeof Masonry === 'function') { 22858 $('.use_masonry').masonry(); 22859 }; 22860 }); 22861 $('#vhsjs_hide_508_1707793552_3358648').hide(); 22862 $('#vhsjs_view_508_1707793552_3358648').show();" 22863 style="display: none" 22864 ><i class="fa fa-caret-down"></i> 22865 <span class="hover_link">Abstract</span></a 22866 > 22867 <div 22868 data-display-control="508_1707793552_3358648" 22869 id="507_1707793552_3358562" 22870 style="display: none" 22871 > 22872 <div class="arrow-slidedown"> 22873 <blockquote> 22874 The simulation-based analysis of process 22875 collaborations introduces significant 22876 challenges, such as the ability to focus on the 22877 interchange of information and data without 22878 disclosing any internal details of collaboration 22879 participants' processes. The use of distributed 22880 simulation (DS) provides good opportunities to 22881 face these challenges. However, properly using 22882 DS standards and technologies requires 22883 significant technical know-how and effort. This 22884 paper introduces a largely automated approach to 22885 carry out distributed simulations of process 22886 collaborations. The DS standard addressed by the 22887 paper is the High Level Architecture (HLA), 22888 which is used to analyze process collaborations 22889 specified by using the Business Process Model 22890 and Notation (BPMN). The degree of automation is 22891 obtained by using a low-code development 22892 paradigm based on automated model 22893 transformations that reduce the amount of manual 22894 effort required to code the HLA-based 22895 simulation. An example application is also 22896 discussed to underline the pros and cons of the 22897 proposed approach. 22898 </blockquote> 22899 </div> 22900 </div> 22901 </div> 22902 </div> 22903 <div class="slot-urls"></div> 22904 <a href="/wsc23papers/212.pdf" target="_blank">pdf</a 22905 ><br /> 22906 </div> 22907 <div class="slot-entry"> 22908 <a name="inv211" tabindex="-1"></a> 22909 <div class="slot-title-line">
22910 <span class="slot-title" 22911 >Incremental Transformation of BPSIM-enriched BPMN 22912 Models into DEVS</span 22913 > 22914 </div> 22915 <div class="slot-authors"> 22916 Mariane El Kassis, Francois Trousset, Gregory 22917 Zacharewicz, and Nicolas Daclin (IMT Mines Alès) 22918 </div> 22919 <div class="slot-abstract"> 22920 <div> 22921 <a 22922 class="clickable no-decoration" 22923 id="vhsjs_view_510_1707793552_3383026" 22924 onclick="$('#vhsjs_view_510_1707793552_3383026').hide(); 22925 $('#vhsjs_hide_510_1707793552_3383026').show(); 22926 $('#509_1707793552_3382943').slideDown(function() { 22927 if (typeof Masonry === 'function') { 22928 $('.use_masonry').masonry(); 22929 }; 22930 22931 });" 22932 ><i class="fa fa-caret-right"></i> 22933 <span class="hover_link">Abstract</span></a 22934 ><a 22935 class="clickable no-decoration" 22936 id="vhsjs_hide_510_1707793552_3383026" 22937 onclick="$('#509_1707793552_3382943').hide(function() { 22938 if (typeof Masonry === 'function') { 22939 $('.use_masonry').masonry(); 22940 }; 22941 }); 22942 $('#vhsjs_hide_510_1707793552_3383026').hide(); 22943 $('#vhsjs_view_510_1707793552_3383026').show();" 22944 style="display: none" 22945 ><i class="fa fa-caret-down"></i> 22946 <span class="hover_link">Abstract</span></a 22947 > 22948 <div 22949 data-display-control="510_1707793552_3383026" 22950 id="509_1707793552_3382943" 22951 style="display: none" 22952 > 22953 <div class="arrow-slidedown"> 22954 <blockquote> 22955 In this paper, we introduce a novel methodology 22956 for business process simulation, focusing on the 22957 incremental transformation of Business Process 22958 Modeling and Notation (BPMN) models enriched 22959 with Business Process Simulation Interchange 22960 Standard (BPSIM) elements into the Discrete 22961 Event System Specification (DEVS) formalism. The 22962 proposed method enhances the precision and 22963 consistency of simulations by systematically 22964 converting BPMN components and BPSIM 22965 characteristics into DEVS representations, using 22966 adaptable rules and templates. A major 22967 contribution of this work is the introduction of 22968 the Interaction Intermediate Model (I2M), a 22969 model that provides a visually lucid 22970 representation with significant semantics, 22971 effectively encapsulating BPMN and BPSIM 22972 simulation aspects. The resulting DEVS model 22973 ensures accurate, reliable, and interoperable 22974 simulations. We provide a thorough analysis of 22975 this methodology, emphasize its advantages, and 22976 validate its efficiency through a case study. 22977 This method, applicable across various sectors 22978 effectively bridging the gap between conceptual 22979 modeling and simulation methodologies. 22980 </blockquote> 22981 </div> 22982 </div> 22983 </div> 22984 </div> 22985 <div class="slot-urls"></div> 22986 <a href="/wsc23papers/213.pdf" target="_blank">pdf</a 22987 ><br /> 22988 </div> 22989 <div class="slot-entry"> 22990 <a name="con356" tabindex="-1"></a> 22991 <div class="slot-title-line">
22992 <span class="slot-title" 22993 >An Approach Towards Predicting the Computational 22994 Runtime Reduction from Discrete-event Simulation Model 22995 Simplification Operations</span 22996 > 22997 </div> 22998 <div class="slot-authors"> 22999 Mohd Shoaib (Indian Institute of Technology Delhi), 23000 Navonil Mustafee (University of Exeter), and Varun 23001 Ramamohan (Indian Institute of Technology Delhi) 23002 </div> 23003 <div class="slot-abstract"> 23004 <div> 23005 <a 23006 class="clickable no-decoration" 23007 id="vhsjs_view_512_1707793552_3406792" 23008 onclick="$('#vhsjs_view_512_1707793552_3406792').hide(); 23009 $('#vhsjs_hide_512_1707793552_3406792').show(); 23010 $('#511_1707793552_3406713').slideDown(function() { 23011 if (typeof Masonry === 'function') { 23012 $('.use_masonry').masonry(); 23013 }; 23014 23015 });" 23016 ><i class="fa fa-caret-right"></i> 23017 <span class="hover_link">Abstract</span></a 23018 ><a 23019 class="clickable no-decoration" 23020 id="vhsjs_hide_512_1707793552_3406792" 23021 onclick="$('#511_1707793552_3406713').hide(function() { 23022 if (typeof Masonry === 'function') { 23023 $('.use_masonry').masonry(); 23024 }; 23025 }); 23026 $('#vhsjs_hide_512_1707793552_3406792').hide(); 23027 $('#vhsjs_view_512_1707793552_3406792').show();" 23028 style="display: none" 23029 ><i class="fa fa-caret-down"></i> 23030 <span class="hover_link">Abstract</span></a 23031 > 23032 <div 23033 data-display-control="512_1707793552_3406792" 23034 id="511_1707793552_3406713" 23035 style="display: none" 23036 > 23037 <div class="arrow-slidedown"> 23038 <blockquote> 23039 Model simplification is the process of 23040 developing a simplified version of an existing 23041 discrete-event simulation (DES) to study the 23042 performance of specific system subcomponents 23043 relevant to the analysis. The simplified model 23044 is referred to as a 'metasimulation'. A widely 23045 used model simplification operation is 23046 abstraction, which involves replacing the 23047 subcomponents, not core to the analysis, from 23048 the parent DES model with random variables 23049 representing the lengths of stay in said 23050 subcomponents. However, the one-time 23051 computational cost of developing metasimulations 23052 via abstraction can itself be considerable, as 23053 the approach necessitates executing the parent 23054 model for generating the necessary data for 23055 developing the metasimulation. Thus, this study 23056 proposes a queuing-theoretic approach for 23057 estimating the computational runtime reduction 23058 (CRR) achieved through abstraction, wherein the 23059 prediction of CRR precedes the development of 23060 the metasimulation. Towards this, we present 23061 preliminary results from applying this approach 23062 for simplification of DES models made up of 23063 M/M/n workstations. 23064 </blockquote> 23065 </div> 23066 </div> 23067 </div> 23068 </div> 23069 <div class="slot-urls"></div> 23070 <a href="/wsc23papers/214.pdf" target="_blank">pdf</a 23071 ><br /> 23072 </div> 23073 </div> 23074 <div class="session-entry">
23075 <span class="session-event-type">Technical Session</span 23076 ><span class="type-track-spacer"> · </span 23077 ><span class="program-track">Modeling Methodology</span 23078 ><br /> 23079 <div class="session-title"> 23080 Panel: Forty Years of Event Graphs in Research and 23081 Education 23082 </div> 23083 <div class="session-chair"> 23084 Chair: Gerd Wagner (Brandenburg University of 23085 Technology)<br /> 23086 </div> 23087 <div class="slot-entry"> 23088 <a name="inv196" tabindex="-1"></a> 23089 <div class="slot-title-line"> 23090 <span class="slot-title" 23091 >Forty Years of Event Graphs in Research and 23092 Education</span 23093 > 23094 </div> 23095 <div class="slot-authors"> 23096 Murat M. Gunal (Fenerbahce University); Yahya Ismail 23097 Osais (King Fahd University of Petroleum and Minerals, 23098 Interdisc. Research Center for Intellig. Secure 23099 Systems); Lee Schruben (University of California, 23100 Berkeley); Gerd Wagner (Brandenburg University of 23101 Technology); and Enver Yücesan (INSEAD) 23102 </div> 23103 <div class="slot-abstract"> 23104 <div> 23105 <a 23106 class="clickable no-decoration" 23107 id="vhsjs_view_514_1707793552_3455105" 23108 onclick="$('#vhsjs_view_514_1707793552_3455105').hide(); 23109 $('#vhsjs_hide_514_1707793552_3455105').show(); 23110 $('#513_1707793552_3455021').slideDown(function() { 23111 if (typeof Masonry === 'function') { 23112 $('.use_masonry').masonry(); 23113 }; 23114 23115 });" 23116 ><i class="fa fa-caret-right"></i> 23117 <span class="hover_link">Abstract</span></a 23118 ><a 23119 class="clickable no-decoration" 23120 id="vhsjs_hide_514_1707793552_3455105" 23121 onclick="$('#513_1707793552_3455021').hide(function() { 23122 if (typeof Masonry === 'function') { 23123 $('.use_masonry').masonry(); 23124 }; 23125 }); 23126 $('#vhsjs_hide_514_1707793552_3455105').hide(); 23127 $('#vhsjs_view_514_1707793552_3455105').show();" 23128 style="display: none" 23129 ><i class="fa fa-caret-down"></i> 23130 <span class="hover_link">Abstract</span></a 23131 > 23132 <div 23133 data-display-control="514_1707793552_3455105" 23134 id="513_1707793552_3455021" 23135 style="display: none" 23136 > 23137 <div class="arrow-slidedown"> 23138 <blockquote> 23139 Forty years ago, in 1983, Lee Schruben proposed 23140 the Event Graph formalism and modeling language, 23141 subsequently defining the paradigm of 23142 Event-Based Simulation, in a precise way, which 23143 had been pioneered 20 years before by SIMSCRIPT. 23144 The purpose of this panel is for a group of 23145 Event Graph researchers both from Operations 23146 Research and Computer Science, including the 23147 inventor of Event Graphs and one of his former 23148 PhD students who has made essential 23149 contributions to their theory, to discu
23149ss their 23150 views on the history and potential of Event 23151 Graph modeling and simulation. In particular, 23152 the adoption of Event Graphs as a discrete 23153 process modeling language in Discrete Event 23154 Simulation and in Computer Science, and their 23155 potential as a foundation for the entire field 23156 of Discrete Event Simulation and for the fields 23157 of process modeling and AI in Computer Science 23158 is debated. 23159 </blockquote> 23160 </div> 23161 </div> 23162 </div> 23163 </div> 23164 <div class="slot-urls"></div> 23165 <a href="/wsc23papers/215.pdf" target="_blank">pdf</a 23166 ><br /> 23167 </div> 23168 </div> 23169 <div class="session-entry"> 23170 <span class="session-event-type">Technical Session</span 23171 ><span class="type-track-spacer"> · </span 23172 ><span class="program-track">Modeling Methodology</span 23173 ><br /> 23174 <div class="session-title">DEVS</div> 23175 <div class="session-chair"> 23176 Chair: Hessam Sarjoughian (Arizona State University)<br /> 23177 </div> 23178 <div class="slot-entry"> 23179 <a name="con341" tabindex="-1"></a> 23180 <div class="slot-title-line"> 23181 <span class="slot-title" 23182 >A Context-Free Grammar for Generating Full Classic 23183 DEVS Models</span 23184 > 23185 </div> 23186 <div class="slot-authors"> 23187 María Julia Blas and Silvio Gonnet (INGAR 23188 (CONICET-UTN)) and Doohwan Kim and Bernard Zeigler 23189 (RTSync Corp.) 23190 </div> 23191 <div class="slot-abstract"> 23192 <div> 23193 <a 23194 class="clickable no-decoration" 23195 id="vhsjs_view_516_1707793552_3510997" 23196 onclick="$('#vhsjs_view_516_1707793552_3510997').hide(); 23197 $('#vhsjs_hide_516_1707793552_3510997').show(); 23198 $('#515_1707793552_3510914').slideDown(function() { 23199 if (typeof Masonry === 'function') { 23200 $('.use_masonry').masonry(); 23201 }; 23202 23203 });" 23204 ><i class="fa fa-caret-right"></i> 23205 <span class="hover_link">Abstract</span></a 23206 ><a 23207 class="clickable no-decoration" 23208 id="vhsjs_hide_516_1707793552_3510997" 23209 onclick="$('#515_1707793552_3510914').hide(function() { 23210 if (typeof Masonry === 'function') { 23211 $('.use_masonry').masonry(); 23212 }; 23213 }); 23214 $('#vhsjs_hide_516_1707793552_3510997').hide(); 23215 $('#vhsjs_view_516_1707793552_3510997').show();" 23216 style="display: none" 23217 ><i class="fa fa-caret-down"></i> 23218 <span class="hover_link">Abstract</span></a 23219 > 23220 <div 23221 data-display-control="516_1707793552_3510997" 23222 id="515_1707793552_3510914" 23223 style="display: none" 23224 > 23225 <div class="arrow-slidedown"> 23226 <blockquote> 23227 Existing grammars generate Finite Deterministic 23228 DEVS models, a restricted subset of DEVS. The 23229 proposed context-free grammar generates the
23230 unrestricted set of Classic DEVS models. The 23231 grammar is implemented in ANTLR, a powerful 23232 parser generator for reading, processing, 23233 executing, or translating structured text or 23234 binary files. ANTLR enables the efficient 23235 processing of the specifications needed for 23236 generating members of Classic DEVS with ports. 23237 Applications include an easier introduction to 23238 DEVS for students and easier translation between 23239 different DEVS implementations. 23240 </blockquote> 23241 </div> 23242 </div> 23243 </div> 23244 </div> 23245 <div class="slot-urls"></div> 23246 <a href="/wsc23papers/216.pdf" target="_blank">pdf</a 23247 ><br /> 23248 </div> 23249 <div class="slot-entry"> 23250 <a name="inv124" tabindex="-1"></a> 23251 <div class="slot-title-line"> 23252 <span class="slot-title" 23253 >CLAVS/ODVS: Combining Class/Object Diagrams and 23254 DEVS</span 23255 > 23256 </div> 23257 <div class="slot-authors"> 23258 Jordan Parezys and Randy Paredis (University of Antwerp) 23259 and Hans Vangheluwe (University of Antwerp, Flanders 23260 Make) 23261 </div> 23262 <div class="slot-abstract"> 23263 <div> 23264 <a 23265 class="clickable no-decoration" 23266 id="vhsjs_view_518_1707793552_3532493" 23267 onclick="$('#vhsjs_view_518_1707793552_3532493').hide(); 23268 $('#vhsjs_hide_518_1707793552_3532493').show(); 23269 $('#517_1707793552_3532407').slideDown(function() { 23270 if (typeof Masonry === 'function') { 23271 $('.use_masonry').masonry(); 23272 }; 23273 23274 });" 23275 ><i class="fa fa-caret-right"></i> 23276 <span class="hover_link">Abstract</span></a 23277 ><a 23278 class="clickable no-decoration" 23279 id="vhsjs_hide_518_1707793552_3532493" 23280 onclick="$('#517_1707793552_3532407').hide(function() { 23281 if (typeof Masonry === 'function') { 23282 $('.use_masonry').masonry(); 23283 }; 23284 }); 23285 $('#vhsjs_hide_518_1707793552_3532493').hide(); 23286 $('#vhsjs_view_518_1707793552_3532493').show();" 23287 style="display: none" 23288 ><i class="fa fa-caret-down"></i> 23289 <span class="hover_link">Abstract</span></a 23290 > 23291 <div 23292 data-display-control="518_1707793552_3532493" 23293 id="517_1707793552_3532407" 23294 style="display: none" 23295 > 23296 <div class="arrow-slidedown"> 23297 <blockquote> 23298 The Discrete Event System Specification (DEVS) 23299 formalism is a modular discrete-event modeling 23300 formalism. It has a formal specification in 23301 terms of systems theory and is supported by 23302 several efficient and usable simulator 23303 implementations. In these implementations, the 23304 DEVS formalism is often “grafted” 23305 onto an existing Object-Oriented programming 23306 language. Examples are C++ in the case of ADEVS 23307 and Python in the case of PythonPDEVS. To match 23308 this grafting, we present CLAVS, the CLAss 23309 diagram and deVS formalism and its instance 23310 counterpart ODVS, the Object Diagram and deVS 23311 formalism, and their visual notations. These 23312 languages use an automaton-like visual notation 23313 for Atomic DEVS models and a Class Diagram 23314 notation augmented with port information and 23315 event structure specification. An implementation 23316 of a visual CLAVS/ODVS modeling environment 23317 built on draw.io is presented. The use and 23318 usefulness of the formalism is demonstrated by 23319 means of a simple traffic model whose detailed 23320 specification is presented. 23321 </blockquote> 23322 </div> 23323 </div> 23324 </div> 23325 </div> 23326 <div class="slot-urls"></div> 23327 <a href="/wsc23papers/217.pdf" target="_blank">pdf</a 23328 ><br /> 23329 </div> 23330 <div class="slot-entry"> 23331 <a name="inv157" tabindex="-1"></a> 23332 <div class="slot-title-line">
23333 <span class="slot-title" 23334 >Project Simulation, Validation and Deployment with 23335 DEVS: IoT Framework for Blooms Monitoring and 23336 Alert</span 23337 > 23338 </div> 23339 <div class="slot-authors"> 23340 Segundo Esteban, Giordy A. Andrade, José L. 23341 Risco-Martín, Jesús Chacón, and Eva 23342 Besada-Portas (Complutense University of Madrid) 23343 </div> 23344 <div class="slot-abstract"> 23345 <div> 23346 <a 23347 class="clickable no-decoration" 23348 id="vhsjs_view_520_1707793552_35567" 23349 onclick="$('#vhsjs_view_520_1707793552_35567').hide(); 23350 $('#vhsjs_hide_520_1707793552_35567').show(); 23351 $('#519_1707793552_355662').slideDown(function() { 23352 if (typeof Masonry === 'function') { 23353 $('.use_masonry').masonry(); 23354 }; 23355 23356 });" 23357 ><i class="fa fa-caret-right"></i> 23358 <span class="hover_link">Abstract</span></a 23359 ><a 23360 class="clickable no-decoration" 23361 id="vhsjs_hide_520_1707793552_35567" 23362 onclick="$('#519_1707793552_355662').hide(function() { 23363 if (typeof Masonry === 'function') { 23364 $('.use_masonry').masonry(); 23365 }; 23366 }); 23367 $('#vhsjs_hide_520_1707793552_35567').hide(); 23368 $('#vhsjs_view_520_1707793552_35567').show();" 23369 style="display: none" 23370 ><i class="fa fa-caret-down"></i> 23371 <span class="hover_link">Abstract</span></a 23372 > 23373 <div 23374 data-display-control="520_1707793552_35567" 23375 id="519_1707793552_355662" 23376 style="display: none" 23377 > 23378 <div class="arrow-slidedown"> 23379 <blockquote> 23380 Harmful Algal and Cyanobacterial Blooms (HABs) 23381 constitute a relevant public health and 23382 ecological hazard due to their frequent 23383 production of toxic metabolites, which is 23384 increased by the current vulnerability of water 23385 resources to environmental changes such as 23386 global warming, population growth, and 23387 eutrophication. These blooms have been typically 23388 assessed by combining predictive models with 23389 manual collection. However, these processes are 23390 generally independent and do not provide data 23391 with sufficient resolution to apply proactive 23392 policies. In this work, we propose a novel and 23393 integrative framework to straightforwardly 23394 combine the conception, design, and deployment 23395 of advanced Early-Warning Systems (EWSs) that 23396 will allow us to automate all the processes 23397 involved in HABs detection and management and 23398 apply proactive policies. The framework is built 23399 upon solid Modeling and Simulation (M&S) 23400 principles, through Model Based Systems 23401 Engineering (MBSE) as the driving methodology 23402 and Discrete Event System Specification (DEVS) 23403 as the M&S formalism. 23404 </blockquote> 23405 </div> 23406 </div> 23407 </div> 23408 </div> 23409 <div class="slot-urls"></div> 23410 <a href="/wsc23papers/218.pdf" target="_blank">pdf</a 23411 ><br /> 23412 </div> 23413 </div> 23414 <div class="session-entry">
23415 <span class="session-event-type">Technical Session</span 23416 ><span class="type-track-spacer"> · </span 23417 ><span class="program-track">Modeling Methodology</span 23418 ><br /> 23419 <div class="session-title">Digital Twins</div> 23420 <div class="session-chair"> 23421 Chair: Claudia Szabo (University of Adelaide, The 23422 University of Adelaide)<br /> 23423 </div> 23424 <div class="slot-entry"> 23425 <a name="con183" tabindex="-1"></a> 23426 <div class="slot-title-line"> 23427 <span class="slot-title" 23428 >Automated Simulation and Virtual Reality Coupling for 23429 Interactive Digital Twins</span 23430 > 23431 </div> 23432 <div class="slot-authors"> 23433 Kai Franke, Jan Marius Stürmer, and Tobias Koch 23434 (German Aerospace Center (DLR), Institute for the 23435 Protection of Terrestrial Infrastructures) 23436 </div> 23437 <div class="slot-abstract"> 23438 <div> 23439 <a 23440 class="clickable no-decoration" 23441 id="vhsjs_view_522_1707793552_3603685" 23442 onclick="$('#vhsjs_view_522_1707793552_3603685').hide(); 23443 $('#vhsjs_hide_522_1707793552_3603685').show(); 23444 $('#521_1707793552_36036').slideDown(function() { 23445 if (typeof Masonry === 'function') { 23446 $('.use_masonry').masonry(); 23447 }; 23448 23449 });" 23450 ><i class="fa fa-caret-right"></i> 23451 <span class="hover_link">Abstract</span></a 23452 ><a 23453 class="clickable no-decoration" 23454 id="vhsjs_hide_522_1707793552_3603685" 23455 onclick="$('#521_1707793552_36036').hide(function() { 23456 if (typeof Masonry === 'function') { 23457 $('.use_masonry').masonry(); 23458 }; 23459 }); 23460 $('#vhsjs_hide_522_1707793552_3603685').hide(); 23461 $('#vhsjs_view_522_1707793552_3603685').show();" 23462 style="display: none" 23463 ><i class="fa fa-caret-down"></i> 23464 <span class="hover_link">Abstract</span></a 23465 > 23466 <div 23467 data-display-control="522_1707793552_3603685" 23468 id="521_1707793552_36036" 23469 style="display: none" 23470 > 23471 <div class="arrow-slidedown"> 23472 <blockquote> 23473 While there are many efforts to simulate 23474 technical systems in virtual environments and 23475 provide a visual interaction for applications 23476 such as training, authoring and analysis, the 23477 process of generating applications still 23478 requires a lot of manual work. This is 23479 particularly critical in the context of 23480 interactive Digital Twins for resilience, where 23481 uncertain events can occur and every malfunction 23482 or mistreatment of any part of the system needs 23483 to be modeled. This paper presents an approach 23484 to model such systems in a modular way by 23485 automating the generation of its components for 23486 a game engine and simulators based on a common 23487 specification. Component instances are then 23488 synchronized bidirectionally across applications 23489 to achieve interaction between the game engine 23490 and simulators. An example hydraulic system is 23491 implemented and tested to demonstrate our 23492 approach, which needs minimal manual work by 23493 using predefined components. The solution can be 23494 extended by integrating more components and 23495 simulations. 23496 </blockquote> 23497 </div> 23498 </div> 23499 </div> 23500 </div> 23501 <div class="slot-urls"></div> 23502 <a href="/wsc23papers/219.pdf" target="_blank">pdf</a 23503 ><br /> 23504 </div> 23505 <div class="slot-entry"> 23506 <a name="con210" tabindex="-1"></a> 23507 <div class="slot-title-line">
23508 <span class="slot-title" 23509 >Cityscape: A City-level Digital Twin Model Generator 23510 for Simulation & Analyses</span 23511 > 23512 </div> 23513 <div class="slot-authors"> 23514 Dhananjai M. Rao (Miami University) 23515 </div> 23516 <div class="slot-abstract"> 23517 <div> 23518 <a 23519 class="clickable no-decoration" 23520 id="vhsjs_view_524_1707793552_362616" 23521 onclick="$('#vhsjs_view_524_1707793552_362616').hide(); 23522 $('#vhsjs_hide_524_1707793552_362616').show(); 23523 $('#523_1707793552_3626077').slideDown(function() { 23524 if (typeof Masonry === 'function') { 23525 $('.use_masonry').masonry(); 23526 }; 23527 23528 });" 23529 ><i class="fa fa-caret-right"></i> 23530 <span class="hover_link">Abstract</span></a 23531 ><a 23532 class="clickable no-decoration" 23533 id="vhsjs_hide_524_1707793552_362616" 23534 onclick="$('#523_1707793552_3626077').hide(function() { 23535 if (typeof Masonry === 'function') { 23536 $('.use_masonry').masonry(); 23537 }; 23538 }); 23539 $('#vhsjs_hide_524_1707793552_362616').hide(); 23540 $('#vhsjs_view_524_1707793552_362616').show();" 23541 style="display: none" 23542 ><i class="fa fa-caret-down"></i> 23543 <span class="hover_link">Abstract</span></a 23544 > 23545 <div 23546 data-display-control="524_1707793552_362616" 23547 id="523_1707793552_3626077" 23548 style="display: none" 23549 > 23550 <div class="arrow-slidedown"> 23551 <blockquote> 23552 Cities and large urban areas face a myriad of 23553 challenges ranging from city planning, 23554 developing sustainable transportation, managing 23555 natural catastrophes, and mitigating 23556 communicable diseases. Addressing these 23557 challenges requires effective analysis and 23558 planning which in turn necessitates the use of 23559 sufficiently detailed models or "digital twins." 23560 Such detailed models that embody multifaceted 23561 demographic and city characteristics are 23562 challenging to generate. This paper presents our 23563 ongoing work to develop a novel model generation 23564 method and software suite called Cityscape, that 23565 fuses diverse real-world data sets to generate a 23566 digital twin for a given city. Specifically, our 23567 method combines data from authoritative sources 23568 including PUMS, PUMAs, and OpenStreet Map to 23569 generate the digital twin. We have used the city 23570 of Chicago (IL, USA) as a case study to verify 23571 and validate (with ~85% confidence) our proposed 23572 method. 23573 </blockquote> 23574 </div> 23575 </div> 23576 </div> 23577 </div> 23578 <div class="slot-urls"></div> 23579 <a href="/wsc23papers/220.pdf" target="_blank">pdf</a 23580 ><br /> 23581 </div> 23582 <div class="slot-entry"> 23583 <a name="inv150" tabindex="-1"></a> 23584 <div class="slot-title-line">
23585 <span class="slot-title" 23586 >Microscopic Vehicular Traffic Simulation: Toward 23587 Online Calibration</span 23588 > 23589 </div> 23590 <div class="slot-authors"> 23591 Yulong Wang and John Miller (University of Georgia) and 23592 Casey Bowman (University of North Georgia) 23593 </div> 23594 <div class="slot-abstract"> 23595 <div> 23596 <a 23597 class="clickable no-decoration" 23598 id="vhsjs_view_526_1707793552_3648274" 23599 onclick="$('#vhsjs_view_526_1707793552_3648274').hide(); 23600 $('#vhsjs_hide_526_1707793552_3648274').show(); 23601 $('#525_1707793552_3648193').slideDown(function() { 23602 if (typeof Masonry === 'function') { 23603 $('.use_masonry').masonry(); 23604 }; 23605 23606 });" 23607 ><i class="fa fa-caret-right"></i> 23608 <span class="hover_link">Abstract</span></a 23609 ><a 23610 class="clickable no-decoration" 23611 id="vhsjs_hide_526_1707793552_3648274" 23612 onclick="$('#525_1707793552_3648193').hide(function() { 23613 if (typeof Masonry === 'function') { 23614 $('.use_masonry').masonry(); 23615 }; 23616 }); 23617 $('#vhsjs_hide_526_1707793552_3648274').hide(); 23618 $('#vhsjs_view_526_1707793552_3648274').show();" 23619 style="display: none" 23620 ><i class="fa fa-caret-down"></i> 23621 <span class="hover_link">Abstract</span></a 23622 > 23623 <div 23624 data-display-control="526_1707793552_3648274" 23625 id="525_1707793552_3648193" 23626 style="display: none" 23627 > 23628 <div class="arrow-slidedown"> 23629 <blockquote> 23630 The modern world requires accurate and efficient 23631 traffic modeling to facilitate commerce and 23632 ensure citizens' safety. Traffic simulations 23633 play an important role in this endeavor by 23634 allowing traffic engineers to test traffic 23635 systems and policies before implementing them. 23636 This requires traffic simulation models that 23637 have the ability to accurately represent 23638 real-world traffic systems, and which are also 23639 capable of re-calibrating model parameters when 23640 needed through online calibration. This work 23641 presents four contributions toward this 23642 endeavor. The data science system ScalaTion was 23643 extended with agent-based modeling and makes use 23644 of virtual threads for each vehicle, which 23645 improves the efficiency of simulations. The 23646 modeling, simulating, and data loading schema 23647 were all optimized to enhance the system 23648 performance as well. Additionally, a new arrival 23649 model strategy was implemented improving the 23650 accuracy of the model calibration phase. 23651 </blockquote> 23652 </div> 23653 </div> 23654 </div> 23655 </div> 23656 <div class="slot-urls"></div> 23657 <a href="/wsc23papers/221.pdf" target="_blank">pdf</a 23658 ><br /> 23659 </div> 23660 </div> 23661 <div class="session-entry">
23662 <span class="session-event-type">Technical Session</span 23663 ><span class="type-track-spacer"> · </span 23664 ><span class="program-track">Modeling Methodology</span 23665 ><br /> 23666 <div class="session-title">Modeling Languages</div> 23667 <div class="session-chair"> 23668 Chair: Andrea D'Ambrogio (University of Roma 23669 TorVergata)<br /> 23670 </div> 23671 <div class="slot-entry"> 23672 <a name="con141" tabindex="-1"></a> 23673 <div class="slot-title-line"> 23674 <span class="slot-title" 23675 >FACT: A Domain Specific Language Based on a 23676 Functional Algebra for Continuous Time Modeling</span 23677 > 23678 </div> 23679 <div class="slot-authors"> 23680 Edil G. Medeiros, Eduardo Lemos, and Eduardo Peixoto 23681 (Universidade de Brasília) 23682 </div> 23683 <div class="slot-abstract"> 23684 <div> 23685 <a 23686 class="clickable no-decoration" 23687 id="vhsjs_view_528_1707793552_3692882" 23688 onclick="$('#vhsjs_view_528_1707793552_3692882').hide(); 23689 $('#vhsjs_hide_528_1707793552_3692882').show(); 23690 $('#527_1707793552_3692799').slideDown(function() { 23691 if (typeof Masonry === 'function') { 23692 $('.use_masonry').masonry(); 23693 }; 23694 23695 });" 23696 ><i class="fa fa-caret-right"></i> 23697 <span class="hover_link">Abstract</span></a 23698 ><a 23699 class="clickable no-decoration" 23700 id="vhsjs_hide_528_1707793552_3692882" 23701 onclick="$('#527_1707793552_3692799').hide(function() { 23702 if (typeof Masonry === 'function') { 23703 $('.use_masonry').masonry(); 23704 }; 23705 }); 23706 $('#vhsjs_hide_528_1707793552_3692882').hide(); 23707 $('#vhsjs_view_528_1707793552_3692882').show();" 23708 style="display: none" 23709 ><i class="fa fa-caret-down"></i> 23710 <span class="hover_link">Abstract</span></a 23711 > 23712 <div 23713 data-display-control="528_1707793552_3692882" 23714 id="527_1707793552_3692799" 23715 style="display: none" 23716 > 23717 <div class="arrow-slidedown"> 23718 <blockquote> 23719 Hybrid and cyber-physical systems create synergy 23720 by combining digital modules with analog 23721 implementations of signal processing operations 23722 typically implemented in the digital domain. We 23723 propose a domain-specific language (DSL), 23724 so-called FACT – Functional Algebra for 23725 Continuous Time, based on the algebraic 23726 properties of the General Purpose Analog 23727 Computer (GPAC), a theoretical model of 23728 computation recently updated as a continuous 23729 time equivalent of the Turing Machine. We lift 23730 the GPAC to a continuous time dynamics inside a 23731 black box semantics for understanding hybrid 23732 systems, which allows us to redefine continuous 23733 time semantics inspired by the functional 23734 reactive programming style. FACT leverages the 23735 type class mechanism from the Haskell functional 23736 programming language to implement operators that 23737 capture the proposed continuous time semantics. 23738 An speed-optimized working open-source 23739 implementation in the Haskell functional 23740 language is provided and was used to demonstrate 23741 how the language supports modeling and 23742 simulation. 23743 </blockquote> 23744 </div> 23745 </div> 23746 </div> 23747 </div> 23748 <div class="slot-urls"></div> 23749 <a href="/wsc23papers/222.pdf" target="_blank">pdf</a 23750 ><br /> 23751 </div> 23752 <div class="slot-entry"> 23753 <a name="con217" tabindex="-1"></a> 23754 <div class="slot-title-line">
23755 <span class="slot-title" 23756 >Transforming Discrete Event Models to Machine 23757 Learning Models</span 23758 > 23759 </div> 23760 <div class="slot-authors"> 23761 Hessam S. Sarjoughian, Forouzan Fallah, and 23762 Seyyedamirhossein Saeidi (Arizona State University) and 23763 Edward J. Yellig (Intel Corporation) 23764 </div> 23765 <div class="slot-abstract"> 23766 <div> 23767 <a 23768 class="clickable no-decoration" 23769 id="vhsjs_view_530_1707793552_371609" 23770 onclick="$('#vhsjs_view_530_1707793552_371609').hide(); 23771 $('#vhsjs_hide_530_1707793552_371609').show(); 23772 $('#529_1707793552_371601').slideDown(function() { 23773 if (typeof Masonry === 'function') { 23774 $('.use_masonry').masonry(); 23775 }; 23776 23777 });" 23778 ><i class="fa fa-caret-right"></i> 23779 <span class="hover_link">Abstract</span></a 23780 ><a 23781 class="clickable no-decoration" 23782 id="vhsjs_hide_530_1707793552_371609" 23783 onclick="$('#529_1707793552_371601').hide(function() { 23784 if (typeof Masonry === 'function') { 23785 $('.use_masonry').masonry(); 23786 }; 23787 }); 23788 $('#vhsjs_hide_530_1707793552_371609').hide(); 23789 $('#vhsjs_view_530_1707793552_371609').show();" 23790 style="display: none" 23791 ><i class="fa fa-caret-down"></i> 23792 <span class="hover_link">Abstract</span></a 23793 > 23794 <div 23795 data-display-control="530_1707793552_371609" 23796 id="529_1707793552_371601" 23797 style="display: none" 23798 > 23799 <div class="arrow-slidedown"> 23800 <blockquote> 23801 Discrete event simulation, formalized as 23802 deductive modeling, has been shown to be 23803 effective for studying dynamical systems. 23804 Development of models, however, is challenging 23805 when numerous interacting components are 23806 involved and should operate under different 23807 conditions. Machine Learning (ML) holds the 23808 promise to help reduce the effort needed to 23809 develop models. Toward this goal, a collection 23810 of ML algorithms, including Automatic Relevance 23811 Determination are used. Parallel Discrete Event 23812 System Specification (PDEVS) models are 23813 developed for Single-stage and Two-stage cascade 23814 factories. Each model is simulated under 23815 different demand profiles. The simulated data 23816 sets are partitioned into subsets, each for one 23817 or more model components. The ML algorithms are 23818 applied to the data sets for generating models. 23819 The throughputs predicted by the ML models 23820 closely match those in the PDEVS simulated data. 23821 This study contributes to modeling by 23822 demonstrating the potential benefits and 23823 complications of utilizing ML for discrete-event 23824 systems. 23825 </blockquote> 23826 </div> 23827 </div> 23828 </div> 23829 </div> 23830 <div class="slot-urls"></div> 23831 <a href="/wsc23papers/223.pdf" target="_blank">pdf</a 23832 ><br /> 23833 </div> 23834 <div class="slot-entry"> 23835 <a name="con249" tabindex="-1"></a> 23836 <div class="slot-title-line">
23837 <span class="slot-title" 23838 >Validation without Data - Formalizing Stylized Facts 23839 of Time Series</span 23840 > 23841 </div> 23842 <div class="slot-authors"> 23843 Pia Wilsdorf, Marian Zuska, Philipp Andelfinger, Florian 23844 Peters, and Adelinde Uhrmacher (University of Rostock) 23845 </div> 23846 <div class="slot-abstract"> 23847 <div> 23848 <a 23849 class="clickable no-decoration" 23850 id="vhsjs_view_532_1707793552_373991" 23851 onclick="$('#vhsjs_view_532_1707793552_373991').hide(); 23852 $('#vhsjs_hide_532_1707793552_373991').show(); 23853 $('#531_1707793552_373983').slideDown(function() { 23854 if (typeof Masonry === 'function') { 23855 $('.use_masonry').masonry(); 23856 }; 23857 23858 });" 23859 ><i class="fa fa-caret-right"></i> 23860 <span class="hover_link">Abstract</span></a 23861 ><a 23862 class="clickable no-decoration" 23863 id="vhsjs_hide_532_1707793552_373991" 23864 onclick="$('#531_1707793552_373983').hide(function() { 23865 if (typeof Masonry === 'function') { 23866 $('.use_masonry').masonry(); 23867 }; 23868 }); 23869 $('#vhsjs_hide_532_1707793552_373991').hide(); 23870 $('#vhsjs_view_532_1707793552_373991').show();" 23871 style="display: none" 23872 ><i class="fa fa-caret-down"></i> 23873 <span class="hover_link">Abstract</span></a 23874 > 23875 <div 23876 data-display-control="532_1707793552_373991" 23877 id="531_1707793552_373983" 23878 style="display: none" 23879 > 23880 <div class="arrow-slidedown"> 23881 <blockquote> 23882 A stylized fact is a simplified presentation of 23883 an empirical finding. When modeling and 23884 simulating complex systems and real data are 23885 sparse, stylized facts have become a key 23886 instrument for building trust in a model as they 23887 represent important requirements regarding the 23888 model’s behavior. However, automatically 23889 validating stylized facts has remained limited 23890 as they are usually expressed in natural 23891 language. Therefore, we develop a formal 23892 language with a custom syntax and tailored 23893 predicates allowing modelers to unambiguously 23894 and succinctly describe important (temporal) 23895 characteristics of simulation traces or 23896 relationships between multiple traces via 23897 statistical tests. The proposed formal language 23898 is able to express numerous facts from the 23899 literature in different application domains, as 23900 well as to automatically check stylized facts. 23901 If stylized facts are defined at the beginning 23902 of a simulation study, formally expressing and 23903 checking them can streamline and guide the 23904 development of simulation models and their 23905 successive revisions. 23906 </blockquote> 23907 </div> 23908 </div> 23909 </div> 23910 </div> 23911 <div class="slot-urls"></div> 23912 <a href="/wsc23papers/224.pdf" target="_blank">pdf</a 23913 ><br /> 23914 </div> 23915 </div> 23916 </div> 23917 <div class="centered"> 23918 <div class="top-link"><a href="#top">Return to Top</a></div> 23919 </div> 23920 <hr /> 23921 </div> 23922 <div class="area-section"> 23923 <div class="centered"> 23924 <a name="ptrack131" tabindex="-1"></a> 23925 <div class="section-title">Professional Development</div> 23926 </div> 23927 <div class="centered track-chair">
23928 <span class="track-chair-role" 23929 >Track Coordinator - Professional Development: </span 23930 ><span class="track-chair-names" 23931 >Thomas Berg (The University of Tennessee, Knoxville), 23932 Weiwei Chen (Rutgers University)</span 23933 > 23934 </div> 23935 <div class="section-entry"> 23936 <div class="session-entry"> 23937 <span class="session-event-type">Technical Session</span 23938 ><span class="type-track-spacer"> · </span 23939 ><span class="program-track">Professional Development</span 23940 ><br /> 23941 <div class="session-title"> 23942 Panel: Navigating Publication Outlets for Simulation 23943 Research: Insights from Journal Editors 23944 </div> 23945 <div class="session-chair"> 23946 Chair: Thomas Berg (The University of Tennessee, 23947 Knoxville)<br /> 23948 </div> 23949 <div class="slot-entry"> 23950 <a name="cea162" tabindex="-1"></a> 23951 <div class="slot-title-line"> 23952 <span class="slot-title" 23953 >Navigating Publication Outlets for Simulation 23954 Research: Insights from Journal Editors</span 23955 > 23956 </div> 23957 <div class="slot-authors"> 23958 Tom Berg (The University of Tennessee, Knoxville); Jose 23959 Blanchet (Stanford University); Christine Currie 23960 (University of Southampton); Weiwei Chen (Rutgers 23961 University); Peter Haas (University of Massachusetts 23962 Amherst); Jeff Hong (Fudan University); Bruno Tuffin 23963 (University of Rennes); and Jie Xu (George Mason 23964 University) 23965 </div> 23966 <div class="slot-abstract"> 23967 <div> 23968 <a 23969 class="clickable no-decoration" 23970 id="vhsjs_view_756_1707793552_87369" 23971 onclick="$('#vhsjs_view_756_1707793552_87369').hide(); 23972 $('#vhsjs_hide_756_1707793552_87369').show(); 23973 $('#755_1707793552_8736815').slideDown(function() { 23974 if (typeof Masonry === 'function') { 23975 $('.use_masonry').masonry(); 23976 }; 23977 23978 });" 23979 ><i class="fa fa-caret-right"></i> 23980 <span class="hover_link">Abstract</span></a 23981 ><a 23982 class="clickable no-decoration" 23983 id="vhsjs_hide_756_1707793552_87369" 23984 onclick="$('#755_1707793552_8736815').hide(function() { 23985 if (typeof Masonry === 'function') { 23986 $('.use_masonry').masonry(); 23987 }; 23988 }); 23989 $('#vhsjs_hide_756_1707793552_87369').hide(); 23990 $('#vhsjs_view_756_1707793552_87369').show();" 23991 style="display: none" 23992 ><i class="fa fa-caret-down"></i> 23993 <span class="hover_link">Abstract</span></a 23994 > 23995 <div 23996 data-display-control="756_1707793552_87369" 23997 id="755_1707793552_8736815" 23998 style="display: none" 23999 > 24000 <div class="arrow-slidedown"> 24001 <blockquote> 24002 This panel discussion is designed to provide 24003 young scholars in the field of simulation with 24004 valuable insights into identifying suitable 24005 publication avenues for their research 24006 endeavors. Senior journal editors will serve as 24007 panelists and share their wealth of experience
24008 and perspectives. Journals represented include 24009 ACM TOMACS, IISE Transactions, INFORMS Journal 24010 on Computing, Journal of Simulation, Operations 24011 Research, and Stochastic Systems. Specifically, 24012 the panelists will introduce preferred topics, 24013 focuses, and future trends for each journal. 24014 Panelists will also share their own experiences 24015 and suggestions on the peer review process, such 24016 as how to navigate through revisions and 24017 rejections, and ethical policies. Young scholars 24018 will also learn the importance of serving the 24019 community as a reviewer, and what senior editors 24020 expect from reviewers. 24021 </blockquote> 24022 </div> 24023 </div> 24024 </div> 24025 </div> 24026 <div class="slot-urls"></div> 24027 <a href="/wsc23papers/cea162.pdf" target="_blank">pdf</a 24028 ><br /> 24029 </div> 24030 </div> 24031 </div> 24032 <div class="centered"> 24033 <div class="top-link"><a href="#top">Return to Top</a></div> 24034 </div> 24035 <hr /> 24036 </div> 24037 24038 <div class="area-section"> 24039 <div class="centered"> 24040 <a name="ptrack118" tabindex="-1"></a> 24041 <div class="section-title"> 24042 Project Management and Construction 24043 </div> 24044 </div> 24045 <div class="centered track-chair"> 24046 <span class="track-chair-role" 24047 >Track Coordinator - Project Management and Construction: </span 24048 ><span class="track-chair-names" 24049 >Jing Du (University of Florida), Joseph Louis (Oregon State 24050 University)</span 24051 > 24052 </div> 24053 <div class="section-entry"> 24054 <div class="session-entry"> 24055 <span class="session-event-type">Technical Session</span 24056 ><span class="type-track-spacer"> · </span 24057 ><span class="program-track" 24058 >Project Management and Construction</span 24059 ><br /> 24060 <div class="session-title"> 24061 Health, Safety, and Sustainability in Construction 24062 </div> 24063 <div class="session-chair"> 24064 Chair: Shuai Li (the University of Tennessee)<br /> 24065 </div> 24066 <div class="slot-entry"> 24067 <a name="con250" tabindex="-1"></a> 24068 <div class="slot-title-line"> 24069 <span class="slot-title" 24070 >Simulation Modeling for Sustainable Construction: A 24071 Case Study to Highlight the Social Aspect</span 24072 > 24073 </div> 24074 <div class="slot-authors"> 24075 Mai Ghazal, Fatemeh Parvaneh, Ahmed Hammad, and Yasser 24076 Mohamed (University of Alberta) 24077 </div> 24078 <div class="slot-abstract"> 24079 <div> 24080 <a 24081 class="clickable no-decoration" 24082 id="vhsjs_view_534_1707793552_383451" 24083 onclick="$('#vhsjs_view_534_1707793552_383451').hide(); 24084 $('#vhsjs_hide_534_1707793552_383451').show(); 24085 $('#533_1707793552_3834426').slideDown(function() { 24086 if (typeof Masonry === 'function') { 24087 $('.use_masonry').masonry(); 24088 }; 24089 24090 });" 24091 ><i class="fa fa-caret-right"></i>
24092 <span class="hover_link">Abstract</span></a 24093 ><a 24094 class="clickable no-decoration" 24095 id="vhsjs_hide_534_1707793552_383451" 24096 onclick="$('#533_1707793552_3834426').hide(function() { 24097 if (typeof Masonry === 'function') { 24098 $('.use_masonry').masonry(); 24099 }; 24100 }); 24101 $('#vhsjs_hide_534_1707793552_383451').hide(); 24102 $('#vhsjs_view_534_1707793552_383451').show();" 24103 style="display: none" 24104 ><i class="fa fa-caret-down"></i> 24105 <span class="hover_link">Abstract</span></a 24106 > 24107 <div 24108 data-display-control="534_1707793552_383451" 24109 id="533_1707793552_3834426" 24110 style="display: none" 24111 > 24112 <div class="arrow-slidedown"> 24113 <blockquote> 24114 To cut costs and drive innovation in product 24115 development, many projects have turned to remote 24116 worksites for construction component 24117 pre-fabrication. Fabricating pipe spools in 24118 shops eliminates delays due to weather and 24119 allows for better resource planning. This paper 24120 aims to optimize labor resource usage in a pipe 24121 spool manufacturing plant that fabricates three 24122 different types of spools. It utilizes 24123 historical data to implement a discrete-event 24124 simulation model. The proposed simulation model 24125 effectively reduced idle time and evenly 24126 distributed the workload. As a result, the 24127 overall fabrication time for all three spools 24128 was reduced, leading to a 22% decrease in active 24129 shop usage. This allowed subsequent jobs to 24130 commence earlier, giving the team more 24131 flexibility in meeting deadlines and addressing 24132 labor constraints. This research provides 24133 insights into how resource allocation plans can 24134 be created to maximize sustainability results, 24135 both socially (through improving working 24136 conditions and reducing workloads) and 24137 economically. 24138 </blockquote> 24139 </div> 24140 </div> 24141 </div> 24142 </div> 24143 <div class="slot-urls"></div> 24144 <a href="/wsc23papers/225.pdf" target="_blank">pdf</a 24145 ><br /> 24146 </div> 24147 <div class="slot-entry"> 24148 <a name="con313" tabindex="-1"></a> 24149 <div class="slot-title-line"> 24150 <span class="slot-title" 24151 >The Impact of Alcohol Use on Construction Safety 24152 Outcomes: An Agent-Based Modeling Investigation</span 24153 > 24154 </div> 24155 <div class="slot-authors"> 24156 Christin Manning and Ehsan Salari (Wichita State 24157 University) 24158 </div> 24159 <div class="slot-abstract"> 24160 <div> 24161 <a 24162 class="clickable no-decoration" 24163 id="vhsjs_view_536_1707793552_3856196" 24164 onclick="$('#vhsjs_view_536_1707793552_3856196').hide(); 24165 $('#vhsjs_hide_536_1707793552_3856196').show(); 24166 $('#535_1707793552_3856115').slideDown(function() { 24167 if (typeof Masonry === 'function') { 24168 $('.use_masonry').masonry(); 24169 }; 24170 24171 });" 24172 ><i class="fa fa-caret-right"></i>
24173 <span class="hover_link">Abstract</span></a 24174 ><a 24175 class="clickable no-decoration" 24176 id="vhsjs_hide_536_1707793552_3856196" 24177 onclick="$('#535_1707793552_3856115').hide(function() { 24178 if (typeof Masonry === 'function') { 24179 $('.use_masonry').masonry(); 24180 }; 24181 }); 24182 $('#vhsjs_hide_536_1707793552_3856196').hide(); 24183 $('#vhsjs_view_536_1707793552_3856196').show();" 24184 style="display: none" 24185 ><i class="fa fa-caret-down"></i> 24186 <span class="hover_link">Abstract</span></a 24187 > 24188 <div 24189 data-display-control="536_1707793552_3856196" 24190 id="535_1707793552_3856115" 24191 style="display: none" 24192 > 24193 <div class="arrow-slidedown"> 24194 <blockquote> 24195 Construction is a notoriously hazardous industry 24196 and heavy alcohol use is common. This project 24197 creates an agent-based modeling (ABM) simulation 24198 exploring the impact of alcohol on safety 24199 outcomes. Simulation modeling is useful in 24200 occupational safety research because it 24201 generates immediate results and bypasses ethical 24202 concerns. Workers and foremen interact on a 24203 virtual jobsite with hazards present. Positive 24204 blood alcohol concentration (BAC) decreases 24205 hazard awareness and reaction time, and 24206 additionally decreases competency of foremen. 24207 Scenarios of baseline, increased, and decreased 24208 alcohol consumption are analyzed for changes in 24209 near misses, injuries, and fatalities. 24210 Additional scenarios of improved training and 24211 engineering controls are explored also for 24212 comparison. A decrease in alcohol consumption 24213 led to a significant reduction in injuries by up 24214 to 12%, and an increase had the opposite effect. 24215 Neither scenario significantly impacted 24216 fatalities due to fatalities' low base rate. 24217 Safety training had a comparable impact but 24218 improving engineering controls outweighed both. 24219 </blockquote> 24220 </div> 24221 </div> 24222 </div> 24223 </div> 24224 <div class="slot-urls"></div> 24225 <a href="/wsc23papers/226.pdf" target="_blank">pdf</a 24226 ><br /> 24227 </div> 24228 <div class="slot-entry"> 24229 <a name="con123" tabindex="-1"></a> 24230 <div class="slot-title-line"> 24231 <span class="slot-title" 24232 >3D Object Detection and Localization within 24233 Healthcare Facilities</span 24234 > 24235 </div> 24236 <div class="slot-authors"> 24237 Da Hu (Kennesaw State University) and Mengjun Wang and 24238 Shuai Li (University of Tennessee) 24239 </div> 24240 <div class="slot-abstract"> 24241 <div> 24242 <a 24243 class="clickable no-decoration" 24244 id="vhsjs_view_538_1707793552_3878474" 24245 onclick="$('#vhsjs_view_538_1707793552_3878474').hide(); 24246 $('#vhsjs_hide_538_1707793552_3878474').show(); 24247 $('#537_1707793552_3878396').slideDown(function() { 24248 if (typeof Masonry === 'function') { 24249 $('.use_masonry').masonry(); 24250 }; 24251 24252 });" 24253 ><i class="fa fa-caret-right"></i>
24254 <span class="hover_link">Abstract</span></a 24255 ><a 24256 class="clickable no-decoration" 24257 id="vhsjs_hide_538_1707793552_3878474" 24258 onclick="$('#537_1707793552_3878396').hide(function() { 24259 if (typeof Masonry === 'function') { 24260 $('.use_masonry').masonry(); 24261 }; 24262 }); 24263 $('#vhsjs_hide_538_1707793552_3878474').hide(); 24264 $('#vhsjs_view_538_1707793552_3878474').show();" 24265 style="display: none" 24266 ><i class="fa fa-caret-down"></i> 24267 <span class="hover_link">Abstract</span></a 24268 > 24269 <div 24270 data-display-control="538_1707793552_3878474" 24271 id="537_1707793552_3878396" 24272 style="display: none" 24273 > 24274 <div class="arrow-slidedown"> 24275 <blockquote> 24276 This study introduces a deep learning-based 24277 method for indoor 3D object detection and 24278 localization in healthcare facilities. This 24279 method incorporates spatial and channel 24280 attention mechanisms into the YOLOv5 24281 architecture, ensuring a balance between 24282 accuracy and computational efficiency. The 24283 network achieves an AP50 of 67.6%, an mAP of 24284 46.7%, and a real-time detection rate with an 24285 FPS of 67. Moreover, the study proposes a novel 24286 mechanism for estimating the 3D coordinates of 24287 detected objects and projecting them onto 3D 24288 maps, with an average error of 0.24 m and 0.28 m 24289 in the x and y directions, respectively. After 24290 being tested and validated with real-world data 24291 from a university campus, the proposed method 24292 shows promise for improving disinfection 24293 efficiency in healthcare facilities by enabling 24294 real-time object detection and localization for 24295 robot navigation. 24296 </blockquote> 24297 </div> 24298 </div> 24299 </div> 24300 </div> 24301 <div class="slot-urls"></div> 24302 <a href="/wsc23papers/227.pdf" target="_blank">pdf</a 24303 ><br /> 24304 </div> 24305 </div> 24306 <div class="session-entry"> 24307 <span class="session-event-type">Technical Session</span 24308 ><span class="type-track-spacer"> · </span 24309 ><span class="program-track" 24310 >Project Management and Construction</span 24311 ><br /> 24312 <div class="session-title"> 24313 Technological Innovations for Enhanced Construction 24314 Operations 24315 </div> 24316 <div class="session-chair"> 24317 Chair: Shuai Li (the University of Tennessee)<br /> 24318 </div> 24319 <div class="slot-entry"> 24320 <a name="con367" tabindex="-1"></a> 24321 <div class="slot-title-line"> 24322 <span class="slot-title" 24323 >Applying Civil Information Modeling and Augmented 24324 Reality to the Construction of Underground 24325 Pipelines</span 24326 > 24327 </div> 24328 <div class="slot-authors"> 24329 Andy Cui (Montgomery Blair High School) and Man Liang 24330 (University of Maryland) 24331 </div> 24332 <div class="slot-abstract"> 24333 <div> 24334 <a 24335 class="clickable no-decoration" 24336 id="vhsjs_view_540_1707793552_393342" 24337 onclick="$('#vhsjs_view_540_1707793552_393342').hide(); 24338 $('#vhsjs_hide_540_1707793552_393342').show(); 24339 $('#539_1707793552_3933334').slideDown(function() { 24340 if (typeof Masonry === 'function') { 24341 $('.use_masonry').masonry(); 24342 }; 24343 24344 });" 24345 ><i class="fa fa-caret-right"></i>
24346 <span class="hover_link">Abstract</span></a 24347 ><a 24348 class="clickable no-decoration" 24349 id="vhsjs_hide_540_1707793552_393342" 24350 onclick="$('#539_1707793552_3933334').hide(function() { 24351 if (typeof Masonry === 'function') { 24352 $('.use_masonry').masonry(); 24353 }; 24354 }); 24355 $('#vhsjs_hide_540_1707793552_393342').hide(); 24356 $('#vhsjs_view_540_1707793552_393342').show();" 24357 style="display: none" 24358 ><i class="fa fa-caret-down"></i> 24359 <span class="hover_link">Abstract</span></a 24360 > 24361 <div 24362 data-display-control="540_1707793552_393342" 24363 id="539_1707793552_3933334" 24364 style="display: none" 24365 > 24366 <div class="arrow-slidedown"> 24367 <blockquote> 24368 Municipal construction projects are often 24369 challenging and risk-prone due to unexpected 24370 underground conditions. Access to As-Built and 24371 As-Design data is essential to avoid budget 24372 overruns, schedule delays, and other 24373 construction disputes. However, coordinating 24374 field conditions with construction drawings can 24375 be difficult and lead to discrepancies. 24376 Traditional methods of denoting information onto 24377 the ground by surveyors and field workers have 24378 been limited in their ability to provide 24379 relevant information and support scaling up. 24380 These methods also create restrictions in data 24381 sharing and communication among workers and 24382 engineering teams. With the development and use 24383 of AR technology, our study proposes an 24384 augmented reality tool leveraging Google ARCore 24385 to assist construction engineers in a 24386 straightforward and efficient manner by 24387 displaying utility information, including pipe 24388 direction, type, slope, diameter, and material. 24389 The campus area of the University of Maryland 24390 College Park is used as a case study to 24391 demonstrate our approach. 24392 </blockquote> 24393 </div> 24394 </div> 24395 </div> 24396 </div> 24397 <div class="slot-urls"></div> 24398 <a href="/wsc23papers/229.pdf" target="_blank">pdf</a 24399 ><br /> 24400 </div> 24401 <div class="slot-entry"> 24402 <a name="con189" tabindex="-1"></a> 24403 <div class="slot-title-line"> 24404 <span class="slot-title" 24405 >A Value Stream Mapping-Based Discrete Event 24406 Simulation Template for Lean Off-Site Construction 24407 Activities</span 24408 > 24409 </div> 24410 <div class="slot-authors"> 24411 Prashanth Kumar Sreram (Indian Institute of Technology 24412 Bombay, NICMAR Hyderabad) and Albert Thomas (Indian 24413 Institute of Technology Bombay) 24414 </div> 24415 <div class="slot-abstract"> 24416 <div> 24417 <a 24418 class="clickable no-decoration" 24419 id="vhsjs_view_542_1707793552_3955271" 24420 onclick="$('#vhsjs_view_542_1707793552_3955271').hide(); 24421 $('#vhsjs_hide_542_1707793552_3955271').show(); 24422 $('#541_1707793552_395519').slideDown(function() { 24423 if (typeof Masonry === 'function') { 24424 $('.use_masonry').masonry(); 24425 }; 24426 24427 });" 24428 ><i class="fa fa-caret-right"></i>
24429 <span class="hover_link">Abstract</span></a 24430 ><a 24431 class="clickable no-decoration" 24432 id="vhsjs_hide_542_1707793552_3955271" 24433 onclick="$('#541_1707793552_395519').hide(function() { 24434 if (typeof Masonry === 'function') { 24435 $('.use_masonry').masonry(); 24436 }; 24437 }); 24438 $('#vhsjs_hide_542_1707793552_3955271').hide(); 24439 $('#vhsjs_view_542_1707793552_3955271').show();" 24440 style="display: none" 24441 ><i class="fa fa-caret-down"></i> 24442 <span class="hover_link">Abstract</span></a 24443 > 24444 <div 24445 data-display-control="542_1707793552_3955271" 24446 id="541_1707793552_395519" 24447 style="display: none" 24448 > 24449 <div class="arrow-slidedown"> 24450 <blockquote> 24451 Lean construction is a promising approach for 24452 performance improvement in the construction 24453 industry. Value stream mapping (VSM) is an 24454 essential lean tool for the process improvement 24455 of construction activities. However, VSM, 24456 regarded as a static pen-and-paper technique, 24457 requires repeating the VSM preparation for every 24458 improvement alternative. Therefore, dynamism can 24459 be introduced into VSM by developing computer 24460 simulation models, which is the study's 24461 objective. A VSM-based discrete event simulation 24462 (DES) template is presented in this paper for 24463 off-site construction activities. The model 24464 provides a virtual testing environment for the 24465 user to decide upon the potential time reduction 24466 in non-value-added (NVA) activities for the 24467 process improvement. The development and 24468 validation of the model is done based on the 24469 actual data from a precast production factory. 24470 The DES-VSM simulation model assists plant 24471 managers with the best possible NVA reduction 24472 strategy and accelerates lean implementation in 24473 the construction industry. 24474 </blockquote> 24475 </div> 24476 </div> 24477 </div> 24478 </div> 24479 <div class="slot-urls"></div> 24480 <a href="/wsc23papers/232.pdf" target="_blank">pdf</a 24481 ><br /> 24482 </div> 24483 </div> 24484 <div class="session-entry"> 24485 <span class="session-event-type">Technical Session</span 24486 ><span class="type-track-spacer"> · </span 24487 ><span class="program-track" 24488 >Project Management and Construction</span 24489 ><br /> 24490 <div class="session-title"> 24491 Advanced Simulation Methods in Construction 24492 </div> 24493 <div class="session-chair"> 24494 Chair: Albert Thomas (Indian Institute of Technology 24495 Bombay)<br /> 24496 </div> 24497 <div class="slot-entry"> 24498 <a name="con162" tabindex="-1"></a> 24499 <div class="slot-title-line">
24500 <span class="slot-title" 24501 >New Functions and Statements to Support Preemption in 24502 the STROBOSCOPE Simulation System</span 24503 > 24504 </div> 24505 <div class="slot-authors"> 24506 Photios G. Ioannou (University of Michigan) and Veerasak 24507 Likhitruangsilp (Chulalongkorn University) 24508 </div> 24509 <div class="slot-abstract"> 24510 <div> 24511 <a 24512 class="clickable no-decoration" 24513 id="vhsjs_view_544_1707793552_399803" 24514 onclick="$('#vhsjs_view_544_1707793552_399803').hide(); 24515 $('#vhsjs_hide_544_1707793552_399803').show(); 24516 $('#543_1707793552_3997943').slideDown(function() { 24517 if (typeof Masonry === 'function') { 24518 $('.use_masonry').masonry(); 24519 }; 24520 24521 });" 24522 ><i class="fa fa-caret-right"></i> 24523 <span class="hover_link">Abstract</span></a 24524 ><a 24525 class="clickable no-decoration" 24526 id="vhsjs_hide_544_1707793552_399803" 24527 onclick="$('#543_1707793552_3997943').hide(function() { 24528 if (typeof Masonry === 'function') { 24529 $('.use_masonry').masonry(); 24530 }; 24531 }); 24532 $('#vhsjs_hide_544_1707793552_399803').hide(); 24533 $('#vhsjs_view_544_1707793552_399803').show();" 24534 style="display: none" 24535 ><i class="fa fa-caret-down"></i> 24536 <span class="hover_link">Abstract</span></a 24537 > 24538 <div 24539 data-display-control="544_1707793552_399803" 24540 id="543_1707793552_3997943" 24541 style="display: none" 24542 > 24543 <div class="arrow-slidedown"> 24544 <blockquote> 24545 The new preemption capabilities added to the 24546 STROBOSCOPE simulation system are described and 24547 illustrated by two examples. The first example 24548 involves moving soil using two wheelbarrows and 24549 two laborers. It investigates the conditions for 24550 preemption to improve production by allowing the 24551 return of an empty wheelbarrow to interrupt 24552 loading and to start hauling a partially loaded 24553 wheelbarrow immediately. In the second example, 24554 two cranes unload barges bringing fill material 24555 for undersea land reclamation. When only one 24556 barge is available, it can unload using both 24557 cranes. When two or more barges become 24558 available, each barge unloads using one crane. 24559 Unloading a barge can switch between using one 24560 and two cranes multiple times, with the 24561 remaining unload time either cut in half or 24562 doubled each time. Modeling the multiple 24563 reallocations of cranes and the required time 24564 adjustments illustrates the new STROBOSCOPE 24565 preemption capabilities. 24566 </blockquote> 24567 </div> 24568 </div> 24569 </div> 24570 </div> 24571 <div class="slot-urls"></div> 24572 <a href="/wsc23papers/230.pdf" target="_blank">pdf</a 24573 ><br /> 24574 </div> 24575 <div class="slot-entry"> 24576 <a name="con163" tabindex="-1"></a> 24577 <div class="slot-title-line">
24578 <span class="slot-title" 24579 >Simulation of Earthmoving for a Dam Using Engineering 24580 Calculations</span 24581 > 24582 </div> 24583 <div class="slot-authors"> 24584 Photios G. Ioannou (University of Michigan) 24585 </div> 24586 <div class="slot-abstract"> 24587 <div> 24588 <a 24589 class="clickable no-decoration" 24590 id="vhsjs_view_546_1707793552_4018786" 24591 onclick="$('#vhsjs_view_546_1707793552_4018786').hide(); 24592 $('#vhsjs_hide_546_1707793552_4018786').show(); 24593 $('#545_1707793552_4018703').slideDown(function() { 24594 if (typeof Masonry === 'function') { 24595 $('.use_masonry').masonry(); 24596 }; 24597 24598 });" 24599 ><i class="fa fa-caret-right"></i> 24600 <span class="hover_link">Abstract</span></a 24601 ><a 24602 class="clickable no-decoration" 24603 id="vhsjs_hide_546_1707793552_4018786" 24604 onclick="$('#545_1707793552_4018703').hide(function() { 24605 if (typeof Masonry === 'function') { 24606 $('.use_masonry').masonry(); 24607 }; 24608 }); 24609 $('#vhsjs_hide_546_1707793552_4018786').hide(); 24610 $('#vhsjs_view_546_1707793552_4018786').show();" 24611 style="display: none" 24612 ><i class="fa fa-caret-down"></i> 24613 <span class="hover_link">Abstract</span></a 24614 > 24615 <div 24616 data-display-control="546_1707793552_4018786" 24617 id="545_1707793552_4018703" 24618 style="display: none" 24619 > 24620 <div class="arrow-slidedown"> 24621 <blockquote> 24622 Detailed STROBOSCOPE simulations of earthmoving 24623 for the construction of a dam use the 24624 engineering calculations typically employed in 24625 heavy construction to estimate equipment 24626 performance based on the characteristics of the 24627 haul and return roads and the mechanical 24628 properties of actual models of heavy loaders and 24629 trucks. Sensitivity analysis investigates the 24630 total cost of truck combinations while 24631 considering the traffic effects of one or two 24632 bridges needed to cross a river along the haul 24633 route. This example can serve as a simulation 24634 model template to facilitate the wider 24635 acceptance of simulation in heavy construction 24636 practice. 24637 </blockquote> 24638 </div> 24639 </div> 24640 </div> 24641 </div> 24642 <div class="slot-urls"></div> 24643 <a href="/wsc23papers/231.pdf" target="_blank">pdf</a 24644 ><br /> 24645 </div> 24646 </div> 24647 <div class="session-entry"> 24648 <span class="session-event-type">Technical Session</span 24649 ><span class="type-track-spacer"> · </span 24650 ><span class="program-track" 24651 >Project Management and Construction</span 24652 ><br /> 24653 <div class="session-title"> 24654 Strategic Modeling and Decision Making in Construction 24655 </div> 24656 <div class="session-chair"> 24657 Chair: Gabriel Castelblanco (University of Florida)<br /> 24658 </div> 24659 <div class="slot-entry"> 24660 <a name="con317" tabindex="-1"></a> 24661 <div class="slot-title-line">
24662 <span class="slot-title" 24663 >Enhancing the Public Investment in Public-Private 24664 Partnerships Using System Dynamics Modeling</span 24665 > 24666 </div> 24667 <div class="slot-authors"> 24668 Sara Biziorek and Alberto De Marco (Politecnico di 24669 Torino), Jose Guevara (Universidad de los Andes), and 24670 Gabriel Castelblanco (University of Florida) 24671 </div> 24672 <div class="slot-abstract"> 24673 <div> 24674 <a 24675 class="clickable no-decoration" 24676 id="vhsjs_view_548_1707793552_4066226" 24677 onclick="$('#vhsjs_view_548_1707793552_4066226').hide(); 24678 $('#vhsjs_hide_548_1707793552_4066226').show(); 24679 $('#547_1707793552_4066145').slideDown(function() { 24680 if (typeof Masonry === 'function') { 24681 $('.use_masonry').masonry(); 24682 }; 24683 24684 });" 24685 ><i class="fa fa-caret-right"></i> 24686 <span class="hover_link">Abstract</span></a 24687 ><a 24688 class="clickable no-decoration" 24689 id="vhsjs_hide_548_1707793552_4066226" 24690 onclick="$('#547_1707793552_4066145').hide(function() { 24691 if (typeof Masonry === 'function') { 24692 $('.use_masonry').masonry(); 24693 }; 24694 }); 24695 $('#vhsjs_hide_548_1707793552_4066226').hide(); 24696 $('#vhsjs_view_548_1707793552_4066226').show();" 24697 style="display: none" 24698 ><i class="fa fa-caret-down"></i> 24699 <span class="hover_link">Abstract</span></a 24700 > 24701 <div 24702 data-display-control="548_1707793552_4066226" 24703 id="547_1707793552_4066145" 24704 style="display: none" 24705 > 24706 <div class="arrow-slidedown"> 24707 <blockquote> 24708 Public-Private Partnership (PPP) programs have 24709 been adopted to leverage private funding for the 24710 development of public infrastructure and 24711 services, thereby relieving public fiscal 24712 pressure. However, the complexity and length of 24713 PPP contracts can lead to higher costs for the 24714 public sector. Using data from more than 700 24715 PPPs that integrate the UK Private Finance 24716 Initiative and Private Finance 2 programs, this 24717 study analyzes the long-term financial 24718 implications of these programs using System 24719 Dynamics. Causal-loop diagrams were developed to 24720 illustrate the causal structures that generate 24721 the long-term financial effects of PPPs on the 24722 public sector. The paper offers potential 24723 strategies to enhance the performance of PPP 24724 programs. This study contributes to closing the 24725 research gap identified in previous research for 24726 more efficient PPP programs by uncovering their 24727 dynamics and offering suitable policies for 24728 governments to improve their outcomes. 24729 </blockquote> 24730 </div> 24731 </div> 24732 </div> 24733 </div> 24734 <div class="slot-urls"></div> 24735 <a href="/wsc23papers/233.pdf" target="_blank">pdf</a 24736 ><br /> 24737 </div> 24738 <div class="slot-entry"> 24739 <a name="con370" tabindex="-1"></a> 24740 <div class="slot-title-line">
24741 <span class="slot-title" 24742 >A Discrete-Event Simulation to Explore Disaggregation 24743 of Biotechnology Research and Development 24744 Workflows</span 24745 > 24746 </div> 24747 <div class="slot-authors"> 24748 Susan S.M. Hanson, Noah Mecikalski, Alex Tobias, Jack 24749 Morris, Neal Wagner, and Rebecca S. Widrick (MITRE 24750 Corporation) and Damon Bayer (University of California 24751 Irvine) 24752 </div> 24753 <div class="slot-abstract"> 24754 <div> 24755 <a 24756 class="clickable no-decoration" 24757 id="vhsjs_view_550_1707793552_4091241" 24758 onclick="$('#vhsjs_view_550_1707793552_4091241').hide(); 24759 $('#vhsjs_hide_550_1707793552_4091241').show(); 24760 $('#549_1707793552_4091156').slideDown(function() { 24761 if (typeof Masonry === 'function') { 24762 $('.use_masonry').masonry(); 24763 }; 24764 24765 });" 24766 ><i class="fa fa-caret-right"></i> 24767 <span class="hover_link">Abstract</span></a 24768 ><a 24769 class="clickable no-decoration" 24770 id="vhsjs_hide_550_1707793552_4091241" 24771 onclick="$('#549_1707793552_4091156').hide(function() { 24772 if (typeof Masonry === 'function') { 24773 $('.use_masonry').masonry(); 24774 }; 24775 }); 24776 $('#vhsjs_hide_550_1707793552_4091241').hide(); 24777 $('#vhsjs_view_550_1707793552_4091241').show();" 24778 style="display: none" 24779 ><i class="fa fa-caret-down"></i> 24780 <span class="hover_link">Abstract</span></a 24781 > 24782 <div 24783 data-display-control="550_1707793552_4091241" 24784 id="549_1707793552_4091156" 24785 style="display: none" 24786 > 24787 <div class="arrow-slidedown"> 24788 <blockquote> 24789 Research and development (R&D) of biotechnology 24790 products is an iterative process typically 24791 characterized by a monolithic workflow in which 24792 a single organization takes a project from start 24793 to finish through many complex operations. This 24794 paper presents a discrete-event simulation 24795 methodology to explore an alternative 24796 disaggregated workflow in which R&D is managed 24797 by a single organization but individual 24798 operations are distributed among multiple 24799 organizations. This methodology is applied to a 24800 protein engineering R&D process to compare the 24801 monolithic and disaggregated workflows over a 24802 range of conditions and scenarios. Based upon a 24803 set of assumed parameters, results identify 24804 conditions favorable to either workflow and 24805 provide a first indication that the 24806 industry’s trend towards disaggregation 24807 may lead to improvements in development 24808 timelines. The methodology also provides a 24809 foundation for decision support tools that 24810 enable decision-makers to manage biotechnology 24811 R&D projects. 24812 </blockquote> 24813 </div> 24814 </div> 24815 </div> 24816 </div> 24817 <div class="slot-urls"></div> 24818 <a href="/wsc23papers/234.pdf" target="_blank">pdf</a 24819 ><br /> 24820 </div> 24821 <div class="slot-entry"> 24822 <a name="cea114" tabindex="-1"></a> 24823 <div class="slot-title-line">
24824 <span class="slot-title" 24825 >Development of a Discrete Event Simulation Based 24826 Framework to Evaluate Six Sigma Implementation in the 24827 Construction Sector</span 24828 > 24829 </div> 24830 <div class="slot-authors"> 24831 Srinivas Rao Jalam (Indian Institute of Technology 24832 Bombay ,Mumbai); Vaishnavi Thumuganti (Stanford 24833 University); and Albert Thomas (Indian Institute of 24834 Technology Bombay ,Mumbai) 24835 </div> 24836 <div class="slot-abstract"> 24837 <div> 24838 <a 24839 class="clickable no-decoration" 24840 id="vhsjs_view_552_1707793552_4114273" 24841 onclick="$('#vhsjs_view_552_1707793552_4114273').hide(); 24842 $('#vhsjs_hide_552_1707793552_4114273').show(); 24843 $('#551_1707793552_4114192').slideDown(function() { 24844 if (typeof Masonry === 'function') { 24845 $('.use_masonry').masonry(); 24846 }; 24847 24848 });" 24849 ><i class="fa fa-caret-right"></i> 24850 <span class="hover_link">Abstract</span></a 24851 ><a 24852 class="clickable no-decoration" 24853 id="vhsjs_hide_552_1707793552_4114273" 24854 onclick="$('#551_1707793552_4114192').hide(function() { 24855 if (typeof Masonry === 'function') { 24856 $('.use_masonry').masonry(); 24857 }; 24858 }); 24859 $('#vhsjs_hide_552_1707793552_4114273').hide(); 24860 $('#vhsjs_view_552_1707793552_4114273').show();" 24861 style="display: none" 24862 ><i class="fa fa-caret-down"></i> 24863 <span class="hover_link">Abstract</span></a 24864 > 24865 <div 24866 data-display-control="552_1707793552_4114273" 24867 id="551_1707793552_4114192" 24868 style="display: none" 24869 > 24870 <div class="arrow-slidedown"> 24871 <blockquote> 24872 Six Sigma is a useful technique adopted in the 24873 construction industry to attain supreme quality 24874 levels by reducing the variability in the 24875 processes. However, rigorous field 24876 implementation of a Six Sigma methodology takes 24877 time, money, resources, and stakeholder 24878 commitment. This study develops a 24879 simulation-based framework that can mimic a Six 24880 Sigma implementation effort in a construction 24881 site using a discrete event simulation 24882 technique. Such a framework helps the decision 24883 makers to check the benefits of Six Sigma by 24884 assessing what-if scenarios for possible system 24885 improvement, even before expending the time and 24886 resources needed for field implementation of Six 24887 Sigma techniques. Therefore, through a 24888 combination of discrete event simulation and Six 24889 Sigma, the variations in a process at a 24890 construction project are eliminated. The results 24891 of this study can inspire construction managers 24892 to use simulation to understand Six Sigma 24893 implementation and improve the process or system 24894 to fulfill customer needs. 24895 </blockquote> 24896 </div> 24897 </div> 24898 </div> 24899 </div> 24900 <div class="slot-urls"></div> 24901 <a href="/wsc23papers/cea114.pdf" target="_blank">pdf</a 24902 ><br /> 24903 </div> 24904 </div> 24905 </div> 24906 <div class="centered"> 24907 <div class="top-link"><a href="#top">Return to Top</a></div> 24908 </div> 24909 <hr /> 24910 </div> 24911 <div class="area-section"> 24912 <div class="centered"> 24913 <a name="ptrack109" tabindex="-1"></a> 24914 <div class="section-title"> 24915 Reliability Modeling and Simulation 24916 </div> 24917 </div> 24918 <div class="centered track-chair">
24919 <span class="track-chair-role" 24920 >Track Coordinator - Reliability Modeling and Simulation: </span 24921 ><span class="track-chair-names" 24922 >Sanja Lazarova-Molnar (University of Southern Denmark, 24923 Karlsruhe Institute of Technology), Xueping Li (University 24924 of Tennessee), Olufemi Omitaomu (Oak Ridge National 24925 Laboratory)</span 24926 > 24927 </div> 24928 <div class="section-entry"> 24929 <div class="session-entry"> 24930 <span class="session-event-type">Technical Session</span 24931 ><span class="type-track-spacer"> · </span 24932 ><span class="program-track" 24933 >Reliability Modeling and Simulation</span 24934 ><br /> 24935 <div class="session-title"> 24936 Simulation of Stochastic Models 24937 </div> 24938 <div class="session-chair"> 24939 Chair: Sophia Gunluk (Mila)<br /> 24940 </div> 24941 <div class="slot-entry"> 24942 <a name="con204" tabindex="-1"></a> 24943 <div class="slot-title-line"> 24944 <span class="slot-title" 24945 >Identifying Quality Mersenne Twister Streams for 24946 Parallel Stochastic Simulations</span 24947 > 24948 </div> 24949 <div class="slot-authors"> 24950 Benjamin Antunes, Claude Mazel, and David Hill (LIMOS) 24951 </div> 24952 <div class="slot-abstract"> 24953 <div> 24954 <a 24955 class="clickable no-decoration" 24956 id="vhsjs_view_554_1707793552_4189909" 24957 onclick="$('#vhsjs_view_554_1707793552_4189909').hide(); 24958 $('#vhsjs_hide_554_1707793552_4189909').show(); 24959 $('#553_1707793552_4189825').slideDown(function() { 24960 if (typeof Masonry === 'function') { 24961 $('.use_masonry').masonry(); 24962 }; 24963 24964 });" 24965 ><i class="fa fa-caret-right"></i> 24966 <span class="hover_link">Abstract</span></a 24967 ><a 24968 class="clickable no-decoration" 24969 id="vhsjs_hide_554_1707793552_4189909" 24970 onclick="$('#553_1707793552_4189825').hide(function() { 24971 if (typeof Masonry === 'function') { 24972 $('.use_masonry').masonry(); 24973 }; 24974 }); 24975 $('#vhsjs_hide_554_1707793552_4189909').hide(); 24976 $('#vhsjs_view_554_1707793552_4189909').show();" 24977 style="display: none" 24978 ><i class="fa fa-caret-down"></i> 24979 <span class="hover_link">Abstract</span></a 24980 > 24981 <div 24982 data-display-control="554_1707793552_4189909" 24983 id="553_1707793552_4189825" 24984 style="display: none" 24985 > 24986 <div class="arrow-slidedown"> 24987 <blockquote> 24988 The Mersenne Twister (MT) is a pseudo-random 24989 number generator (PRNG) widely used in High 24990 Performance Computing for parallel stochastic 24991 simulations. We aim to assess the quality of 24992 common parallelization techniques used to 24993 generate large streams of MT pseudo-random 24994 numbers. We compare three techniques: sequence 24995 splitting, random spacing and MT indexed 24996 sequence. The TestU01 Big Crush battery is used 24997 to evaluate the quality of 4096 streams for each 24998 technique on three different hardware 24999 configurations. Surprisingly, all techniques 25000 exhibited almost 30% of defects with no 25001 technique showing better quality than the 25002 others. While all 106 Big Crush tests showed 25003 failures, the failure rate was limited to a 25004 small number of tests (maximum of 6 tests failed 25005 per stream, resulting in over 94% success rate). 25006 Thanks to 33 CPU years, high-quality streams 25007 identified are given. They can be used for 25008 sensitive parallel simulations such as nuclear 25009 medicine and precise high-energy physics 25010 applications. 25011 </blockquote> 25012 </div> 25013 </div> 25014 </div> 25015 </div> 25016 <div class="slot-urls"></div> 25017 <a href="/wsc23papers/235.pdf" target="_blank">pdf</a 25018 ><br /> 25019 </div> 25020 <div class="slot-entry"> 25021 <a name="con331" tabindex="-1"></a> 25022 <div class="slot-title-line">
25023 <span class="slot-title" 25024 >Simulating Justice: Simulation of Stochastic Models 25025 for Community Bail Funds</span 25026 > 25027 </div> 25028 <div class="slot-authors"> 25029 Sophia Gunluk (Mila) and Yidan Zhang and Jamol Pender 25030 (Cornell University) 25031 </div> 25032 <div class="slot-abstract"> 25033 <div> 25034 <a 25035 class="clickable no-decoration" 25036 id="vhsjs_view_556_1707793552_42122" 25037 onclick="$('#vhsjs_view_556_1707793552_42122').hide(); 25038 $('#vhsjs_hide_556_1707793552_42122').show(); 25039 $('#555_1707793552_421212').slideDown(function() { 25040 if (typeof Masonry === 'function') { 25041 $('.use_masonry').masonry(); 25042 }; 25043 25044 });" 25045 ><i class="fa fa-caret-right"></i> 25046 <span class="hover_link">Abstract</span></a 25047 ><a 25048 class="clickable no-decoration" 25049 id="vhsjs_hide_556_1707793552_42122" 25050 onclick="$('#555_1707793552_421212').hide(function() { 25051 if (typeof Masonry === 'function') { 25052 $('.use_masonry').masonry(); 25053 }; 25054 }); 25055 $('#vhsjs_hide_556_1707793552_42122').hide(); 25056 $('#vhsjs_view_556_1707793552_42122').show();" 25057 style="display: none" 25058 ><i class="fa fa-caret-down"></i> 25059 <span class="hover_link">Abstract</span></a 25060 > 25061 <div 25062 data-display-control="556_1707793552_42122" 25063 id="555_1707793552_421212" 25064 style="display: none" 25065 > 25066 <div class="arrow-slidedown"> 25067 <blockquote> 25068 Bail funds have a long history of helping those 25069 who cannot afford bail in order to wait for 25070 trial at home. They have also had a large impact 25071 on the verdict of the defendant. In this paper, 25072 we present the first stochastic model for 25073 capturing the dynamics of a community bail fund. 25074 Our bail fund model integrates traditional 25075 queueing models with classic insurance/risk 25076 models to represent the bail fund’s 25077 intricate dynamics. We employ simulation 25078 techniques to assess Gaussian-based 25079 approximations that estimate the probability of 25080 a defendant being denied access to the bail fund 25081 when it lacks the adequate funds to support 25082 them. Additionally, we propose a new 25083 simulation-based algorithm that leverages a 25084 deterministic infusion of capital as a control 25085 variable to stabilize the probability that 25086 defendants have access to the bail fund. Our 25087 simulation results reveal that our 25088 Gaussian-based approximations are suitable for 25089 moderately and highly active bail funds. 25090 </blockquote> 25091 </div> 25092 </div> 25093 </div> 25094 </div> 25095 <div class="slot-urls"></div> 25096 <a href="/wsc23papers/236.pdf" target="_blank">pdf</a 25097 ><br /> 25098 </div> 25099 <div class="slot-entry"> 25100 <a name="con372" tabindex="-1"></a> 25101 <div class="slot-title-line">
25102 <span class="slot-title" 25103 >Sensor Fusion DEVS for Angle Estimation on Inertial 25104 Measurement Unit</span 25105 > 25106 </div> 25107 <div class="slot-authors"> 25108 Gabriel Wainer, Joseph Boi-Ukeme, and Vedant Paranjape 25109 (Carleton University) 25110 </div> 25111 <div class="slot-abstract"> 25112 <div> 25113 <a 25114 class="clickable no-decoration" 25115 id="vhsjs_view_558_1707793552_4233623" 25116 onclick="$('#vhsjs_view_558_1707793552_4233623').hide(); 25117 $('#vhsjs_hide_558_1707793552_4233623').show(); 25118 $('#557_1707793552_4233541').slideDown(function() { 25119 if (typeof Masonry === 'function') { 25120 $('.use_masonry').masonry(); 25121 }; 25122 25123 });" 25124 ><i class="fa fa-caret-right"></i> 25125 <span class="hover_link">Abstract</span></a 25126 ><a 25127 class="clickable no-decoration" 25128 id="vhsjs_hide_558_1707793552_4233623" 25129 onclick="$('#557_1707793552_4233541').hide(function() { 25130 if (typeof Masonry === 'function') { 25131 $('.use_masonry').masonry(); 25132 }; 25133 }); 25134 $('#vhsjs_hide_558_1707793552_4233623').hide(); 25135 $('#vhsjs_view_558_1707793552_4233623').show();" 25136 style="display: none" 25137 ><i class="fa fa-caret-down"></i> 25138 <span class="hover_link">Abstract</span></a 25139 > 25140 <div 25141 data-display-control="558_1707793552_4233623" 25142 id="557_1707793552_4233541" 25143 style="display: none" 25144 > 25145 <div class="arrow-slidedown"> 25146 <blockquote> 25147 We explore the application of a Sensor Fusion 25148 Framework, called SAFE (Simple, Applicable, 25149 Extensible, and Flexible) to improve the 25150 reliability of measurements obtained from 25151 Inertial Measurement Unit (IMU) sensors. SAFE is 25152 built using a DEVS specification and the Cadmium 25153 tool. Measuring angular position is a difficult 25154 task due to the unreliability of gyroscopes and 25155 accelerometers, two sensors widely used to 25156 measure angles. Although angular position can be 25157 measured using imaging systems, these are 25158 costly, and not ideal for handheld and portable 25159 devices. An alternative solution is to use 25160 sensor fusion to fuse the readings of both 25161 accelerometer and gyroscope, obtaining reliable 25162 readings. We show the application of the SAFE 25163 methodology and the results of our case study 25164 showing the potential of this method. 25165 </blockquote> 25166 </div> 25167 </div> 25168 </div> 25169 </div> 25170 <div class="slot-urls"></div> 25171 <a href="/wsc23papers/237.pdf" target="_blank">pdf</a 25172 ><br /> 25173 </div> 25174 </div> 25175 <div class="session-entry">
25176 <span class="session-event-type">Technical Session</span 25177 ><span class="type-track-spacer"> · </span 25178 ><span class="program-track" 25179 >Reliability Modeling and Simulation</span 25180 ><br /> 25181 <div class="session-title">Cyber-physical Systems</div> 25182 <div class="session-chair"> 25183 Chair: Olufemi Omitaomu (Oak Ridge National Laboratory)<br /> 25184 </div> 25185 <div class="slot-entry"> 25186 <a name="con139" tabindex="-1"></a> 25187 <div class="slot-title-line"> 25188 <span class="slot-title" 25189 >A Virtual Testbed for the Development and 25190 Verification of Cyber-Physical Systems</span 25191 > 25192 </div> 25193 <div class="slot-authors"> 25194 Jan Reitz, David Böken, and Jürgen 25195 Roßmann (RWTH Aachen University) 25196 </div> 25197 <div class="slot-abstract"> 25198 <div> 25199 <a 25200 class="clickable no-decoration" 25201 id="vhsjs_view_560_1707793552_4280877" 25202 onclick="$('#vhsjs_view_560_1707793552_4280877').hide(); 25203 $('#vhsjs_hide_560_1707793552_4280877').show(); 25204 $('#559_1707793552_4280791').slideDown(function() { 25205 if (typeof Masonry === 'function') { 25206 $('.use_masonry').masonry(); 25207 }; 25208 25209 });" 25210 ><i class="fa fa-caret-right"></i> 25211 <span class="hover_link">Abstract</span></a 25212 ><a 25213 class="clickable no-decoration" 25214 id="vhsjs_hide_560_1707793552_4280877" 25215 onclick="$('#559_1707793552_4280791').hide(function() { 25216 if (typeof Masonry === 'function') { 25217 $('.use_masonry').masonry(); 25218 }; 25219 }); 25220 $('#vhsjs_hide_560_1707793552_4280877').hide(); 25221 $('#vhsjs_view_560_1707793552_4280877').show();" 25222 style="display: none" 25223 ><i class="fa fa-caret-down"></i> 25224 <span class="hover_link">Abstract</span></a 25225 > 25226 <div 25227 data-display-control="560_1707793552_4280877" 25228 id="559_1707793552_4280791" 25229 style="display: none" 25230 > 25231 <div class="arrow-slidedown"> 25232 <blockquote> 25233 This paper presents a virtual testbed for the 25234 development and verification of cyber-physical 25235 systems, integrating network simulation, 25236 physics, and hardware emulation within the 25237 multi-domain simulation framework VEROSIM. The 25238 testbed facilitates comprehensive 25239 software-in-the-loop testing, enabling accurate 25240 and reliable evaluation of control algorithms in 25241 complex, interconnected systems. The integrated 25242 approach simplifies simulation setup and model 25243 management, while allowing natural treatment of 25244 mobility and the use of sophisticated physical 25245 radio wave propagation models. The testbed also 25246 enables the simulation of various fault 25247 scenarios, supporting the assessment of system 25248 resilience and fault-tolerant strategies. A case 25249 study involving a capsule approaching the 25250 International Space Station demonstrates the 25251 effectiveness of the proposed testbed in 25252 capturing the interactions between software, 25253 hardware, and physical elements, and verifying 25254 the overall behavior of a cyber-physical system 25255 under adverse conditions. 25256 </blockquote> 25257 </div> 25258 </div> 25259 </div> 25260 </div> 25261 <div class="slot-urls"></div> 25262 <a href="/wsc23papers/238.pdf" target="_blank">pdf</a 25263 ><br /> 25264 </div> 25265 <div class="slot-entry"> 25266 <a name="con156" tabindex="-1"></a> 25267 <div class="slot-title-line">
25268 <span class="slot-title" 25269 >Multi-Agent Simulation Based Framework for Power 25270 Restoration Time Estimation at Distribution 25271 Level</span 25272 > 25273 </div> 25274 <div class="slot-authors"> 25275 Yang Chen (North Carolina Agricultural and Technical 25276 State University), Olufemi Omitaomu (Oak Ridge National 25277 Laboratory), Nicholas Roberts (Dewberry), and Bandana 25278 Kar (U.S. Department of Energy) 25279 </div> 25280 <div class="slot-abstract"> 25281 <div> 25282 <a 25283 class="clickable no-decoration" 25284 id="vhsjs_view_562_1707793552_430452" 25285 onclick="$('#vhsjs_view_562_1707793552_430452').hide(); 25286 $('#vhsjs_hide_562_1707793552_430452').show(); 25287 $('#561_1707793552_430444').slideDown(function() { 25288 if (typeof Masonry === 'function') { 25289 $('.use_masonry').masonry(); 25290 }; 25291 25292 });" 25293 ><i class="fa fa-caret-right"></i> 25294 <span class="hover_link">Abstract</span></a 25295 ><a 25296 class="clickable no-decoration" 25297 id="vhsjs_hide_562_1707793552_430452" 25298 onclick="$('#561_1707793552_430444').hide(function() { 25299 if (typeof Masonry === 'function') { 25300 $('.use_masonry').masonry(); 25301 }; 25302 }); 25303 $('#vhsjs_hide_562_1707793552_430452').hide(); 25304 $('#vhsjs_view_562_1707793552_430452').show();" 25305 style="display: none" 25306 ><i class="fa fa-caret-down"></i> 25307 <span class="hover_link">Abstract</span></a 25308 > 25309 <div 25310 data-display-control="562_1707793552_430452" 25311 id="561_1707793552_430444" 25312 style="display: none" 25313 > 25314 <div class="arrow-slidedown"> 25315 <blockquote> 25316 The growing frequency of power outages has 25317 prompted increased interest in developing a more 25318 resilient power grid that can quickly recover 25319 from weather-related damage. At the distribution 25320 level, power restoration is a complex, 25321 multi-stage process involving multiple response 25322 entities. Providing utility stakeholders, 25323 government regulators, and the public with 25324 information about outage duration and estimated 25325 time to restoration is crucial. The research 25326 employs a multi-agent simulation approach, which 25327 allows for the simulation of decision-making 25328 behaviors among different entities and the 25329 incorporation of various uncertainties. 25330 Specifically, the study uses the open-source 25331 simulation package Mesa-Geo in conjunction with 25332 the Python language and constructs a road 25333 network using the open-source network extension 25334 pgRouting for routing queries. The research 25335 design includes several experiments focused on 25336 Florida as a case study, comparing repair crew 25337 sizes, power outage numbers, and road damage 25338 scenarios. The findings could offer valuable 25339 managerial guidance on resource allocation in 25340 the restoration process. 25341 </blockquote> 25342 </div> 25343 </div> 25344 </div> 25345 </div> 25346 <div class="slot-urls"></div> 25347 <a href="/wsc23papers/239.pdf" target="_blank">pdf</a 25348 ><br /> 25349 </div> 25350 <div class="slot-entry"> 25351 <a name="con358" tabindex="-1"></a> 25352 <div class="slot-title-line">
25353 <span class="slot-title" 25354 >A Framework for Validating Data-Driven Discrete-Event 25355 Simulation Models of Cyber-Physical Production 25356 Systems</span 25357 > 25358 </div> 25359 <div class="slot-authors"> 25360 Jonas Friederich (University of Southern Denmark) and 25361 Sanja Lazarova-Molnar (Karlsruhe Institute of 25362 Technology) 25363 </div> 25364 <div class="slot-abstract"> 25365 <div> 25366 <a 25367 class="clickable no-decoration" 25368 id="vhsjs_view_564_1707793552_432608" 25369 onclick="$('#vhsjs_view_564_1707793552_432608').hide(); 25370 $('#vhsjs_hide_564_1707793552_432608').show(); 25371 $('#563_1707793552_4326').slideDown(function() { 25372 if (typeof Masonry === 'function') { 25373 $('.use_masonry').masonry(); 25374 }; 25375 25376 });" 25377 ><i class="fa fa-caret-right"></i> 25378 <span class="hover_link">Abstract</span></a 25379 ><a 25380 class="clickable no-decoration" 25381 id="vhsjs_hide_564_1707793552_432608" 25382 onclick="$('#563_1707793552_4326').hide(function() { 25383 if (typeof Masonry === 'function') { 25384 $('.use_masonry').masonry(); 25385 }; 25386 }); 25387 $('#vhsjs_hide_564_1707793552_432608').hide(); 25388 $('#vhsjs_view_564_1707793552_432608').show();" 25389 style="display: none" 25390 ><i class="fa fa-caret-down"></i> 25391 <span class="hover_link">Abstract</span></a 25392 > 25393 <div 25394 data-display-control="564_1707793552_432608" 25395 id="563_1707793552_4326" 25396 style="display: none" 25397 > 25398 <div class="arrow-slidedown"> 25399 <blockquote> 25400 In recent years, there has been a significant 25401 increase in the deployment of Cyber-physical 25402 Production Systems (CPPS) across various 25403 industries. CPPS consist of interconnected 25404 devices and systems that combine physical and 25405 digital elements to enhance the efficiency, 25406 productivity, and reliability of manufacturing 25407 processes. Due to the continuous and fast-paced 25408 evolution of the behavior of CPPS, there is an 25409 increasing interest in generating data-driven 25410 Discrete-event Simulation (DES) models of such 25411 systems. The validation of these models, 25412 however, remains a challenge, and traditional 25413 approaches may be insufficient to ensure their 25414 accuracy. To address this challenge, we propose 25415 a framework for validating data-driven DES 25416 models of CPPS. We emphasize the importance of 25417 continuously monitoring the validity of 25418 data-driven DES models and updating them when 25419 necessary to ensure their accuracy over time. 25420 We, furthermore, demonstrate our proposed 25421 approach through a case study in reliability 25422 assessment and discuss challenges and 25423 limitations of our framework. 25424 </blockquote> 25425 </div> 25426 </div> 25427 </div> 25428 </div> 25429 <div class="slot-urls"></div> 25430 <a href="/wsc23papers/240.pdf" target="_blank">pdf</a 25431 ><br /> 25432 </div> 25433 </div> 25434 </div> 25435 <div class="centered"> 25436 <div class="top-link"><a href="#top">Return to Top</a></div> 25437 </div> 25438 <hr /> 25439 </div> 25440 <div class="area-section"> 25441 <div class="centered"> 25442 <a name="ptrack107" tabindex="-1"></a> 25443 <div class="section-title">Scientific Applications</div> 25444 </div> 25445 <div class="centered track-chair">
25446 <span class="track-chair-role" 25447 >Track Coordinator - Scientific Applications: </span 25448 ><span class="track-chair-names" 25449 >Rafael Mayo-García (CIEMAT), Esteban Mocskos 25450 (University of Buenos Aires (AR), CSC-CONICET)</span 25451 > 25452 </div> 25453 <div class="section-entry"> 25454 <div class="session-entry"> 25455 <span class="session-event-type">Technical Session</span 25456 ><span class="type-track-spacer"> · </span 25457 ><span class="program-track">Scientific Applications</span 25458 ><br /> 25459 <div class="session-title"> 25460 Computer Science for Simulations 25461 </div> 25462 <div class="session-chair"> 25463 Chair: Rafael Mayo-García (CIEMAT)<br /> 25464 </div> 25465 <div class="slot-entry"> 25466 <a name="inv187" tabindex="-1"></a> 25467 <div class="slot-title-line"> 25468 <span class="slot-title" 25469 >Strong Scaling of the SVD Algorithm for HPC Science: 25470 A PETSc-based Approach</span 25471 > 25472 </div> 25473 <div class="slot-authors"> 25474 Paula Ferrero-Roza (Universidad de La Coruña), 25475 José A. Moríñigo (CIEMAT), and Filippo 25476 Terragni (UC3M) 25477 </div> 25478 <div class="slot-abstract"> 25479 <div> 25480 <a 25481 class="clickable no-decoration" 25482 id="vhsjs_view_566_1707793552_441019" 25483 onclick="$('#vhsjs_view_566_1707793552_441019').hide(); 25484 $('#vhsjs_hide_566_1707793552_441019').show(); 25485 $('#565_1707793552_4410107').slideDown(function() { 25486 if (typeof Masonry === 'function') { 25487 $('.use_masonry').masonry(); 25488 }; 25489 25490 });" 25491 ><i class="fa fa-caret-right"></i> 25492 <span class="hover_link">Abstract</span></a 25493 ><a 25494 class="clickable no-decoration" 25495 id="vhsjs_hide_566_1707793552_441019" 25496 onclick="$('#565_1707793552_4410107').hide(function() { 25497 if (typeof Masonry === 'function') { 25498 $('.use_masonry').masonry(); 25499 }; 25500 }); 25501 $('#vhsjs_hide_566_1707793552_441019').hide(); 25502 $('#vhsjs_view_566_1707793552_441019').show();" 25503 style="display: none" 25504 ><i class="fa fa-caret-down"></i> 25505 <span class="hover_link">Abstract</span></a 25506 > 25507 <div 25508 data-display-control="566_1707793552_441019" 25509 id="565_1707793552_4410107" 25510 style="display: none" 25511 > 25512 <div class="arrow-slidedown"> 25513 <blockquote> 25514 The Singular Value Decomposition (SVD) algorithm 25515 is ubiquitous in many fields of science and 25516 technology. It may be used embedded into other 25517 advanced algorithms, solvers or data processing 25518 chains. In those scenarios dealing with large 25519 data volumes expressed as a huge matrix, there 25520 is the need of a parallel SVD version to process 25521 it efficiently. We present some ideas and 25522 results obtained within the PETSc framework,
25523 which enable to design promising HPC scalable 25524 solvers. The focused SVD implementations have 25525 been taken from the SLEPc library, which is 25526 seamless plugged into PETSc to extend its 25527 capabilities. Besides, there is also a 25528 randomized SVD and wrappers to interface 25529 ScaLAPACK and others packages to extract 25530 singular triplets. This work assesses the strong 25531 scaling attained with these SVD implementations 25532 at extracting the leading singular values of a 25533 population of both sparse and dense matrices. A 25534 comparison of performance is provided. 25535 </blockquote> 25536 </div> 25537 </div> 25538 </div> 25539 </div> 25540 <div class="slot-urls"></div> 25541 <a href="/wsc23papers/241.pdf" target="_blank">pdf</a 25542 ><br /> 25543 </div> 25544 <div class="slot-entry"> 25545 <a name="con171" tabindex="-1"></a> 25546 <div class="slot-title-line"> 25547 <span class="slot-title" 25548 >nbSimGen: Jupyter Notebook Extension for Generating 25549 Simulation Experiments</span 25550 > 25551 </div> 25552 <div class="slot-authors"> 25553 Pia Wilsdorf, Anton Willy Kirchhübel, and Adelinde 25554 M. Uhrmacher (University of Rostock) 25555 </div> 25556 <div class="slot-abstract"> 25557 <div> 25558 <a 25559 class="clickable no-decoration" 25560 id="vhsjs_view_568_1707793552_4432216" 25561 onclick="$('#vhsjs_view_568_1707793552_4432216').hide(); 25562 $('#vhsjs_hide_568_1707793552_4432216').show(); 25563 $('#567_1707793552_443213').slideDown(function() { 25564 if (typeof Masonry === 'function') { 25565 $('.use_masonry').masonry(); 25566 }; 25567 25568 });" 25569 ><i class="fa fa-caret-right"></i> 25570 <span class="hover_link">Abstract</span></a 25571 ><a 25572 class="clickable no-decoration" 25573 id="vhsjs_hide_568_1707793552_4432216" 25574 onclick="$('#567_1707793552_443213').hide(function() { 25575 if (typeof Masonry === 'function') { 25576 $('.use_masonry').masonry(); 25577 }; 25578 }); 25579 $('#vhsjs_hide_568_1707793552_4432216').hide(); 25580 $('#vhsjs_view_568_1707793552_4432216').show();" 25581 style="display: none" 25582 ><i class="fa fa-caret-down"></i> 25583 <span class="hover_link">Abstract</span></a 25584 > 25585 <div 25586 data-display-control="568_1707793552_4432216" 25587 id="567_1707793552_443213" 25588 style="display: none" 25589 > 25590 <div class="arrow-slidedown"> 25591 <blockquote> 25592 Simulation experiments are crucial in conducting 25593 simulation studies. With simulation studies 25594 growing increasingly complex, simulation 25595 experiments are intertwined with steps of 25596 conceptual modeling, model building, analyzing 25597 data, and visualizing and interpreting results. 25598 Making the products of these various steps 25599 (assumptions, requirements, data, model 25600 components, and experiments) explicit has been 25601 shown to increase the reproducibility of 25602 simulation studies. Moreover, using an 25603 integrated environment that allows developing, 25604 organizing and documenting those products can 25605 facilitate their automatic reuse and 25606 exploitation. We explore Jupyter Notebook as an 25607 all-in-one solution for conducting and 25608 documenting a simulation study, and we present 25609 nbSimGen. This Jupyter Notebook extension lends 25610 support to modelers by automatically specifying 25611 and running suitable simulation experiments. It 25612 is based on an annotation vocabulary that, 25613 during the development of the conceptual model 25614 and the simulation model, allows users to mark 25615 portions of their notebook deemed relevant to 25616 the various simulation experiments to come. 25617 </blockquote> 25618 </div> 25619 </div> 25620 </div> 25621 </div> 25622 <div class="slot-urls"></div> 25623 <a href="/wsc23papers/242.pdf" target="_blank">pdf</a 25624 ><br /> 25625 </div> 25626 <div class="slot-entry"> 25627 <a name="con288" tabindex="-1"></a> 25628 <div class="slot-title-line">
25629 <span class="slot-title" 25630 >A Facilitated Discrete Event Simulation Framework to 25631 Support Online Studies: An Intervention in a Small 25632 Enterprise</span 25633 > 25634 </div> 25635 <div class="slot-authors"> 25636 Milena Silva Oliveira, Carlos Henrique Santos, Gustavo 25637 Teodoro Gabriel, Fabiano Leal, and José Arnaldo 25638 Barra Montevechi (Federal University of Itajuba) 25639 </div> 25640 <div class="slot-abstract"> 25641 <div> 25642 <a 25643 class="clickable no-decoration" 25644 id="vhsjs_view_570_1707793552_445604" 25645 onclick="$('#vhsjs_view_570_1707793552_445604').hide(); 25646 $('#vhsjs_hide_570_1707793552_445604').show(); 25647 $('#569_1707793552_4455957').slideDown(function() { 25648 if (typeof Masonry === 'function') { 25649 $('.use_masonry').masonry(); 25650 }; 25651 25652 });" 25653 ><i class="fa fa-caret-right"></i> 25654 <span class="hover_link">Abstract</span></a 25655 ><a 25656 class="clickable no-decoration" 25657 id="vhsjs_hide_570_1707793552_445604" 25658 onclick="$('#569_1707793552_4455957').hide(function() { 25659 if (typeof Masonry === 'function') { 25660 $('.use_masonry').masonry(); 25661 }; 25662 }); 25663 $('#vhsjs_hide_570_1707793552_445604').hide(); 25664 $('#vhsjs_view_570_1707793552_445604').show();" 25665 style="display: none" 25666 ><i class="fa fa-caret-down"></i> 25667 <span class="hover_link">Abstract</span></a 25668 > 25669 <div 25670 data-display-control="570_1707793552_445604" 25671 id="569_1707793552_4455957" 25672 style="display: none" 25673 > 25674 <div class="arrow-slidedown"> 25675 <blockquote> 25676 Considering some challenges that prevent the 25677 expansion of discrete event simulation studies, 25678 such as financial constraints to invest in the 25679 data collection of large samples and to hire 25680 qualified people for data analysis and for 25681 developing complex models, this paper aims to 25682 propose a framework to support simulation 25683 studies where it is not widely used, adopting 25684 facilitated modeling. Since the facilitated DES 25685 frameworks in the literature focus on healthcare 25686 and face-to-face meetings, the present work 25687 offers a framework for simulation projects in 25688 production systems, which also supports online 25689 interventions. After its development, the 25690 FaMoSim (Facilitated Modeling Simulation) 25691 framework was applied in a real case to evaluate 25692 its applicability. In the application, it was 25693 possible to carry out a faster and more flexible 25694 online modeling process, create a simple 25695 computer model that does not require a complex 25696 data collection structure nor a specialist team, 25697 and assist the stakeholders in identifying 25698 improvements. 25699 </blockquote> 25700 </div> 25701 </div> 25702 </div> 25703 </div> 25704 <div class="slot-urls"></div> 25705 <a href="/wsc23papers/243.pdf" target="_blank">pdf</a 25706 ><br /> 25707 </div> 25708 </div> 25709 <div class="session-entry">
25710 <span class="session-event-type">Technical Session</span 25711 ><span class="type-track-spacer"> · </span 25712 ><span class="program-track">Scientific Applications</span 25713 ><br /> 25714 <div class="session-title">AI-oriented Simulations</div> 25715 <div class="session-chair"> 25716 Chair: Rafael Mayo-García (CIEMAT)<br /> 25717 </div> 25718 <div class="slot-entry"> 25719 <a name="con376" tabindex="-1"></a> 25720 <div class="slot-title-line"> 25721 <span class="slot-title" 25722 >Emotion Classification Through Speech Data 25723 Analysis</span 25724 > 25725 </div> 25726 <div class="slot-authors"> 25727 Luzalen Marcos, Abdolreza Abhari, and Kristiina Mai 25728 (Toronto Metropolitan University) 25729 </div> 25730 <div class="slot-abstract"> 25731 <div> 25732 <a 25733 class="clickable no-decoration" 25734 id="vhsjs_view_572_1707793552_4516509" 25735 onclick="$('#vhsjs_view_572_1707793552_4516509').hide(); 25736 $('#vhsjs_hide_572_1707793552_4516509').show(); 25737 $('#571_1707793552_4516423').slideDown(function() { 25738 if (typeof Masonry === 'function') { 25739 $('.use_masonry').masonry(); 25740 }; 25741 25742 });" 25743 ><i class="fa fa-caret-right"></i> 25744 <span class="hover_link">Abstract</span></a 25745 ><a 25746 class="clickable no-decoration" 25747 id="vhsjs_hide_572_1707793552_4516509" 25748 onclick="$('#571_1707793552_4516423').hide(function() { 25749 if (typeof Masonry === 'function') { 25750 $('.use_masonry').masonry(); 25751 }; 25752 }); 25753 $('#vhsjs_hide_572_1707793552_4516509').hide(); 25754 $('#vhsjs_view_572_1707793552_4516509').show();" 25755 style="display: none" 25756 ><i class="fa fa-caret-down"></i> 25757 <span class="hover_link">Abstract</span></a 25758 > 25759 <div 25760 data-display-control="572_1707793552_4516509" 25761 id="571_1707793552_4516423" 25762 style="display: none" 25763 > 25764 <div class="arrow-slidedown"> 25765 <blockquote> 25766 Good quality healthcare services require 25767 effective communication between the patient and 25768 the healthcare provider. This work will help 25769 improve the areas of healthcare systems 25770 automation and optimization by applying Speech 25771 Emotion Recognition (SER) in health 25772 consultations to prevent miscommunication 25773 between patients and healthcare providers. 25774 Crowd-Sourced Emotional Multimodal Actors 25775 Dataset (CREMA-D) was used to compare the 25776 performances of different machine learning 25777 models in classifying emotions. Before feeding 25778 the raw dataset to the models, exploratory data 25779 analysis was done to determine features that 25780 should be considered for future analysis. Our 25781 results showed that depending on the emotion, 25782 there are some syllables in the text that were 25783 emphasized or took time to be pronounced by the 25784 speaker. After data analysis, the dataset was 25785 fed into different models and determined that
25786 the Support Vector Machine (SVM) is a 25787 machine-learning model for SER. 25788 </blockquote> 25789 </div> 25790 </div> 25791 </div> 25792 </div> 25793 <div class="slot-urls"></div> 25794 <a href="/wsc23papers/244.pdf" target="_blank">pdf</a 25795 ><br /> 25796 </div> 25797 <div class="slot-entry"> 25798 <a name="inv189" tabindex="-1"></a> 25799 <div class="slot-title-line"> 25800 <span class="slot-title" 25801 >GPT-Based Models Meet Simulation: How to Efficiently 25802 Use Large-Scale Pre-Trained Language Models Across 25803 Simulation Tasks</span 25804 > 25805 </div> 25806 <div class="slot-authors"> 25807 Philippe J. Giabbanelli (Miami University) 25808 </div> 25809 <div class="slot-abstract"> 25810 <div> 25811 <a 25812 class="clickable no-decoration" 25813 id="vhsjs_view_574_1707793552_453865" 25814 onclick="$('#vhsjs_view_574_1707793552_453865').hide(); 25815 $('#vhsjs_hide_574_1707793552_453865').show(); 25816 $('#573_1707793552_4538565').slideDown(function() { 25817 if (typeof Masonry === 'function') { 25818 $('.use_masonry').masonry(); 25819 }; 25820 25821 });" 25822 ><i class="fa fa-caret-right"></i> 25823 <span class="hover_link">Abstract</span></a 25824 ><a 25825 class="clickable no-decoration" 25826 id="vhsjs_hide_574_1707793552_453865" 25827 onclick="$('#573_1707793552_4538565').hide(function() { 25828 if (typeof Masonry === 'function') { 25829 $('.use_masonry').masonry(); 25830 }; 25831 }); 25832 $('#vhsjs_hide_574_1707793552_453865').hide(); 25833 $('#vhsjs_view_574_1707793552_453865').show();" 25834 style="display: none" 25835 ><i class="fa fa-caret-down"></i> 25836 <span class="hover_link">Abstract</span></a 25837 > 25838 <div 25839 data-display-control="574_1707793552_453865" 25840 id="573_1707793552_4538565" 25841 style="display: none" 25842 > 25843 <div class="arrow-slidedown"> 25844 <blockquote> 25845 The disruptive technology provided by 25846 large-scale pre-trained language models (LLMs) 25847 such as ChatGPT or GPT-4 has received 25848 significant attention in several application 25849 domains, often with an emphasis on high-level 25850 opportunities and concerns. This paper is the 25851 first examination regarding the use of LLMs for 25852 scientific simulations. We focus on four 25853 modeling and simulation tasks, each time 25854 assessing the expected benefits and limitations 25855 of LLMs while providing practical guidance for 25856 modelers regarding the steps involved. The first 25857 task is devoted to explaining the structure of a 25858 conceptual model to promote the engagement of 25859 participants in the modeling process. The second 25860 task focuses on summarizing simulation outputs, 25861 so that model users can identify a preferred 25862 scenario. The third task seeks to broaden 25863 accessibility to simulation platforms by 25864 conveying the insights of simulation 25865 visualizations via text. Finally, the last task 25866 evokes the possibility of explaining simulation 25867 errors and providing guidance to resolve them. 25868 </blockquote> 25869 </div> 25870 </div> 25871 </div> 25872 </div> 25873 <div class="slot-urls"></div> 25874 <a href="/wsc23papers/245.pdf" target="_blank">pdf</a 25875 ><br /> 25876 </div> 25877 </div> 25878 <div class="session-entry">
25879 <span class="session-event-type">Technical Session</span 25880 ><span class="type-track-spacer"> · </span 25881 ><span class="program-track">Scientific Applications</span 25882 ><br /> 25883 <div class="session-title">Multi-physics Simulations</div> 25884 <div class="session-chair"> 25885 Chair: Rafael Mayo-García (CIEMAT)<br /> 25886 </div> 25887 <div class="slot-entry"> 25888 <a name="cea127" tabindex="-1"></a> 25889 <div class="slot-title-line"> 25890 <span class="slot-title" 25891 >An Integrated Multi-Physics Optimization Framework 25892 for Particle Accelerator Design</span 25893 > 25894 </div> 25895 <div class="slot-authors"> 25896 Gongxiaohui Chen, Tyler Chang, and John Power (Argonne 25897 National Laboratory) and Chungunag Jing (Euclid Techlabs 25898 LLC) 25899 </div> 25900 <div class="slot-abstract"> 25901 <div> 25902 <a 25903 class="clickable no-decoration" 25904 id="vhsjs_view_576_1707793552_4579587" 25905 onclick="$('#vhsjs_view_576_1707793552_4579587').hide(); 25906 $('#vhsjs_hide_576_1707793552_4579587').show(); 25907 $('#575_1707793552_4579504').slideDown(function() { 25908 if (typeof Masonry === 'function') { 25909 $('.use_masonry').masonry(); 25910 }; 25911 25912 });" 25913 ><i class="fa fa-caret-right"></i> 25914 <span class="hover_link">Abstract</span></a 25915 ><a 25916 class="clickable no-decoration" 25917 id="vhsjs_hide_576_1707793552_4579587" 25918 onclick="$('#575_1707793552_4579504').hide(function() { 25919 if (typeof Masonry === 'function') { 25920 $('.use_masonry').masonry(); 25921 }; 25922 }); 25923 $('#vhsjs_hide_576_1707793552_4579587').hide(); 25924 $('#vhsjs_view_576_1707793552_4579587').show();" 25925 style="display: none" 25926 ><i class="fa fa-caret-down"></i> 25927 <span class="hover_link">Abstract</span></a 25928 > 25929 <div 25930 data-display-control="576_1707793552_4579587" 25931 id="575_1707793552_4579504" 25932 style="display: none" 25933 > 25934 <div class="arrow-slidedown"> 25935 <blockquote> 25936 The overarching goal of beamline design is to 25937 achieve a high brightness electron beam from the 25938 beamline. Traditional beamline design studies 25939 involved separate optimizations of 25940 radio-frequency cavities, magnets, and beam 25941 dynamics using different codes and pursuing 25942 various intermediate objectives. In this work, 25943 we present a novel unified global optimization 25944 framework that integrates multiple physics 25945 modules for beamline design as simulation 25946 functions for a two-stage global optimization 25947 solver. 25948 </blockquote> 25949 </div> 25950 </div> 25951 </div> 25952 </div> 25953 <div class="slot-urls"></div> 25954 <a href="/wsc23papers/cea127.pdf" target="_blank">pdf</a 25955 ><br /> 25956 </div> 25957 <div class="slot-entry"> 25958 <a name="inv156" tabindex="-1"></a> 25959 <div class="slot-title-line">
25960 <span class="slot-title" 25961 >The Cloud-Based Implementation and Standardisation of 25962 Anthropomorphic Phantoms and their Applications</span 25963 > 25964 </div> 25965 <div class="slot-authors"> 25966 Osiris Núñez-Chongo and Manuel Carretero 25967 (Universidad Carlos III de Madrid); Rafael 25968 Mayo-García (Centro de Investigaciones 25969 Energéticas, Medioambientales y Tecnológicas 25970 (CIEMAT)); and Hernán Asorey (Comisión 25971 Nacional de Energía Atómica, Centro 25972 Atómico Bariloche) 25973 </div> 25974 <div class="slot-abstract"> 25975 <div> 25976 <a 25977 class="clickable no-decoration" 25978 id="vhsjs_view_578_1707793552_461231" 25979 onclick="$('#vhsjs_view_578_1707793552_461231').hide(); 25980 $('#vhsjs_hide_578_1707793552_461231').show(); 25981 $('#577_1707793552_4612222').slideDown(function() { 25982 if (typeof Masonry === 'function') { 25983 $('.use_masonry').masonry(); 25984 }; 25985 25986 });" 25987 ><i class="fa fa-caret-right"></i> 25988 <span class="hover_link">Abstract</span></a 25989 ><a 25990 class="clickable no-decoration" 25991 id="vhsjs_hide_578_1707793552_461231" 25992 onclick="$('#577_1707793552_4612222').hide(function() { 25993 if (typeof Masonry === 'function') { 25994 $('.use_masonry').masonry(); 25995 }; 25996 }); 25997 $('#vhsjs_hide_578_1707793552_461231').hide(); 25998 $('#vhsjs_view_578_1707793552_461231').show();" 25999 style="display: none" 26000 ><i class="fa fa-caret-down"></i> 26001 <span class="hover_link">Abstract</span></a 26002 > 26003 <div 26004 data-display-control="578_1707793552_461231" 26005 id="577_1707793552_4612222" 26006 style="display: none" 26007 > 26008 <div class="arrow-slidedown"> 26009 <blockquote> 26010 Radiation protection applications often require 26011 the creation of a large number of precise 26012 simulations of radiation-human body 26013 interactions. Our research is focused on 26014 creating RadPhantom, a new Geant4 application 26015 that constructs voxelized anthropomorphic 26016 phantom models. This allows for the standardized 26017 and reproducible generation of Geant4 26018 simulations in cloud-based environments. We have 26019 incorporated existing and publicly accessible 26020 models into Meiga, a framework designed for the 26021 integration of Geant4-based applications. To 26022 standardize these simulations, guarantee their 26023 reproducibility, and adhere to the FAIR 26024 principles, we have developed an extended 26025 vocabulary schema using metadata and ontologies 26026 that align with current standards. By employing 26027 virtualization containers, we capitalize on the 26028 scalability and adaptability of public and 26029 federated clouds. In this paper, we detail our 26030 implementation, present some benchmarking 26031 results and comparisons with current 26032 methodologies, and discuss the potential 26033 applications for evaluating doses on commercial 26034 flights or assessing radiation shielding in 26035 neutron production facilities. 26036 </blockquote> 26037 </div> 26038 </div> 26039 </div> 26040 </div> 26041 <div class="slot-urls"></div> 26042 <a href="/wsc23papers/246.pdf" target="_blank">pdf</a 26043 ><br /> 26044 </div> 26045 </div> 26046 </div> 26047 <div class="centered"> 26048 <div class="top-link"><a href="#top">Return to Top</a></div> 26049 </div> 26050 <hr /> 26051 </div> 26052 <div class="area-section"> 26053 <div class="centered"> 26054 <a name="ptrack130" tabindex="-1"></a> 26055 <div class="section-title">Simulation Around the World</div> 26056 </div> 26057 <div class="centered track-chair">
26058 <span class="track-chair-role" 26059 >Track Coordinator - Simulation Around the World: </span 26060 ><span class="track-chair-names" 26061 >Seong-Hee Kim (Georgia Institute of Technology), Theresa 26062 Roeder (San Francisco State University), John Shortle 26063 (George Mason University)</span 26064 > 26065 </div> 26066 <div class="section-entry"> 26067 <div class="session-entry"> 26068 <span class="session-event-type">Technical Session</span 26069 ><span class="type-track-spacer"> · </span 26070 ><span class="program-track" 26071 >Simulation Around the World</span 26072 ><br /> 26073 <div class="session-title"> 26074 Construction and Project Management 26075 </div> 26076 <div class="session-chair"> 26077 Chair: Gabriel Wainer (Carleton University)<br /> 26078 </div> 26079 <div class="slot-entry"> 26080 <a name="satwcont106" tabindex="-1"></a> 26081 <div class="slot-title-line"> 26082 <span class="slot-title" 26083 >DEVS Modeling and Simulation of the Loading and 26084 Hauling Process in Open Pit Mines</span 26085 > 26086 </div> 26087 <div class="slot-authors"> 26088 Joel Santana and Alonso Inostrosa-Psijas (Universidad de 26089 Valparaíso), Francisco Moreno (Universidad de 26090 Santiago), Mauricio Oyarzún (Universidad Arturo 26091 Prat), and Gabriel Wainer (Carleton University) 26092 </div> 26093 <div class="slot-abstract"> 26094 <div> 26095 <a 26096 class="clickable no-decoration" 26097 id="vhsjs_view_580_1707793552_4718595" 26098 onclick="$('#vhsjs_view_580_1707793552_4718595').hide(); 26099 $('#vhsjs_hide_580_1707793552_4718595').show(); 26100 $('#579_1707793552_471851').slideDown(function() { 26101 if (typeof Masonry === 'function') { 26102 $('.use_masonry').masonry(); 26103 }; 26104 26105 });" 26106 ><i class="fa fa-caret-right"></i> 26107 <span class="hover_link">Abstract</span></a 26108 ><a 26109 class="clickable no-decoration" 26110 id="vhsjs_hide_580_1707793552_4718595" 26111 onclick="$('#579_1707793552_471851').hide(function() { 26112 if (typeof Masonry === 'function') { 26113 $('.use_masonry').masonry(); 26114 }; 26115 }); 26116 $('#vhsjs_hide_580_1707793552_4718595').hide(); 26117 $('#vhsjs_view_580_1707793552_4718595').show();" 26118 style="display: none" 26119 ><i class="fa fa-caret-down"></i> 26120 <span class="hover_link">Abstract</span></a 26121 > 26122 <div 26123 data-display-control="580_1707793552_4718595" 26124 id="579_1707793552_471851" 26125 style="display: none" 26126 > 26127 <div class="arrow-slidedown"> 26128 <blockquote> 26129 Chile is the world's leading copper producer, 26130 with more than 5.6 million tons produced in 26131 2020. Most of the produced ore comes from open 26132 pit mines, whose extraction process consists of 26133 different subprocesses, with ore hauling 26134 incurring the highest operational cost. Tools to 26135 improve this subprocess are of paramount 26136 importance. Most tools use approaches that rely 26137 on optimization based on analytical methods. 26138 However, these fail to capture human behavior or 26139 to consider fine-grained details. To this end, 26140 we present a DEVS (Discrete-Event System 26141 Specification) simulation model. The formal 26142 definition of DEVS helps with the design and 26143 experimentation. DEVS modular interfaces allow 26144 users to extend the model easily to consider 26145 more entities, mine layouts, and dispatching 26146 policies. Simulations of the model delivered 26147 precise results compared to the literature, 26148 providing a valuable tool for decision-making in 26149 the mining industry. 26150 </blockquote> 26151 </div> 26152 </div> 26153 </div> 26154 </div> 26155 <div class="slot-urls"></div> 26156 <a href="/wsc23papers/259.pdf" target="_blank">pdf</a 26157 ><br /> 26158 </div> 26159 <div class="slot-entry"> 26160 <a name="con338" tabindex="-1"></a> 26161 <div class="slot-title-line">
26162 <span class="slot-title" 26163 >A Hybrid Simulation-based Optimization Framework for 26164 Managing Modular Bridge Construction Projects: A 26165 Cable-Stayed Bridge Case Study</span 26166 > 26167 </div> 26168 <div class="slot-authors"> 26169 Mohamed Assaf, Sena Assaf, William Correa, Rafik 26170 Lemouchi, and Yasser Mohamed (University of Alberta) 26171 </div> 26172 <div class="slot-abstract"> 26173 <div> 26174 <a 26175 class="clickable no-decoration" 26176 id="vhsjs_view_582_1707793552_4936278" 26177 onclick="$('#vhsjs_view_582_1707793552_4936278').hide(); 26178 $('#vhsjs_hide_582_1707793552_4936278').show(); 26179 $('#581_1707793552_4936197').slideDown(function() { 26180 if (typeof Masonry === 'function') { 26181 $('.use_masonry').masonry(); 26182 }; 26183 26184 });" 26185 ><i class="fa fa-caret-right"></i> 26186 <span class="hover_link">Abstract</span></a 26187 ><a 26188 class="clickable no-decoration" 26189 id="vhsjs_hide_582_1707793552_4936278" 26190 onclick="$('#581_1707793552_4936197').hide(function() { 26191 if (typeof Masonry === 'function') { 26192 $('.use_masonry').masonry(); 26193 }; 26194 }); 26195 $('#vhsjs_hide_582_1707793552_4936278').hide(); 26196 $('#vhsjs_view_582_1707793552_4936278').show();" 26197 style="display: none" 26198 ><i class="fa fa-caret-down"></i> 26199 <span class="hover_link">Abstract</span></a 26200 > 26201 <div 26202 data-display-control="582_1707793552_4936278" 26203 id="581_1707793552_4936197" 26204 style="display: none" 26205 > 26206 <div class="arrow-slidedown"> 26207 <blockquote> 26208 Generally, bridge construction is one of the 26209 most complex structures in the construction 26210 industry due to the higher scalability and 26211 supply chain complexity. The modular bridge 26212 construction (MBC) technique is considered more 26213 advantageous in providing higher productivity, 26214 shorter schedules, and better quality. Current 26215 practices in managing MBC projects overlook 26216 dynamic behaviors among the relevant 26217 stakeholders and the interactions among various 26218 interacting systems, including manufacturing, 26219 logistics, and onsite assembly. To this end, 26220 this paper proposes a simulation-optimization 26221 framework to enhance MBC projects planning. The 26222 simulation module comprises discrete event 26223 simulation and agent-based modeling to model the 26224 interconnected behaviors of the MBC systems. The 26225 optimization module aims to improve the key 26226 performance indicators (KPIs) of MBC projects, 26227 including project cost, schedule, and 26228 sustainability. The proposed framework is 26229 validated by introducing an MBC case of a 26230 cable-stayed bridge. The generated solutions by 26231 the optimization model show possible significant 26232 enhancements in the identified KPIs. 26233 </blockquote> 26234 </div> 26235 </div> 26236 </div> 26237 </div> 26238 <div class="slot-urls"></div> 26239 <a href="/wsc23papers/260.pdf" target="_blank">pdf</a 26240 ><br /> 26241 </div> 26242 <div class="slot-entry"> 26243 <a name="con330" tabindex="-1"></a> 26244 <div class="slot-title-line">
26245 <span class="slot-title" 26246 >Integrated Analysis and Simulation for Enhancing Wall 26247 Assembly Process Efficiency by Resolving 26248 Bottlenecks</span 26249 > 26250 </div> 26251 <div class="slot-authors"> 26252 Zeyu Mao, Alejandro Ramon Rivera, and Yasser Mohamed 26253 (University of Alberta) 26254 </div> 26255 <div class="slot-abstract"> 26256 <div> 26257 <a 26258 class="clickable no-decoration" 26259 id="vhsjs_view_584_1707793552_4960127" 26260 onclick="$('#vhsjs_view_584_1707793552_4960127').hide(); 26261 $('#vhsjs_hide_584_1707793552_4960127').show(); 26262 $('#583_1707793552_496005').slideDown(function() { 26263 if (typeof Masonry === 'function') { 26264 $('.use_masonry').masonry(); 26265 }; 26266 26267 });" 26268 ><i class="fa fa-caret-right"></i> 26269 <span class="hover_link">Abstract</span></a 26270 ><a 26271 class="clickable no-decoration" 26272 id="vhsjs_hide_584_1707793552_4960127" 26273 onclick="$('#583_1707793552_496005').hide(function() { 26274 if (typeof Masonry === 'function') { 26275 $('.use_masonry').masonry(); 26276 }; 26277 }); 26278 $('#vhsjs_hide_584_1707793552_4960127').hide(); 26279 $('#vhsjs_view_584_1707793552_4960127').show();" 26280 style="display: none" 26281 ><i class="fa fa-caret-down"></i> 26282 <span class="hover_link">Abstract</span></a 26283 > 26284 <div 26285 data-display-control="584_1707793552_4960127" 26286 id="583_1707793552_496005" 26287 style="display: none" 26288 > 26289 <div class="arrow-slidedown"> 26290 <blockquote> 26291 Unbalanced production rates of activities and 26292 abundant resource allocation are the leading 26293 reason behind bottlenecks in processes and have 26294 been one of the causes that negatively affect 26295 projects leading to wasted resources. Many 26296 industries suffer from unbalanced resource 26297 workloads, where manufacturing takt times at 26298 some workstations are out of sync with preceding 26299 stations, consequently leading to an abruption 26300 in the workflow between activities. This 26301 research aims to assess the current state of the 26302 manufacturing process of a wall assembly line 26303 from material cutting to installation, 26304 identifying bottlenecks, and creating a 26305 framework that would contrast both cycles to 26306 finally propose a solution through simulation. A 26307 case was studied to propose innovative methods 26308 to improve the process flow and to eliminate any 26309 waste generated by bottlenecks. This will not 26310 only reduce the process duration but will also 26311 significantly increase cost expenditure since 26312 the amount of idle time and resources will be 26313 reduced. 26314 </blockquote> 26315 </div> 26316 </div> 26317 </div> 26318 </div> 26319 <div class="slot-urls"></div> 26320 <a href="/wsc23papers/261.pdf" target="_blank">pdf</a 26321 ><br /> 26322 </div> 26323 </div> 26324 <div class="session-entry">
26325 <span class="session-event-type">Technical Session</span 26326 ><span class="type-track-spacer"> · </span 26327 ><span class="program-track" 26328 >Simulation Around the World</span 26329 ><br /> 26330 <div class="session-title"> 26331 Facilitating Business Decisions 26332 </div> 26333 <div class="session-chair"> 26334 Chair: Christos Alexopoulos (Georgia Institute of 26335 Technology)<br /> 26336 </div> 26337 <div class="slot-entry"> 26338 <a name="satwcea103" tabindex="-1"></a> 26339 <div class="slot-title-line"> 26340 <span class="slot-title" 26341 >Impactful Simulation Models from a Brazilian 26342 Simulation Consultancy</span 26343 > 26344 </div> 26345 <div class="slot-authors"> 26346 Wilson Pereira and Leonardo Chwif (Simulate) 26347 </div> 26348 <div class="slot-abstract"> 26349 <div> 26350 <a 26351 class="clickable no-decoration" 26352 id="vhsjs_view_586_1707793552_5044959" 26353 onclick="$('#vhsjs_view_586_1707793552_5044959').hide(); 26354 $('#vhsjs_hide_586_1707793552_5044959').show(); 26355 $('#585_1707793552_5044878').slideDown(function() { 26356 if (typeof Masonry === 'function') { 26357 $('.use_masonry').masonry(); 26358 }; 26359 26360 });" 26361 ><i class="fa fa-caret-right"></i> 26362 <span class="hover_link">Abstract</span></a 26363 ><a 26364 class="clickable no-decoration" 26365 id="vhsjs_hide_586_1707793552_5044959" 26366 onclick="$('#585_1707793552_5044878').hide(function() { 26367 if (typeof Masonry === 'function') { 26368 $('.use_masonry').masonry(); 26369 }; 26370 }); 26371 $('#vhsjs_hide_586_1707793552_5044959').hide(); 26372 $('#vhsjs_view_586_1707793552_5044959').show();" 26373 style="display: none" 26374 ><i class="fa fa-caret-down"></i> 26375 <span class="hover_link">Abstract</span></a 26376 > 26377 <div 26378 data-display-control="586_1707793552_5044959" 26379 id="585_1707793552_5044878" 26380 style="display: none" 26381 > 26382 <div class="arrow-slidedown"> 26383 <blockquote> 26384 Simulate Simulation Technology is a Brazilian 26385 consultancy company focused on developing 26386 discrete event simulation models and providing 26387 simulation training. Some of the simulation 26388 models developed over the last 20 years are 26389 classified by us as successful and impactful, 26390 with no relationship to their complexity, 26391 applicability level, or purpose. This article 26392 presents some of these models. 26393 </blockquote> 26394 </div> 26395 </div> 26396 </div> 26397 </div> 26398 <div class="slot-urls"></div> 26399 <a href="/wsc23papers/satwcea103.pdf" target="_blank" 26400 >pdf</a 26401 ><br /> 26402 </div> 26403 <div class="slot-entry"> 26404 <a name="satwcea101" tabindex="-1"></a> 26405 <div class="slot-title-line">
26406 <span class="slot-title" 26407 >Using System Dynamics to Adapt Business Models to 26408 Changing Conditions</span 26409 > 26410 </div> 26411 <div class="slot-authors"> 26412 Marisa Analia Sanchez (Universidad Nacional del Sur) and 26413 Javier García Fronti (Universidad de Buenos Aires) 26414 </div> 26415 <div class="slot-abstract"> 26416 <div> 26417 <a 26418 class="clickable no-decoration" 26419 id="vhsjs_view_588_1707793552_5229924" 26420 onclick="$('#vhsjs_view_588_1707793552_5229924').hide(); 26421 $('#vhsjs_hide_588_1707793552_5229924').show(); 26422 $('#587_1707793552_5229845').slideDown(function() { 26423 if (typeof Masonry === 'function') { 26424 $('.use_masonry').masonry(); 26425 }; 26426 26427 });" 26428 ><i class="fa fa-caret-right"></i> 26429 <span class="hover_link">Abstract</span></a 26430 ><a 26431 class="clickable no-decoration" 26432 id="vhsjs_hide_588_1707793552_5229924" 26433 onclick="$('#587_1707793552_5229845').hide(function() { 26434 if (typeof Masonry === 'function') { 26435 $('.use_masonry').masonry(); 26436 }; 26437 }); 26438 $('#vhsjs_hide_588_1707793552_5229924').hide(); 26439 $('#vhsjs_view_588_1707793552_5229924').show();" 26440 style="display: none" 26441 ><i class="fa fa-caret-down"></i> 26442 <span class="hover_link">Abstract</span></a 26443 > 26444 <div 26445 data-display-control="588_1707793552_5229924" 26446 id="587_1707793552_5229845" 26447 style="display: none" 26448 > 26449 <div class="arrow-slidedown"> 26450 <blockquote> 26451 This paper addresses the problem of determining 26452 organizational adaptations to ensure business 26453 continuity. We propose a methodology to assess 26454 the impact of disruptions on a business model 26455 and evaluate interventions using System Dynamics 26456 archetypes. The methodology aims to contribute 26457 to making decision-making more effective and 26458 efficient in an uncertain scenario. 26459 </blockquote> 26460 </div> 26461 </div> 26462 </div> 26463 </div> 26464 <div class="slot-urls"></div> 26465 <a href="/wsc23papers/satwcea101.pdf" target="_blank" 26466 >pdf</a 26467 ><br /> 26468 </div> 26469 <div class="slot-entry"> 26470 <a name="satwcea109" tabindex="-1"></a> 26471 <div class="slot-title-line"> 26472 <span class="slot-title" 26473 >Simulation-Based Immersive Analytics Toward Advanced 26474 Decision Making</span 26475 > 26476 </div> 26477 <div class="slot-authors"> 26478 Gisela Belen Confalonieri, Ezequiel Pecker-Marcosig, 26479 Esteban Lanzarotti, and Rodrigo Daniel Castro 26480 (Departamento de Computación, FCEyN-UBA / Instituto 26481 de Ciencias de la Computación (ICC-CONICET)) 26482 </div> 26483 <div class="slot-abstract"> 26484 <div> 26485 <a 26486 class="clickable no-decoration" 26487 id="vhsjs_view_590_1707793552_5251224" 26488 onclick="$('#vhsjs_view_590_1707793552_5251224').hide(); 26489 $('#vhsjs_hide_590_1707793552_5251224').show(); 26490 $('#589_1707793552_5251145').slideDown(function() { 26491 if (typeof Masonry === 'function') { 26492 $('.use_masonry').masonry(); 26493 }; 26494 26495 });" 26496 ><i class="fa fa-caret-right"></i>
26497 <span class="hover_link">Abstract</span></a 26498 ><a 26499 class="clickable no-decoration" 26500 id="vhsjs_hide_590_1707793552_5251224" 26501 onclick="$('#589_1707793552_5251145').hide(function() { 26502 if (typeof Masonry === 'function') { 26503 $('.use_masonry').masonry(); 26504 }; 26505 }); 26506 $('#vhsjs_hide_590_1707793552_5251224').hide(); 26507 $('#vhsjs_view_590_1707793552_5251224').show();" 26508 style="display: none" 26509 ><i class="fa fa-caret-down"></i> 26510 <span class="hover_link">Abstract</span></a 26511 > 26512 <div 26513 data-display-control="590_1707793552_5251224" 26514 id="589_1707793552_5251145" 26515 style="display: none" 26516 > 26517 <div class="arrow-slidedown"> 26518 <blockquote> 26519 Managing effective visualisations for data 26520 analysis is critical to support informed 26521 decision making across multiple domains, which 26522 also requires the ability to interact with the 26523 data. This includes understanding data from 26524 real-world scenarios enriched with simulated 26525 virtual data, and the ability to assess the 26526 impact of user interventions on concurrently 26527 running simulation models. To address this, we 26528 propose a framework that combines a DEVS 26529 simulator with a game engine, allowing users to 26530 interact directly with the model during 26531 simulation runtime, while observing realistic 26532 visualisations of the generated data and system 26533 behaviour. 26534 </blockquote> 26535 </div> 26536 </div> 26537 </div> 26538 </div> 26539 <div class="slot-urls"></div> 26540 <a href="/wsc23papers/satwcea109.pdf" target="_blank" 26541 >pdf</a 26542 ><br /> 26543 </div> 26544 </div> 26545 <div class="session-entry"> 26546 <span class="session-event-type">Technical Session</span 26547 ><span class="type-track-spacer"> · </span 26548 ><span class="program-track" 26549 >Simulation Around the World</span 26550 ><br /> 26551 <div class="session-title"> 26552 Discrete-event Simulation Language and Platforms 26553 </div> 26554 <div class="session-chair"> 26555 Chair: María Julia Blas (INGAR CONICET UTN)<br /> 26556 </div> 26557 <div class="slot-entry"> 26558 <a name="satwcont104" tabindex="-1"></a> 26559 <div class="slot-title-line"> 26560 <span class="slot-title" 26561 >RustSim: A Process-Oriented Simulation Framework for 26562 the Rust Language</span 26563 > 26564 </div> 26565 <div class="slot-authors"> 26566 Kevin Frez and Mauricio Oyarzun (Universidad Arturo 26567 Prat), Alonso Inostrosa-Psijas (Universidad de 26568 Valparaíso), Francisco Moreno (Universidad de 26569 Santiago), and Gabriel Wainer (Carleton University) 26570 </div> 26571 <div class="slot-abstract"> 26572 <div> 26573 <a 26574 class="clickable no-decoration" 26575 id="vhsjs_view_592_1707793552_530882" 26576 onclick="$('#vhsjs_view_592_1707793552_530882').hide(); 26577 $('#vhsjs_hide_592_1707793552_530882').show(); 26578 $('#591_1707793552_5308738').slideDown(function() { 26579 if (typeof Masonry === 'function') { 26580 $('.use_masonry').masonry(); 26581 }; 26582 26583 });" 26584 ><i class="fa fa-caret-right"></i>
26585 <span class="hover_link">Abstract</span></a 26586 ><a 26587 class="clickable no-decoration" 26588 id="vhsjs_hide_592_1707793552_530882" 26589 onclick="$('#591_1707793552_5308738').hide(function() { 26590 if (typeof Masonry === 'function') { 26591 $('.use_masonry').masonry(); 26592 }; 26593 }); 26594 $('#vhsjs_hide_592_1707793552_530882').hide(); 26595 $('#vhsjs_view_592_1707793552_530882').show();" 26596 style="display: none" 26597 ><i class="fa fa-caret-down"></i> 26598 <span class="hover_link">Abstract</span></a 26599 > 26600 <div 26601 data-display-control="592_1707793552_530882" 26602 id="591_1707793552_5308738" 26603 style="display: none" 26604 > 26605 <div class="arrow-slidedown"> 26606 <blockquote> 26607 We present RustSim, a library for discrete-event 26608 process-oriented simulations designed and 26609 implemented in Rust programming language. It 26610 includes a broad set of classes to allow the 26611 user to implement simulation processes and 26612 process-oriented primitives. The flexible 26613 modular design of RustSim allows users to extend 26614 its functionality. In addition, RustSim includes 26615 mechanisms to avoid inconsistencies when 26616 applying state-changing primitives that other 26617 libraries in the language's ecosystem do not 26618 provide. We take advantage of Rust generators 26619 (coroutine equivalent) to implement 26620 process-oriented simulation primitives. Finally, 26621 the library's internal process handling 26622 structure is discussed in detail, including its 26623 implementation, how simulations are executed, 26624 and a case study with a highly detailed example 26625 of its use. 26626 </blockquote> 26627 </div> 26628 </div> 26629 </div> 26630 </div> 26631 <div class="slot-urls"></div> 26632 <a href="/wsc23papers/262.pdf" target="_blank">pdf</a 26633 ><br /> 26634 </div> 26635 <div class="slot-entry"> 26636 <a name="satwcont105" tabindex="-1"></a> 26637 <div class="slot-title-line"> 26638 <span class="slot-title" 26639 >Modeling and Simulating Stream Processing 26640 Platforms</span 26641 > 26642 </div> 26643 <div class="slot-authors"> 26644 Alonso Inostrosa-Psijas (Universidad de 26645 Valparaíso); Veronica Gil-Costa (UNSL, CONICET); 26646 Roberto Solar and Mauricio Marin (Universidad de 26647 Santiago de Chile); and Gabriel Wainer (Carleton 26648 University) 26649 </div> 26650 <div class="slot-abstract"> 26651 <div> 26652 <a 26653 class="clickable no-decoration" 26654 id="vhsjs_view_594_1707793552_5332785" 26655 onclick="$('#vhsjs_view_594_1707793552_5332785').hide(); 26656 $('#vhsjs_hide_594_1707793552_5332785').show(); 26657 $('#593_1707793552_5332701').slideDown(function() { 26658 if (typeof Masonry === 'function') { 26659 $('.use_masonry').masonry(); 26660 }; 26661 26662 });" 26663 ><i class="fa fa-caret-right"></i>
26664 <span class="hover_link">Abstract</span></a 26665 ><a 26666 class="clickable no-decoration" 26667 id="vhsjs_hide_594_1707793552_5332785" 26668 onclick="$('#593_1707793552_5332701').hide(function() { 26669 if (typeof Masonry === 'function') { 26670 $('.use_masonry').masonry(); 26671 }; 26672 }); 26673 $('#vhsjs_hide_594_1707793552_5332785').hide(); 26674 $('#vhsjs_view_594_1707793552_5332785').show();" 26675 style="display: none" 26676 ><i class="fa fa-caret-down"></i> 26677 <span class="hover_link">Abstract</span></a 26678 > 26679 <div 26680 data-display-control="594_1707793552_5332785" 26681 id="593_1707793552_5332701" 26682 style="display: none" 26683 > 26684 <div class="arrow-slidedown"> 26685 <blockquote> 26686 Stream processing platforms allow processing and 26687 analyzing real-time data. Several tools have 26688 been developed for these platforms to guarantee 26689 that the applications running on them are 26690 scalable, fast, and fault-tolerant and that they 26691 can be deployed on many processors. However, 26692 determining the proper number of processors 26693 suitable to hold a given stream processing-based 26694 software application is challenging, especially 26695 if the application is intended to serve a large 26696 user community. In this paper, we propose to 26697 model and simulate stream processing platforms 26698 for performance evaluation purposes. In our case 26699 study, we simulated a commonly used application 26700 for the analysis of Twitter streams with Storm. 26701 We evaluate its performance under different 26702 workloads. Our simulator supports profiling to 26703 measure various aspects of the application's 26704 performance. Results show that the simulator can 26705 replicate the metrics reported by the 26706 application running on a real platform with 26707 minimal error. 26708 </blockquote> 26709 </div> 26710 </div> 26711 </div> 26712 </div> 26713 <div class="slot-urls"></div> 26714 <a href="/wsc23papers/263.pdf" target="_blank">pdf</a 26715 ><br /> 26716 </div> 26717 <div class="slot-entry"> 26718 <a name="satwcea111" tabindex="-1"></a> 26719 <div class="slot-title-line"> 26720 <span class="slot-title" 26721 >Using a Software Design Pattern for Redesign Routed 26722 DEVS Formalism</span 26723 > 26724 </div> 26725 <div class="slot-authors"> 26726 Mateo Toniolo, María Julia Blas, and Silvio Gonnet 26727 (Universidad Tecnológica Nacional - Facultad 26728 Regional Santa Fe) 26729 </div> 26730 <div class="slot-abstract"> 26731 <div> 26732 <a 26733 class="clickable no-decoration" 26734 id="vhsjs_view_596_1707793552_593096" 26735 onclick="$('#vhsjs_view_596_1707793552_593096').hide(); 26736 $('#vhsjs_hide_596_1707793552_593096').show(); 26737 $('#595_1707793552_5930824').slideDown(function() { 26738 if (typeof Masonry === 'function') { 26739 $('.use_masonry').masonry(); 26740 }; 26741 26742 });" 26743 ><i class="fa fa-caret-right"></i>
26744 <span class="hover_link">Abstract</span></a 26745 ><a 26746 class="clickable no-decoration" 26747 id="vhsjs_hide_596_1707793552_593096" 26748 onclick="$('#595_1707793552_5930824').hide(function() { 26749 if (typeof Masonry === 'function') { 26750 $('.use_masonry').masonry(); 26751 }; 26752 }); 26753 $('#vhsjs_hide_596_1707793552_593096').hide(); 26754 $('#vhsjs_view_596_1707793552_593096').show();" 26755 style="display: none" 26756 ><i class="fa fa-caret-down"></i> 26757 <span class="hover_link">Abstract</span></a 26758 > 26759 <div 26760 data-display-control="596_1707793552_593096" 26761 id="595_1707793552_5930824" 26762 style="display: none" 26763 > 26764 <div class="arrow-slidedown"> 26765 <blockquote> 26766 Routed DEVS (RDEVS) models improve traditional 26767 discrete-event models by enhancing the 26768 development of routing processes over predefined 26769 behaviors. In this paper, we demonstrate how a 26770 Software Engineering design pattern, 26771 specifically the Decorator pattern, was applied 26772 to the RDEVS formalism design to include event 26773 tracking into the models without altering their 26774 expected behavior. As a result, we provide a 26775 solution that allows getting structured data 26776 from RDEVS models at execution time. 26777 </blockquote> 26778 </div> 26779 </div> 26780 </div> 26781 </div> 26782 <div class="slot-urls"></div> 26783 <a href="/wsc23papers/satwcea111.pdf" target="_blank" 26784 >pdf</a 26785 ><br /> 26786 </div> 26787 </div> 26788 <div class="session-entry"> 26789 <span class="session-event-type">Technical Session</span 26790 ><span class="type-track-spacer"> · </span 26791 ><span class="program-track" 26792 >Simulation Around the World</span 26793 ><br /> 26794 <div class="session-title"> 26795 Agent-based and Healthcare Applications 26796 </div> 26797 <div class="session-chair"> 26798 Chair: Alonso Inostrosa Psijas (Universidad de 26799 Valparaíso)<br /> 26800 </div> 26801 <div class="slot-entry"> 26802 <a name="satwcont103" tabindex="-1"></a> 26803 <div class="slot-title-line"> 26804 <span class="slot-title" 26805 >Using a Hybrid ABMS to Study the Propagation of 26806 Vector-Borne Diseases in an Urban Area with 26807 Heterogenous Geospatial Conditions</span 26808 > 26809 </div> 26810 <div class="slot-authors"> 26811 Paula Escudero, Mariajose Franco, María Sofía 26812 Uribe, Susana Álvarez, and Rafael Mateus 26813 (Universidad EAFIT) 26814 </div> 26815 <div class="slot-abstract"> 26816 <div> 26817 <a 26818 class="clickable no-decoration" 26819 id="vhsjs_view_598_1707793552_5988085" 26820 onclick="$('#vhsjs_view_598_1707793552_5988085').hide(); 26821 $('#vhsjs_hide_598_1707793552_5988085').show(); 26822 $('#597_1707793552_598801').slideDown(function() { 26823 if (typeof Masonry === 'function') { 26824 $('.use_masonry').masonry(); 26825 }; 26826 26827 });" 26828 ><i class="fa fa-caret-right"></i>
26829 <span class="hover_link">Abstract</span></a 26830 ><a 26831 class="clickable no-decoration" 26832 id="vhsjs_hide_598_1707793552_5988085" 26833 onclick="$('#597_1707793552_598801').hide(function() { 26834 if (typeof Masonry === 'function') { 26835 $('.use_masonry').masonry(); 26836 }; 26837 }); 26838 $('#vhsjs_hide_598_1707793552_5988085').hide(); 26839 $('#vhsjs_view_598_1707793552_5988085').show();" 26840 style="display: none" 26841 ><i class="fa fa-caret-down"></i> 26842 <span class="hover_link">Abstract</span></a 26843 > 26844 <div 26845 data-display-control="598_1707793552_5988085" 26846 id="597_1707793552_598801" 26847 style="display: none" 26848 > 26849 <div class="arrow-slidedown"> 26850 <blockquote> 26851 Agent-Based Modeling and Simulation (ABMS) is a 26852 valuable tool for understanding infectious 26853 disease propagation. This study presents a 26854 hybrid ABMS approach to explore the transmission 26855 dynamics of vector-borne diseases (Dengue, Zika, 26856 and Chikungunya) in Bello, Colombia, 26857 incorporating geospatial characteristics. The 26858 model was developed with specific assumptions to 26859 validate its alignment with theoretical 26860 behavior. Our results demonstrate the 26861 temperature’s significant impact on 26862 disease spread. Particularly, Chikungunya 26863 exhibits distinct behavior compared to Dengue 26864 and Zika. While major infection peaks occur 26865 early in the simulation, subsequent spread 26866 diminishes due to the absence of reinfection 26867 considerations. This research represents an 26868 early stage of a larger project, laying the 26869 groundwork for future research to address 26870 computational challenges, enabling statistical 26871 analysis with multiple runs, and enhancing the 26872 model’s realism with seasonal temperature 26873 variations and geographical distributions. These 26874 findings will provide valuable insights for 26875 policymakers and disease control strategies in 26876 Colombia. 26877 </blockquote> 26878 </div> 26879 </div> 26880 </div> 26881 </div> 26882 <div class="slot-urls"></div> 26883 <a href="/wsc23papers/264.pdf" target="_blank">pdf</a 26884 ><br /> 26885 </div> 26886 <div class="slot-entry"> 26887 <a name="satwcea108" tabindex="-1"></a> 26888 <div class="slot-title-line"> 26889 <span class="slot-title" 26890 >Agent-Based Model for Analysis of Cervical Cancer 26891 Detection</span 26892 > 26893 </div> 26894 <div class="slot-authors"> 26895 Juan F. Galindo Jaramillo (University of Campinas, 26896 Hermínio Ometto Foundation) and Leonardo Grando, 26897 José Roberto Emiliano Leite, Diama Bhadra Vale, and 26898 Edson Ursini (University of Campinas) 26899 </div> 26900 <div class="slot-abstract"> 26901 <div> 26902 <a 26903 class="clickable no-decoration" 26904 id="vhsjs_view_600_1707793552_600959" 26905 onclick="$('#vhsjs_view_600_1707793552_600959').hide(); 26906 $('#vhsjs_hide_600_1707793552_600959').show(); 26907 $('#599_1707793552_6009512').slideDown(function() { 26908 if (typeof Masonry === 'function') { 26909 $('.use_masonry').masonry(); 26910 }; 26911 26912 });" 26913 ><i class="fa fa-caret-right"></i>
26914 <span class="hover_link">Abstract</span></a 26915 ><a 26916 class="clickable no-decoration" 26917 id="vhsjs_hide_600_1707793552_600959" 26918 onclick="$('#599_1707793552_6009512').hide(function() { 26919 if (typeof Masonry === 'function') { 26920 $('.use_masonry').masonry(); 26921 }; 26922 }); 26923 $('#vhsjs_hide_600_1707793552_600959').hide(); 26924 $('#vhsjs_view_600_1707793552_600959').show();" 26925 style="display: none" 26926 ><i class="fa fa-caret-down"></i> 26927 <span class="hover_link">Abstract</span></a 26928 > 26929 <div 26930 data-display-control="600_1707793552_600959" 26931 id="599_1707793552_6009512" 26932 style="display: none" 26933 > 26934 <div class="arrow-slidedown"> 26935 <blockquote> 26936 Using Agent-Based Models (ABM) for disease 26937 incidence may help decision-making processes. 26938 This work shows an ABM for cervical cancer 26939 detection. Our results show the relevance of 26940 social indicators. 26941 </blockquote> 26942 </div> 26943 </div> 26944 </div> 26945 </div> 26946 <div class="slot-urls"></div> 26947 <a href="/wsc23papers/satwcea108.pdf" target="_blank" 26948 >pdf</a 26949 ><br /> 26950 </div> 26951 <div class="slot-entry"> 26952 <a name="con114" tabindex="-1"></a> 26953 <div class="slot-title-line"> 26954 <span class="slot-title" 26955 >Coordination of Hospital Parking and Transportation 26956 Services: A Simulation-based Approach</span 26957 > 26958 </div> 26959 <div class="slot-authors"> 26960 Tomer Schmid, Dror Neustatel, and Noa Zychlinski 26961 (Technion–Israel Institute of Technology) 26962 </div> 26963 <div class="slot-abstract"> 26964 <div> 26965 <a 26966 class="clickable no-decoration" 26967 id="vhsjs_view_602_1707793552_6032376" 26968 onclick="$('#vhsjs_view_602_1707793552_6032376').hide(); 26969 $('#vhsjs_hide_602_1707793552_6032376').show(); 26970 $('#601_1707793552_6032293').slideDown(function() { 26971 if (typeof Masonry === 'function') { 26972 $('.use_masonry').masonry(); 26973 }; 26974 26975 });" 26976 ><i class="fa fa-caret-right"></i> 26977 <span class="hover_link">Abstract</span></a 26978 ><a 26979 class="clickable no-decoration" 26980 id="vhsjs_hide_602_1707793552_6032376" 26981 onclick="$('#601_1707793552_6032293').hide(function() { 26982 if (typeof Masonry === 'function') { 26983 $('.use_masonry').masonry(); 26984 }; 26985 }); 26986 $('#vhsjs_hide_602_1707793552_6032376').hide(); 26987 $('#vhsjs_view_602_1707793552_6032376').show();" 26988 style="display: none" 26989 ><i class="fa fa-caret-down"></i> 26990 <span class="hover_link">Abstract</span></a 26991 > 26992 <div
26993 data-display-control="602_1707793552_6032376" 26994 id="601_1707793552_6032293" 26995 style="display: none" 26996 > 26997 <div class="arrow-slidedown"> 26998 <blockquote> 26999 Motivated by hospital parking problems that 27000 limit the access of patients and visitors, we 27001 study a hospital parking setting comprising an 27002 on-site parking lot with an occupancy-based 27003 dynamic tariff and a free shuttle service from 27004 an off-site free parking lot. We developed a 27005 discrete event simulation model to study the 27006 system’s dynamics and find the preferable 27007 coordinated tariff and shuttle schedule that 27008 maximize revenue for the contractor operating 27009 the hospital’s parking services under a 27010 predefined service level. We use a case study 27011 from Hadassah Medical Center in Ein Kerem, 27012 Jerusalem, to demonstrate the effectiveness of 27013 our method. Our results show that the 27014 coordinated solution provides significantly 27015 better performance: more than a 30% increase in 27016 service level, a 25% (about $5,000) increase in 27017 daily revenue, and a 53% decrease in average 27018 waiting time for a shuttle. 27019 </blockquote> 27020 </div> 27021 </div> 27022 </div> 27023 </div> 27024 <div class="slot-urls"></div> 27025 <a href="/wsc23papers/143.pdf" target="_blank">pdf</a 27026 ><br /> 27027 </div> 27028 </div> 27029 <div class="session-entry"> 27030 <span class="session-event-type">Technical Session</span 27031 ><span class="type-track-spacer"> · </span 27032 ><span class="program-track" 27033 >Simulation Around the World</span 27034 ><br /> 27035 <div class="session-title"> 27036 Simulation Applications in Africa 27037 </div> 27038 <div class="session-chair"> 27039 Chair: Simon J. E. Taylor (Brunel University London)<br /> 27040 </div> 27041 <div class="slot-entry"> 27042 <a name="satwcea110" tabindex="-1"></a> 27043 <div class="slot-title-line"> 27044 <span class="slot-title" 27045 >Weather Prediction Simulations for East Africa</span 27046 > 27047 </div> 27048 <div class="slot-authors"> 27049 Julianne Sansa-Otim (Makerere University), Isaac Mugume 27050 (Uganda National Meteorological Authority), and Mary 27051 Nsabagwa (Makerere University) 27052 </div> 27053 <div class="slot-abstract"> 27054 <div> 27055 <a 27056 class="clickable no-decoration" 27057 id="vhsjs_view_604_1707793552_6083174" 27058 onclick="$('#vhsjs_view_604_1707793552_6083174').hide(); 27059 $('#vhsjs_hide_604_1707793552_6083174').show(); 27060 $('#603_1707793552_6083093').slideDown(function() { 27061 if (typeof Masonry === 'function') { 27062 $('.use_masonry').masonry(); 27063 }; 27064 27065 });" 27066 ><i class="fa fa-caret-right"></i>
27067 <span class="hover_link">Abstract</span></a 27068 ><a 27069 class="clickable no-decoration" 27070 id="vhsjs_hide_604_1707793552_6083174" 27071 onclick="$('#603_1707793552_6083093').hide(function() { 27072 if (typeof Masonry === 'function') { 27073 $('.use_masonry').masonry(); 27074 }; 27075 }); 27076 $('#vhsjs_hide_604_1707793552_6083174').hide(); 27077 $('#vhsjs_view_604_1707793552_6083174').show();" 27078 style="display: none" 27079 ><i class="fa fa-caret-down"></i> 27080 <span class="hover_link">Abstract</span></a 27081 > 27082 <div 27083 data-display-control="604_1707793552_6083174" 27084 id="603_1707793552_6083093" 27085 style="display: none" 27086 > 27087 <div class="arrow-slidedown"> 27088 <blockquote> 27089 Numerical weather prediction (NWP) contributes 27090 significantly in the production of appropriate 27091 weather forecasts. These critical capabilities 27092 were still largely lacking in East Africa in the 27093 early 2010s and were recently established under 27094 the auspices of the WIMEA-ICT Project. The 27095 project introduced the use of the Weather 27096 Research and Forecasting (WRF) model in the 27097 region. This model was adopted by the National 27098 Hydro-meteorological Agencies and is largely 27099 being used as guidance in the operations. 27100 However, due to advances in technology, there is 27101 a need to build capacity in NWP data 27102 assimilation as well as Machine Learning to 27103 further improve the accuracy of weather and 27104 climatic predictions. Additional crop weather 27105 modelling studies will further inform 27106 agricultural productivity enhancement in the 27107 region. 27108 </blockquote> 27109 </div> 27110 </div> 27111 </div> 27112 </div> 27113 <div class="slot-urls"></div> 27114 <a href="/wsc23papers/satwcea110.pdf" target="_blank" 27115 >pdf</a 27116 ><br /> 27117 </div> 27118 <div class="slot-entry"> 27119 <a name="satwcea106" tabindex="-1"></a> 27120 <div class="slot-title-line"> 27121 <span class="slot-title" 27122 >Challenges of Using Simulation for Healthcare 27123 Operations Management in Developing Countries: The 27124 Case of Ethiopia</span 27125 > 27126 </div> 27127 <div class="slot-authors"> 27128 Tesfamariam M. Abuhay (University of Gondar, Queen's 27129 University); Mihret Woldesemayat Tereda, Lomi Eyachew 27130 Adane, and Malefia Demilie Melesse (University of 27131 Gondar); Stewart Robinson (Newcastle University); and 27132 Vedat Verter (Queen's University) 27133 </div> 27134 <div class="slot-abstract"> 27135 <div> 27136 <a 27137 class="clickable no-decoration" 27138 id="vhsjs_view_606_1707793552_6106775" 27139 onclick="$('#vhsjs_view_606_1707793552_6106775').hide(); 27140 $('#vhsjs_hide_606_1707793552_6106775').show(); 27141 $('#605_1707793552_6106694').slideDown(function() { 27142 if (typeof Masonry === 'function') { 27143 $('.use_masonry').masonry(); 27144 }; 27145 27146 });" 27147 ><i class="fa fa-caret-right"></i>
27148 <span class="hover_link">Abstract</span></a 27149 ><a 27150 class="clickable no-decoration" 27151 id="vhsjs_hide_606_1707793552_6106775" 27152 onclick="$('#605_1707793552_6106694').hide(function() { 27153 if (typeof Masonry === 'function') { 27154 $('.use_masonry').masonry(); 27155 }; 27156 }); 27157 $('#vhsjs_hide_606_1707793552_6106775').hide(); 27158 $('#vhsjs_view_606_1707793552_6106775').show();" 27159 style="display: none" 27160 ><i class="fa fa-caret-down"></i> 27161 <span class="hover_link">Abstract</span></a 27162 > 27163 <div 27164 data-display-control="606_1707793552_6106775" 27165 id="605_1707793552_6106694" 27166 style="display: none" 27167 > 27168 <div class="arrow-slidedown"> 27169 <blockquote> 27170 Simulation models have been employed in 27171 developed countries for healthcare service 27172 operations management. However, leveraging 27173 simulation in developing countries is limited 27174 because healthcare operations management 27175 challenges are quite different due to scarcity 27176 of resources, high population numbers, high 27177 healthcare demand, and poor planning, 27178 implementation, monitoring and evaluation. This 27179 study, hence, aims to investigate the usage and 27180 adoption of simulation for healthcare operations 27181 management in developing countries and the 27182 challenges of using simulation in this context 27183 by studying the case of Ethiopia through a 27184 systematic literature review and survey. 27185 </blockquote> 27186 </div> 27187 </div> 27188 </div> 27189 </div> 27190 <div class="slot-urls"></div> 27191 <a href="/wsc23papers/satwcea106.pdf" target="_blank" 27192 >pdf</a 27193 ><br /> 27194 </div> 27195 <div class="slot-entry"> 27196 <a name="con300" tabindex="-1"></a> 27197 <div class="slot-title-line"> 27198 <span class="slot-title" 27199 >Hybrid Approaches for Handling Mobile Crane Location 27200 Problems in Construction Sites</span 27201 > 27202 </div> 27203 <div class="slot-authors"> 27204 Khaoula Boutouhami, Rafik Lemouchi, and Mohamed Assaf 27205 (University of Alberta); Ahmed Bouferguene (university 27206 of alberta); Mohamed Al-Hussein (University of Alberta); 27207 and Joe Kosa (NCSG Crane and Heavy Haul Services) 27208 </div> 27209 <div class="slot-abstract"> 27210 <div> 27211 <a 27212 class="clickable no-decoration" 27213 id="vhsjs_view_608_1707793552_6132512" 27214 onclick="$('#vhsjs_view_608_1707793552_6132512').hide(); 27215 $('#vhsjs_hide_608_1707793552_6132512').show(); 27216 $('#607_1707793552_613243').slideDown(function() { 27217 if (typeof Masonry === 'function') { 27218 $('.use_masonry').masonry(); 27219 }; 27220 27221 });" 27222 ><i class="fa fa-caret-right"></i>
27223 <span class="hover_link">Abstract</span></a 27224 ><a 27225 class="clickable no-decoration" 27226 id="vhsjs_hide_608_1707793552_6132512" 27227 onclick="$('#607_1707793552_613243').hide(function() { 27228 if (typeof Masonry === 'function') { 27229 $('.use_masonry').masonry(); 27230 }; 27231 }); 27232 $('#vhsjs_hide_608_1707793552_6132512').hide(); 27233 $('#vhsjs_view_608_1707793552_6132512').show();" 27234 style="display: none" 27235 ><i class="fa fa-caret-down"></i> 27236 <span class="hover_link">Abstract</span></a 27237 > 27238 <div 27239 data-display-control="608_1707793552_6132512" 27240 id="607_1707793552_613243" 27241 style="display: none" 27242 > 27243 <div class="arrow-slidedown"> 27244 <blockquote> 27245 Mobile crane location (MCL) in modular 27246 construction is a complex problem that affects 27247 both construction safety and efficiency. 27248 Sub-optimal MCL planning increases the number of 27249 crane relocations and the overall project cost. 27250 Interestingly, recently, research on crane 27251 operation planning and analysis focused on 27252 determining crane configurations, boom lengths, 27253 and radii to enable lifting given a crane 27254 location. However, with a large number of 27255 feasible locations, finding the best solution 27256 becomes a harder task. In this respect, finding 27257 a single crane location ensures an optimal lift 27258 plan, e.g., minimizing the number of 27259 pick-location. As a result, this paper aims to 27260 bridge this gap by providing a hybrid approach 27261 using heuristics, grid-based, and combinatorial 27262 optimization algorithms to find the least 27263 required lifting points. The proposed approach 27264 is tested on a case study of a modular building. 27265 The study contributes by minimizing the number 27266 of crane relocations to enhance budget and cost 27267 planning. 27268 </blockquote> 27269 </div> 27270 </div> 27271 </div> 27272 </div> 27273 <div class="slot-urls"></div> 27274 <a href="/wsc23papers/228.pdf" target="_blank">pdf</a 27275 ><br /> 27276 </div> 27277 </div> 27278 <div class="session-entry"> 27279 <span class="session-event-type">Technical Session</span 27280 ><span class="type-track-spacer"> · </span 27281 ><span class="program-track" 27282 >Simulation Around the World</span 27283 ><br /> 27284 <div class="session-title"> 27285 Decision Making with Discrete-event Simulation I 27286 </div> 27287 <div class="session-chair"> 27288 Chair: Stewart Robinson (Newcastle University)<br /> 27289 </div> 27290 <div class="slot-entry"> 27291 <a name="satwcea107" tabindex="-1"></a> 27292 <div class="slot-title-line"> 27293 <span class="slot-title" 27294 >Modeling and Simulation for Farming Drone Battery 27295 Recharging</span 27296 > 27297 </div> 27298 <div class="slot-authors"> 27299 Leonardo Grando (University of Campinas); Juan F. 27300 Galindo Jaramillo (University of Campinas, Herminio 27301 Ometto Foundation); and José Roberto Emiliano Leite 27302 and Edson Luiz Ursini (University of Campinas) 27303 </div> 27304 <div class="slot-abstract"> 27305 <div> 27306 <a 27307 class="clickable no-decoration" 27308 id="vhsjs_view_610_1707793552_6183665" 27309 onclick="$('#vhsjs_view_610_1707793552_6183665').hide(); 27310 $('#vhsjs_hide_610_1707793552_6183665').show(); 27311 $('#609_1707793552_6183586').slideDown(function() { 27312 if (typeof Masonry === 'function') { 27313 $('.use_masonry').masonry(); 27314 }; 27315 27316 });" 27317 ><i class="fa fa-caret-right"></i>
27318 <span class="hover_link">Abstract</span></a 27319 ><a 27320 class="clickable no-decoration" 27321 id="vhsjs_hide_610_1707793552_6183665" 27322 onclick="$('#609_1707793552_6183586').hide(function() { 27323 if (typeof Masonry === 'function') { 27324 $('.use_masonry').masonry(); 27325 }; 27326 }); 27327 $('#vhsjs_hide_610_1707793552_6183665').hide(); 27328 $('#vhsjs_view_610_1707793552_6183665').show();" 27329 style="display: none" 27330 ><i class="fa fa-caret-down"></i> 27331 <span class="hover_link">Abstract</span></a 27332 > 27333 <div 27334 data-display-control="610_1707793552_6183665" 27335 id="609_1707793552_6183586" 27336 style="display: none" 27337 > 27338 <div class="arrow-slidedown"> 27339 <blockquote> 27340 The Connected Farm is composed of several 27341 elements that communicate with each other 27342 through a 4G/5G Radio Base Station (RBS) placed 27343 in the middle of the farm. This RBS is connected 27344 to the Internet, allowing communication for all 27345 kinds of autonomous devices, performing 27346 uninterrupted tasks. This work simulates the 27347 Connected Farm environment for an autonomous 27348 drone. Our model intends to define when each 27349 drone needs to recharge its batteries, with no 27350 collusion regarding this recharging decision, 27351 reducing the drone's battery usage due to the 27352 absence of this communication. 27353 </blockquote> 27354 </div> 27355 </div> 27356 </div> 27357 </div> 27358 <div class="slot-urls"></div> 27359 <a href="/wsc23papers/satwcea107.pdf" target="_blank" 27360 >pdf</a 27361 ><br /> 27362 </div> 27363 <div class="slot-entry"> 27364 <a name="satwcont101" tabindex="-1"></a> 27365 <div class="slot-title-line"> 27366 <span class="slot-title" 27367 >Simulating the Social Influence in Transport Mode 27368 Choices</span 27369 > 27370 </div> 27371 <div class="slot-authors"> 27372 Kathleen Salazar-Serna (Pontificia Universidad 27373 Javeriana, Universidad Nacional de Colombia); Lynnette 27374 Hui Xian Ng (Carnegie Mellon University); Lorena Cadavid 27375 and Carlos Jaime Franco (Universidad Nacional de 27376 Colombia); and Kathleen M. Carley (Carnegie Mellon 27377 University) 27378 </div> 27379 <div class="slot-abstract"> 27380 <div> 27381 <a 27382 class="clickable no-decoration" 27383 id="vhsjs_view_612_1707793552_6208467" 27384 onclick="$('#vhsjs_view_612_1707793552_6208467').hide(); 27385 $('#vhsjs_hide_612_1707793552_6208467').show(); 27386 $('#611_1707793552_6208389').slideDown(function() { 27387 if (typeof Masonry === 'function') { 27388 $('.use_masonry').masonry(); 27389 }; 27390 27391 });" 27392 ><i class="fa fa-caret-right"></i>
27393 <span class="hover_link">Abstract</span></a 27394 ><a 27395 class="clickable no-decoration" 27396 id="vhsjs_hide_612_1707793552_6208467" 27397 onclick="$('#611_1707793552_6208389').hide(function() { 27398 if (typeof Masonry === 'function') { 27399 $('.use_masonry').masonry(); 27400 }; 27401 }); 27402 $('#vhsjs_hide_612_1707793552_6208467').hide(); 27403 $('#vhsjs_view_612_1707793552_6208467').show();" 27404 style="display: none" 27405 ><i class="fa fa-caret-down"></i> 27406 <span class="hover_link">Abstract</span></a 27407 > 27408 <div 27409 data-display-control="612_1707793552_6208467" 27410 id="611_1707793552_6208389" 27411 style="display: none" 27412 > 27413 <div class="arrow-slidedown"> 27414 <blockquote> 27415 Agent-based simulations have been used in 27416 modeling transportation systems for traffic 27417 management and passenger flows. In this work, we 27418 hope to shed light on the complex factors that 27419 influence transportation mode decisions within 27420 developing countries, using Colombia as a case 27421 study. We model an ecosystem of human agents 27422 that decide at each time step on the mode of 27423 transportation they would take to work. Their 27424 decision is based on a combination of their 27425 personal satisfaction with the journey they had 27426 just taken, which is evaluated across a personal 27427 vector of needs, the information they 27428 crowdsource from their prevailing social 27429 network, and their personal uncertainty about 27430 the discomfort of trying a new transport 27431 solution. We simulate different network 27432 structures to analyze the social influence for 27433 different decision-makers. We find that in 27434 low/medium connected groups inquisitive people 27435 actively change modes cyclically over the years 27436 while imitators cluster rapidly and change less 27437 frequently. 27438 </blockquote> 27439 </div> 27440 </div> 27441 </div> 27442 </div> 27443 <div class="slot-urls"></div> 27444 <a href="/wsc23papers/265.pdf" target="_blank">pdf</a 27445 ><br /> 27446 </div> 27447 </div> 27448 <div class="session-entry"> 27449 <span class="session-event-type">Technical Session</span 27450 ><span class="type-track-spacer"> · </span 27451 ><span class="program-track" 27452 >Simulation Around the World</span 27453 ><br /> 27454 <div class="session-title"> 27455 Decision Making with Discrete-event Simulation II 27456 </div> 27457 <div class="session-chair"> 27458 Chair: Cristina Ruiz-Martín (Carleton University)<br /> 27459 </div> 27460 <div class="slot-entry"> 27461 <a name="satwcont102" tabindex="-1"></a> 27462 <div class="slot-title-line">
27463 <span class="slot-title" 27464 >A Simulation-Optimization Approach for Designing 27465 Resilient Hyperconnected Physical Internet Supply 27466 Chains</span 27467 > 27468 </div> 27469 <div class="slot-authors"> 27470 Rafael D. Tordecilla, Jairo R. Montoya-Torres, and 27471 William J. Guerrero (Universidad de La Sabana) 27472 </div> 27473 <div class="slot-abstract"> 27474 <div> 27475 <a 27476 class="clickable no-decoration" 27477 id="vhsjs_view_614_1707793552_6259508" 27478 onclick="$('#vhsjs_view_614_1707793552_6259508').hide(); 27479 $('#vhsjs_hide_614_1707793552_6259508').show(); 27480 $('#613_1707793552_6259425').slideDown(function() { 27481 if (typeof Masonry === 'function') { 27482 $('.use_masonry').masonry(); 27483 }; 27484 27485 });" 27486 ><i class="fa fa-caret-right"></i> 27487 <span class="hover_link">Abstract</span></a 27488 ><a 27489 class="clickable no-decoration" 27490 id="vhsjs_hide_614_1707793552_6259508" 27491 onclick="$('#613_1707793552_6259425').hide(function() { 27492 if (typeof Masonry === 'function') { 27493 $('.use_masonry').masonry(); 27494 }; 27495 }); 27496 $('#vhsjs_hide_614_1707793552_6259508').hide(); 27497 $('#vhsjs_view_614_1707793552_6259508').show();" 27498 style="display: none" 27499 ><i class="fa fa-caret-down"></i> 27500 <span class="hover_link">Abstract</span></a 27501 > 27502 <div 27503 data-display-control="614_1707793552_6259508" 27504 id="613_1707793552_6259425" 27505 style="display: none" 27506 > 27507 <div class="arrow-slidedown"> 27508 <blockquote> 27509 The Physical Internet (PI) is a recent paradigm 27510 in the supply chain management that proposes a 27511 framework in which standardization and 27512 optimization are key factors to raise supply 27513 chain efficiency, resilience, and 27514 sustainability. Strategic decisions are included 27515 in the PI, including the supply chain network 27516 design (SCND). In fact, structuring a (near) 27517 optimal design is essential to achieve the PI 27518 objectives. Additionally, disruptive events such 27519 as the COVID-19 pandemic, earthquakes, or 27520 terrorist attacks threaten the supply chains. 27521 These events are difficult to predict, but their 27522 effects can be simulated when addressing this 27523 problem. Hence, we propose a 27524 simulation-optimization approach that hybridizes 27525 a multi-objective multi-period mixed-integer 27526 program with discrete-event simulation to 27527 optimize both cost and resilience in the SCND. 27528 Furthermore, a network hyperconnection strategy 27529 is tested. Results show that both resilience and 27530 risk are improved after hyperconnecting the 27531 supply chain, especially when active edges are 27532 disturbed, but incur higher costs. 27533 </blockquote> 27534 </div> 27535 </div> 27536 </div> 27537 </div> 27538 <div class="slot-urls"></div> 27539 <a href="/wsc23papers/267.pdf" target="_blank">pdf</a 27540 ><br /> 27541 </div> 27542 <div class="slot-entry"> 27543 <a name="satwcea105" tabindex="-1"></a> 27544 <div class="slot-title-line">
27545 <span class="slot-title" 27546 >Formal Modeling and Simulation of Economic Complexity 27547 Networks with Emergent Behavior-DEVS</span 27548 > 27549 </div> 27550 <div class="slot-authors"> 27551 Tobias Carreira Munich and Rodrigo Castro (Departamento 27552 de Computación, FCEyN-UBA / Instituto de Ciencias 27553 de la Computación (ICC-CONICET)) 27554 </div> 27555 <div class="slot-abstract"> 27556 <div> 27557 <a 27558 class="clickable no-decoration" 27559 id="vhsjs_view_616_1707793552_6279607" 27560 onclick="$('#vhsjs_view_616_1707793552_6279607').hide(); 27561 $('#vhsjs_hide_616_1707793552_6279607').show(); 27562 $('#615_1707793552_6279523').slideDown(function() { 27563 if (typeof Masonry === 'function') { 27564 $('.use_masonry').masonry(); 27565 }; 27566 27567 });" 27568 ><i class="fa fa-caret-right"></i> 27569 <span class="hover_link">Abstract</span></a 27570 ><a 27571 class="clickable no-decoration" 27572 id="vhsjs_hide_616_1707793552_6279607" 27573 onclick="$('#615_1707793552_6279523').hide(function() { 27574 if (typeof Masonry === 'function') { 27575 $('.use_masonry').masonry(); 27576 }; 27577 }); 27578 $('#vhsjs_hide_616_1707793552_6279607').hide(); 27579 $('#vhsjs_view_616_1707793552_6279607').show();" 27580 style="display: none" 27581 ><i class="fa fa-caret-down"></i> 27582 <span class="hover_link">Abstract</span></a 27583 > 27584 <div 27585 data-display-control="616_1707793552_6279607" 27586 id="615_1707793552_6279523" 27587 style="display: none" 27588 > 27589 <div class="arrow-slidedown"> 27590 <blockquote> 27591 We present an application of the EB-DEVS 27592 modelling framework for agent-based complex 27593 adaptive systems to a systematic study of the 27594 international Product Space network in the field 27595 of Economic Complexity. The evolution of the 27596 production structure of agents (countries) 27597 becomes mutually determined by an emerging 27598 macroscopic network (resulting from the 27599 worldwide trade). This framework allows to make 27600 prospective analysis about the productive 27601 structure of countries. 27602 </blockquote> 27603 </div> 27604 </div> 27605 </div> 27606 </div> 27607 <div class="slot-urls"></div> 27608 <a href="/wsc23papers/satwcea105.pdf" target="_blank" 27609 >pdf</a 27610 ><br /> 27611 </div> 27612 <div class="slot-entry"> 27613 <a name="satwcea112" tabindex="-1"></a> 27614 <div class="slot-title-line"> 27615 <span class="slot-title" 27616 >Predicting Job Waiting Times in a Stochastic 27617 Scheduling Environment Using Simulation and Regression 27618 Machine Learning Models</span 27619 > 27620 </div> 27621 <div class="slot-authors"> 27622 Ivan Kristianto Singgih (University of Surabaya, The 27623 Indonesian Researcher Association in South Korea) and 27624 Stefanus Soegiharto (University of Surabaya) 27625 </div> 27626 <div class="slot-abstract"> 27627 <div> 27628 <a 27629 class="clickable no-decoration" 27630 id="vhsjs_view_618_1707793552_6300383" 27631 onclick="$('#vhsjs_view_618_1707793552_6300383').hide(); 27632 $('#vhsjs_hide_618_1707793552_6300383').show(); 27633 $('#617_1707793552_6300302').slideDown(function() { 27634 if (typeof Masonry === 'function') { 27635 $('.use_masonry').masonry(); 27636 }; 27637 27638 });" 27639 ><i class="fa fa-caret-right"></i>
27640 <span class="hover_link">Abstract</span></a 27641 ><a 27642 class="clickable no-decoration" 27643 id="vhsjs_hide_618_1707793552_6300383" 27644 onclick="$('#617_1707793552_6300302').hide(function() { 27645 if (typeof Masonry === 'function') { 27646 $('.use_masonry').masonry(); 27647 }; 27648 }); 27649 $('#vhsjs_hide_618_1707793552_6300383').hide(); 27650 $('#vhsjs_view_618_1707793552_6300383').show();" 27651 style="display: none" 27652 ><i class="fa fa-caret-down"></i> 27653 <span class="hover_link">Abstract</span></a 27654 > 27655 <div 27656 data-display-control="618_1707793552_6300383" 27657 id="617_1707793552_6300302" 27658 style="display: none" 27659 > 27660 <div class="arrow-slidedown"> 27661 <blockquote> 27662 Scheduling real systems is complicated because 27663 of the consideration of various working 27664 conditions. Although various combinatorial 27665 optimization methods, ranging from mathematical 27666 models, heuristics, metaheuristics, etc., have 27667 been developed, these methods could require a 27668 long computational time due to the complexity of 27669 the problems. This study proposes a framework to 27670 understand the system’s behavior using 27671 regression machine learning techniques. The 27672 considered system could be any type, e.g., the 27673 flow shop, job shop, and their variants, with a 27674 certain scheduling method. The framework 27675 consists of (1) the development of the 27676 simulation for generating the data and (2) how 27677 the data could be used for training the 27678 regression machine learning models. An example 27679 of the stochastic single-machine problem with 27680 the First-In-First-Out rule is considered. The 27681 framework could be used to simplify the process 27682 of understanding the system’s behavior 27683 without any necessity to solve the optimization 27684 problem, which could be time-consuming. 27685 </blockquote> 27686 </div> 27687 </div> 27688 </div> 27689 </div> 27690 <div class="slot-urls"></div> 27691 <a href="/wsc23papers/satwcea112.pdf" target="_blank" 27692 >pdf</a 27693 ><br /> 27694 </div> 27695 </div> 27696 <div class="session-entry"> 27697 <span class="session-event-type">Technical Session</span 27698 ><span class="type-track-spacer"> · </span 27699 ><span class="program-track" 27700 >Simulation Around the World</span 27701 ><br /> 27702 <div class="session-title">Post-disaster Relief</div> 27703 <div class="session-chair"> 27704 Chair: Enver Yucesan (INSEAD)<br /> 27705 </div> 27706 <div class="slot-entry"> 27707 <a name="satwcea116" tabindex="-1"></a> 27708 <div class="slot-title-line"> 27709 <span class="slot-title" 27710 >An Agent-Based Modeling to Simulate the Dynamics of 27711 First Responders and Evacuees in Post-Disaster 27712 Scenarios</span 27713 > 27714 </div> 27715 <div class="slot-authors"> 27716 Amirreza Pashapour and F. Sibel Salman (Koc University), 27717 Sridhar R. Tayur (Carnegie Mellon University), and 27718 Barış Yıldız (Koc University) 27719 </div> 27720 <div class="slot-abstract"> 27721 <div> 27722 <a 27723 class="clickable no-decoration" 27724 id="vhsjs_view_620_1707793552_6363072" 27725 onclick="$('#vhsjs_view_620_1707793552_6363072').hide(); 27726 $('#vhsjs_hide_620_1707793552_6363072').show(); 27727 $('#619_1707793552_6362987').slideDown(function() { 27728 if (typeof Masonry === 'function') { 27729 $('.use_masonry').masonry(); 27730 }; 27731 27732 });" 27733 ><i class="fa fa-caret-right"></i>
27734 <span class="hover_link">Abstract</span></a 27735 ><a 27736 class="clickable no-decoration" 27737 id="vhsjs_hide_620_1707793552_6363072" 27738 onclick="$('#619_1707793552_6362987').hide(function() { 27739 if (typeof Masonry === 'function') { 27740 $('.use_masonry').masonry(); 27741 }; 27742 }); 27743 $('#vhsjs_hide_620_1707793552_6363072').hide(); 27744 $('#vhsjs_view_620_1707793552_6363072').show();" 27745 style="display: none" 27746 ><i class="fa fa-caret-down"></i> 27747 <span class="hover_link">Abstract</span></a 27748 > 27749 <div 27750 data-display-control="620_1707793552_6363072" 27751 id="619_1707793552_6362987" 27752 style="display: none" 27753 > 27754 <div class="arrow-slidedown"> 27755 <blockquote> 27756 In the aftermath of a sudden catastrophe, First 27757 Responders (FR) strive to promptly reach and 27758 rescue victims. Simultaneously, individuals take 27759 roads to evacuate the affected region, access 27760 medical facilities or shelters, and reunite with 27761 their relatives. The escalated traffic 27762 congestion significantly hinders critical FR 27763 operations. In this study, we construct an 27764 Agent-Based Simulation (ABS) model that extends 27765 the existing models by incorporating FR agents, 27766 their allocated road map, and their interaction 27767 with evacuees in the network. Our model 27768 investigates individuals' evacuation times as 27769 well as FRs' rescue operation performance, 27770 provided that a subset of road segments are 27771 reserved for the explicit use of FRs. The 27772 decision-maker can allocate these segments 27773 manually within the simulation interface. 27774 Subsequently, the consequences are discovered 27775 through the earthquake scenario outputs of the 27776 ABS model, casting light on its real-world 27777 impact. 27778 </blockquote> 27779 </div> 27780 </div> 27781 </div> 27782 </div> 27783 <div class="slot-urls"></div> 27784 <a href="/wsc23papers/satwcea116.pdf" target="_blank" 27785 >pdf</a 27786 ><br /> 27787 </div> 27788 <div class="slot-entry"> 27789 <a name="satwcea115" tabindex="-1"></a> 27790 <div class="slot-title-line"> 27791 <span class="slot-title" 27792 >Optimization of Battery Allocation for 27793 Post-Earthquake Damage Assessment Using Drones</span 27794 > 27795 </div> 27796 <div class="slot-authors"> 27797 Selver Tugba Yaldiz (Marmara University) and Elvin Coban 27798 (Ozyegin University) 27799 </div> 27800 <div class="slot-abstract"> 27801 <div> 27802 <a 27803 class="clickable no-decoration" 27804 id="vhsjs_view_622_1707793552_6383545" 27805 onclick="$('#vhsjs_view_622_1707793552_6383545').hide(); 27806 $('#vhsjs_hide_622_1707793552_6383545').show(); 27807 $('#621_1707793552_6383462').slideDown(function() { 27808 if (typeof Masonry === 'function') { 27809 $('.use_masonry').masonry(); 27810 }; 27811 27812 });" 27813 ><i class="fa fa-caret-right"></i>
27814 <span class="hover_link">Abstract</span></a 27815 ><a 27816 class="clickable no-decoration" 27817 id="vhsjs_hide_622_1707793552_6383545" 27818 onclick="$('#621_1707793552_6383462').hide(function() { 27819 if (typeof Masonry === 'function') { 27820 $('.use_masonry').masonry(); 27821 }; 27822 }); 27823 $('#vhsjs_hide_622_1707793552_6383545').hide(); 27824 $('#vhsjs_view_622_1707793552_6383545').show();" 27825 style="display: none" 27826 ><i class="fa fa-caret-down"></i> 27827 <span class="hover_link">Abstract</span></a 27828 > 27829 <div 27830 data-display-control="622_1707793552_6383545" 27831 id="621_1707793552_6383462" 27832 style="display: none" 27833 > 27834 <div class="arrow-slidedown"> 27835 <blockquote> 27836 Earthquakes are one of the most common natural 27837 disasters and assessing the hazard levels of the 27838 affected regions and planning post-disaster 27839 operations, including search and rescue 27840 operations, are very critical. As the roads can 27841 be blocked due to an earthquake and debris 27842 removal may take time preventing critical rescue 27843 operations from starting, drone utilization has 27844 been increasing. Since the drones fly, it will 27845 be easier to assess the damage levels. However, 27846 drones have a major drawback, their batteries. 27847 In this study, we propose a scenario-based 27848 mathematical model to allocate a limited of 27849 batteries before the earthquake while computing 27850 the drones’ paths for each scenario 27851 maximizing the total expected priority scores. 27852 Our preliminary analysis shows that small 27853 instances can be solved very efficiently. 27854 </blockquote> 27855 </div> 27856 </div> 27857 </div> 27858 </div> 27859 <div class="slot-urls"></div> 27860 <a href="/wsc23papers/satwcea115.pdf" target="_blank" 27861 >pdf</a 27862 ><br /> 27863 </div> 27864 </div> 27865 </div> 27866 <div class="centered"> 27867 <div class="top-link"><a href="#top">Return to Top</a></div> 27868 </div> 27869 <hr /> 27870 </div> 27871 <div class="area-section"> 27872 <div class="centered"> 27873 <a name="ptrack119" tabindex="-1"></a> 27874 <div class="section-title"> 27875 Simulation and Artificial Intelligence 27876 </div> 27877 </div> 27878 <div class="centered track-chair"> 27879 <span class="track-chair-role" 27880 >Track Coordinator - Simulation and Artificial Intelligence: </span 27881 ><span class="track-chair-names" 27882 >Edward Y. Hua (MITRE Corporation), Yijie Peng (Peking 27883 University), Simon J. E. Taylor (Brunel University 27884 London)</span 27885 > 27886 </div> 27887 <div class="section-entry"> 27888 <div class="session-entry"> 27889 <span class="session-event-type">Technical Session</span 27890 ><span class="type-track-spacer"> · </span 27891 ><span class="program-track" 27892 >Simulation and Artificial Intelligence</span 27893 ><br /> 27894 <div class="session-title">Simulation Methodologies</div> 27895 <div class="session-chair"> 27896 Chair: Yifan Lin (Georgia Institute of Technology)<br /> 27897 </div> 27898 <div class="slot-entry"> 27899 <a name="con179" tabindex="-1"></a> 27900 <div class="slot-title-line">
27901 <span class="slot-title" 27902 >Generating Population Synthesis Using a Diffusion 27903 Model</span 27904 > 27905 </div> 27906 <div class="slot-authors"> 27907 Jaewoong Kang, Young Kim, Muhammad Mu’az Imran, 27908 Gi-sun Jung, and Yun Bae Kim (Sungkyunkwan University) 27909 </div> 27910 <div class="slot-abstract"> 27911 <div> 27912 <a 27913 class="clickable no-decoration" 27914 id="vhsjs_view_624_1707793552_6461422" 27915 onclick="$('#vhsjs_view_624_1707793552_6461422').hide(); 27916 $('#vhsjs_hide_624_1707793552_6461422').show(); 27917 $('#623_1707793552_6461344').slideDown(function() { 27918 if (typeof Masonry === 'function') { 27919 $('.use_masonry').masonry(); 27920 }; 27921 27922 });" 27923 ><i class="fa fa-caret-right"></i> 27924 <span class="hover_link">Abstract</span></a 27925 ><a 27926 class="clickable no-decoration" 27927 id="vhsjs_hide_624_1707793552_6461422" 27928 onclick="$('#623_1707793552_6461344').hide(function() { 27929 if (typeof Masonry === 'function') { 27930 $('.use_masonry').masonry(); 27931 }; 27932 }); 27933 $('#vhsjs_hide_624_1707793552_6461422').hide(); 27934 $('#vhsjs_view_624_1707793552_6461422').show();" 27935 style="display: none" 27936 ><i class="fa fa-caret-down"></i> 27937 <span class="hover_link">Abstract</span></a 27938 > 27939 <div 27940 data-display-control="624_1707793552_6461422" 27941 id="623_1707793552_6461344" 27942 style="display: none" 27943 > 27944 <div class="arrow-slidedown"> 27945 <blockquote> 27946 Owing to the increase in computing power, 27947 large-scale agent-based modeling (ABM) has been 27948 increasingly used in various fields. However, a 27949 complete and detailed individual population is 27950 challenging to obtain because of confidentiality 27951 concerns. Thus, modelers must adopt population 27952 synthesis to emulate the joint distribution of 27953 individual-level attributes of the actual 27954 population in the region of interest. 27955 Traditional population synthesis methods often 27956 exhibit issues regarding scalability and 27957 sampling zero. Therefore, this paper presents 27958 the use of a deep generative model called the 27959 denoising diffusion probabilistic model to 27960 generate new samples. Our proposed method uses 27961 the characteristics of deep generative model of 27962 generation from noise to generate a synthetic 27963 population, including sampling zero. In the 27964 experimental results, the standardized root mean 27965 squared error of our proposed model performed 27966 2.130, which outperformed 2.381 of the deep 27967 learning-based population synthesis method, VAE, 27968 and 7.620 of the traditional population 27969 synthesis method, MCMC. 27970 </blockquote> 27971 </div> 27972 </div> 27973 </div> 27974 </div> 27975 <div class="slot-urls"></div> 27976 <a href="/wsc23papers/247.pdf" target="_blank">pdf</a 27977 ><br /> 27978 </div> 27979 <div class="slot-entry"> 27980 <a name="con192" tabindex="-1"></a> 27981 <div class="slot-title-line">
27982 <span class="slot-title" 27983 >Quantum Embedding Framework of Industrial Data for 27984 Quantum Deep Learning</span 27985 > 27986 </div> 27987 <div class="slot-authors"> 27988 Hyunsoo Lee (Kumoh National Institute of Technology) and 27989 Amarnath Banerjee (Texas A&M University) 27990 </div> 27991 <div class="slot-abstract"> 27992 <div> 27993 <a 27994 class="clickable no-decoration" 27995 id="vhsjs_view_626_1707793552_6482742" 27996 onclick="$('#vhsjs_view_626_1707793552_6482742').hide(); 27997 $('#vhsjs_hide_626_1707793552_6482742').show(); 27998 $('#625_1707793552_6482658').slideDown(function() { 27999 if (typeof Masonry === 'function') { 28000 $('.use_masonry').masonry(); 28001 }; 28002 28003 });" 28004 ><i class="fa fa-caret-right"></i> 28005 <span class="hover_link">Abstract</span></a 28006 ><a 28007 class="clickable no-decoration" 28008 id="vhsjs_hide_626_1707793552_6482742" 28009 onclick="$('#625_1707793552_6482658').hide(function() { 28010 if (typeof Masonry === 'function') { 28011 $('.use_masonry').masonry(); 28012 }; 28013 }); 28014 $('#vhsjs_hide_626_1707793552_6482742').hide(); 28015 $('#vhsjs_view_626_1707793552_6482742').show();" 28016 style="display: none" 28017 ><i class="fa fa-caret-down"></i> 28018 <span class="hover_link">Abstract</span></a 28019 > 28020 <div 28021 data-display-control="626_1707793552_6482742" 28022 id="625_1707793552_6482658" 28023 style="display: none" 28024 > 28025 <div class="arrow-slidedown"> 28026 <blockquote> 28027 Quantum computing is a contemporary engineering 28028 discipline that innovatively overcomes 28029 computational burdens. This study applies 28030 quantum computing techniques to data analyses 28031 with input data issues. When a dataset has 28032 insufficient attributes and uncertainties, 28033 quantum embedding techniques contribute to the 28034 dimensional expansion of input vectors and the 28035 quantification of uncertainties. The converted 28036 qubits are linked to subsequent deep learning 28037 modules, and this architecture is used for 28038 accurate data analysis. This study proposes a 28039 quantum embedding technique and a corresponding 28040 quantum neural network (QNN) to better 28041 understand these processes. In this QNN 28042 architecture, input data are converted into 28043 corresponding qubits, which are transformed with 28044 quantum phase-operating modules. The quantum 28045 features pass through subsequent deep learning 28046 layers for more accurate data analyses. To 28047 demonstrate the effectiveness of the proposed 28048 model, a process model and relevant analyses are 28049 presented and compared with existing deep 28050 learning methods. 28051 </blockquote> 28052 </div> 28053 </div> 28054 </div> 28055 </div> 28056 <div class="slot-urls"></div> 28057 <a href="/wsc23papers/248.pdf" target="_blank">pdf</a 28058 ><br /> 28059 </div> 28060 <div class="slot-entry"> 28061 <a name="con299" tabindex="-1"></a> 28062 <div class="slot-title-line">
28063 <span class="slot-title" 28064 >Simulation of a Novel, Low Swap, Sparse 28065 Hyper-Dimensional Neural Network Architecture for 28066 Anomaly Detection AI at the Edge</span 28067 > 28068 </div> 28069 <div class="slot-authors"> 28070 Dean C. Mumme (RAM Laboratories, Inc.) and Ksenia Burova 28071 (RAM Laboratories, Inc) 28072 </div> 28073 <div class="slot-abstract"> 28074 <div> 28075 <a 28076 class="clickable no-decoration" 28077 id="vhsjs_view_628_1707793552_650467" 28078 onclick="$('#vhsjs_view_628_1707793552_650467').hide(); 28079 $('#vhsjs_hide_628_1707793552_650467').show(); 28080 $('#627_1707793552_6504583').slideDown(function() { 28081 if (typeof Masonry === 'function') { 28082 $('.use_masonry').masonry(); 28083 }; 28084 28085 });" 28086 ><i class="fa fa-caret-right"></i> 28087 <span class="hover_link">Abstract</span></a 28088 ><a 28089 class="clickable no-decoration" 28090 id="vhsjs_hide_628_1707793552_650467" 28091 onclick="$('#627_1707793552_6504583').hide(function() { 28092 if (typeof Masonry === 'function') { 28093 $('.use_masonry').masonry(); 28094 }; 28095 }); 28096 $('#vhsjs_hide_628_1707793552_650467').hide(); 28097 $('#vhsjs_view_628_1707793552_650467').show();" 28098 style="display: none" 28099 ><i class="fa fa-caret-down"></i> 28100 <span class="hover_link">Abstract</span></a 28101 > 28102 <div 28103 data-display-control="628_1707793552_650467" 28104 id="627_1707793552_6504583" 28105 style="display: none" 28106 > 28107 <div class="arrow-slidedown"> 28108 <blockquote> 28109 This paper details the simulation and 28110 performance results of a Sparse 28111 Hyper-Distributed Robust Efficient Neural 28112 Network (SpHyRE-Net) architecture that performs 28113 anomaly detection for real-world time-series 28114 data. SpHyRE-Net is an innovative, novel, low 28115 size, weight and power (SWaP) machine learning 28116 solution for devices operating at the tactical 28117 edge. It utilizes bit operations and sparse 28118 hyper-dimensional representations for 28119 bio-inspired learning via a Hebbian-like rule 28120 that results in a combined power-latency 28121 reduction of 2-orders of magnitude over ordinary 28122 deep networks. The paper details the application 28123 of SpHyRE-Net to real-world cell-traffic 28124 datasets as well as simulation requirements to 28125 minimize latency and memory use. Also discussed 28126 are the mechanisms necessary for implementing 28127 the architecture on an FPGA as a precursor to 28128 realization on a neuro-morphic ASIC with 28129 ultra-low power profile. 28130 </blockquote> 28131 </div> 28132 </div> 28133 </div> 28134 </div> 28135 <div class="slot-urls"></div> 28136 <a href="/wsc23papers/249.pdf" target="_blank">pdf</a 28137 ><br /> 28138 </div> 28139 </div> 28140 <div class="session-entry">
28141 <span class="session-event-type">Technical Session</span 28142 ><span class="type-track-spacer"> · </span 28143 ><span class="program-track" 28144 >Simulation and Artificial Intelligence</span 28145 ><br /> 28146 <div class="session-title"> 28147 Applications in Energy, Climate, and Finance 28148 </div> 28149 <div class="session-chair"> 28150 Chair: Dean Mumme (RAM Laboratories, Inc.)<br /> 28151 </div> 28152 <div class="slot-entry"> 28153 <a name="con244" tabindex="-1"></a> 28154 <div class="slot-title-line"> 28155 <span class="slot-title" 28156 >A Conversational Human-Computer Interface for Smart 28157 Energy System Simulation Environments</span 28158 > 28159 </div> 28160 <div class="slot-authors"> 28161 Gabriel Dengler (FAU Erlangen-Nuremberg, Laboratory of 28162 Computer Networks and Communication Systems); Pooia 28163 Lalbakhsh (Monash University); Peter Bazan (FAU 28164 Erlangen-Nuremberg, Laboratory of Computer Networks and 28165 Communication Systems); Ariel Liebmann (Monash 28166 University); and Reinhard German (FAU 28167 Erlangen-Nuremberg, Laboratory of Computer Networks and 28168 Communication Systems) 28169 </div> 28170 <div class="slot-abstract"> 28171 <div> 28172 <a 28173 class="clickable no-decoration" 28174 id="vhsjs_view_630_1707793552_6554554" 28175 onclick="$('#vhsjs_view_630_1707793552_6554554').hide(); 28176 $('#vhsjs_hide_630_1707793552_6554554').show(); 28177 $('#629_1707793552_6554465').slideDown(function() { 28178 if (typeof Masonry === 'function') { 28179 $('.use_masonry').masonry(); 28180 }; 28181 28182 });" 28183 ><i class="fa fa-caret-right"></i> 28184 <span class="hover_link">Abstract</span></a 28185 ><a 28186 class="clickable no-decoration" 28187 id="vhsjs_hide_630_1707793552_6554554" 28188 onclick="$('#629_1707793552_6554465').hide(function() { 28189 if (typeof Masonry === 'function') { 28190 $('.use_masonry').masonry(); 28191 }; 28192 }); 28193 $('#vhsjs_hide_630_1707793552_6554554').hide(); 28194 $('#vhsjs_view_630_1707793552_6554554').show();" 28195 style="display: none" 28196 ><i class="fa fa-caret-down"></i> 28197 <span class="hover_link">Abstract</span></a 28198 > 28199 <div 28200 data-display-control="630_1707793552_6554554" 28201 id="629_1707793552_6554465" 28202 style="display: none" 28203 > 28204 <div class="arrow-slidedown"> 28205 <blockquote> 28206 This paper introduces a conversational framework 28207 that enhances the usability of smart energy 28208 system simulations. This study is centered 28209 around OpenAI's Generative Pre-trained 28210 Transformer (GPT), a fine-tuned conversational 28211 model that allows users to communicate with the 28212 system in a natural way. Therefore, users can 28213 describe their simulation scenarios in plain 28214 language and GPT seamlessly translates these 28215 descriptions into Python scripts, used as inputs 28216 to the simulation environment, in our
28216case, 28217 AnyLogic Simulation Software. Our framework is 28218 based on the i7-AnyEnergy core framework to 28219 compute distribution flows and relevant 28220 statistics. The proposed human-machine interface 28221 facilitates and accelerates simulation modeling, 28222 as demonstrated through the two scenarios we 28223 have provided in this paper. Overall, our 28224 conversational framework has the potential to 28225 significantly improve the user experience of 28226 smart energy system simulation environments. By 28227 simplifying the interaction between users and 28228 complex simulation models, we enable users to 28229 obtain valuable insights rapidly and more 28230 easily. 28231 </blockquote> 28232 </div> 28233 </div> 28234 </div> 28235 </div> 28236 <div class="slot-urls"></div> 28237 <a href="/wsc23papers/250.pdf" target="_blank">pdf</a 28238 ><br /> 28239 </div> 28240 <div class="slot-entry"> 28241 <a name="con272" tabindex="-1"></a> 28242 <div class="slot-title-line"> 28243 <span class="slot-title" 28244 >A Machine Learning Framework to Explain Complex 28245 Geospatial Simulations: A Climate Change Case 28246 Study</span 28247 > 28248 </div> 28249 <div class="slot-authors"> 28250 Tanvir Ferdousi (University of Virginia); Mingliang Liu, 28251 Kirti Rajagopalan, and Jennifer Adam (Washington State 28252 University); and Abhijin Adiga, Mandy Wilson, S. S. 28253 Ravi, Anil Vullikanti, Madhav Marathe, and Samarth 28254 Swarup (University of Virginia) 28255 </div> 28256 <div class="slot-abstract"> 28257 <div> 28258 <a 28259 class="clickable no-decoration" 28260 id="vhsjs_view_632_1707793552_6580503" 28261 onclick="$('#vhsjs_view_632_1707793552_6580503').hide(); 28262 $('#vhsjs_hide_632_1707793552_6580503').show(); 28263 $('#631_1707793552_6580422').slideDown(function() { 28264 if (typeof Masonry === 'function') { 28265 $('.use_masonry').masonry(); 28266 }; 28267 28268 });" 28269 ><i class="fa fa-caret-right"></i> 28270 <span class="hover_link">Abstract</span></a 28271 ><a 28272 class="clickable no-decoration" 28273 id="vhsjs_hide_632_1707793552_6580503" 28274 onclick="$('#631_1707793552_6580422').hide(function() { 28275 if (typeof Masonry === 'function') { 28276 $('.use_masonry').masonry(); 28277 }; 28278 }); 28279 $('#vhsjs_hide_632_1707793552_6580503').hide(); 28280 $('#vhsjs_view_632_1707793552_6580503').show();" 28281 style="display: none" 28282 ><i class="fa fa-caret-down"></i> 28283 <span class="hover_link">Abstract</span></a 28284 > 28285 <div 28286 data-display-control="632_1707793552_6580503" 28287 id="631_1707793552_6580422" 28288 style="display: none" 28289 > 28290 <div class="arrow-slidedown"> 28291 <blockquote> 28292 The explainability of large and complex 28293 simulation models is an open problem. We present 28294 a framework to analyze such models by processing 28295 multidimensional data through a pipeline of 28296 target variable computation, clustering, 28297 supervised classification, and feature 28298 importance analysis. As a use case, the 28299 well-known large-scale hydrology and crop 28300 systems simulator VIC-CropSyst is utilized to 28301 evaluate how climate change may affect water 28302 availability in Washington, United States. We 28303 study how snowmelt varies with climate variables 28304 (temperature, precipitation) to identify 28305 different response characteristics. Based on 28306 these characteristics, spatial units are 28307 clustered into six distinct classes. A random 28308 forest classifier is used with Shapley values to 28309 rank static soil and land parameters that help 28310 detect each class. The results also include an 28311 analysis of risk across different classes to 28312 identify areas vulnerable to climate change. 28313 This paper demonstrates the usefulness of the 28314 proposed framework in providing explainability 28315 for large and complex simulations. 28316 </blockquote> 28317 </div> 28318 </div> 28319 </div> 28320 </div> 28321 <div class="slot-urls"></div> 28322 <a href="/wsc23papers/251.pdf" target="_blank">pdf</a 28323 ><br /> 28324 </div> 28325 <div class="slot-entry"> 28326 <a name="con377" tabindex="-1"></a> 28327 <div class="slot-title-line">
28328 <span class="slot-title" 28329 >Cutting through the Noise: Machine Learning Proxies 28330 for High Dimensional Nested Simulation</span 28331 > 28332 </div> 28333 <div class="slot-authors"> 28334 Xintong Li, Ben Mingbin Feng, and Tony Wirjanto 28335 (University of Waterloo) 28336 </div> 28337 <div class="slot-abstract"> 28338 <div> 28339 <a 28340 class="clickable no-decoration" 28341 id="vhsjs_view_634_1707793552_6604586" 28342 onclick="$('#vhsjs_view_634_1707793552_6604586').hide(); 28343 $('#vhsjs_hide_634_1707793552_6604586').show(); 28344 $('#633_1707793552_6604502').slideDown(function() { 28345 if (typeof Masonry === 'function') { 28346 $('.use_masonry').masonry(); 28347 }; 28348 28349 });" 28350 ><i class="fa fa-caret-right"></i> 28351 <span class="hover_link">Abstract</span></a 28352 ><a 28353 class="clickable no-decoration" 28354 id="vhsjs_hide_634_1707793552_6604586" 28355 onclick="$('#633_1707793552_6604502').hide(function() { 28356 if (typeof Masonry === 'function') { 28357 $('.use_masonry').masonry(); 28358 }; 28359 }); 28360 $('#vhsjs_hide_634_1707793552_6604586').hide(); 28361 $('#vhsjs_view_634_1707793552_6604586').show();" 28362 style="display: none" 28363 ><i class="fa fa-caret-down"></i> 28364 <span class="hover_link">Abstract</span></a 28365 > 28366 <div 28367 data-display-control="634_1707793552_6604586" 28368 id="633_1707793552_6604502" 28369 style="display: none" 28370 > 28371 <div class="arrow-slidedown"> 28372 <blockquote> 28373 Deep learning models have gained great success 28374 in many applications, but their adoption in 28375 financial and actuarial applications have been 28376 received by regulators with some treprdation. 28377 The lack of transparency and interpretability of 28378 these models leads to skepticism about their 28379 resilience and reliability, which are important 28380 factors to ensure financial stability and 28381 insurance benefit fulfillment. In this study, we 28382 use stochastic simulation as a data generator to 28383 examine deep learning models under controlled 28384 settings. Our study shows interesting findings 28385 in fundamental questions like "What do deep 28386 learning models learn from noisy data?'' and 28387 "How well do they learn from noisy data?''. 28388 Based on our findings, we propose an efficient 28389 nested simulation procedure that uses deep 28390 learning models as proxies to estimate tail risk 28391 measures of hedging errors for variable 28392 annuities. The proposed procedure uses deep 28393 learning models to concentrate simulation budget 28394 on tail scenarios while maintaining transparency 28395 in the estimation. 28396 </blockquote> 28397 </div> 28398 </div> 28399 </div> 28400 </div> 28401 <div class="slot-urls"></div> 28402 <a href="/wsc23papers/252.pdf" target="_blank">pdf</a 28403 ><br /> 28404 </div> 28405 </div> 28406 <div class="session-entry">
28407 <span class="session-event-type">Technical Session</span 28408 ><span class="type-track-spacer"> · </span 28409 ><span class="program-track" 28410 >Simulation and Artificial Intelligence</span 28411 ><br /> 28412 <div class="session-title">Reinforcement Learning</div> 28413 <div class="session-chair"> 28414 Chair: Gabriel Dengler (FAU Erlangen-Nuremberg, Laboratory 28415 of Computer Networks and Communication Systems)<br /> 28416 </div> 28417 <div class="slot-entry"> 28418 <a name="con325" tabindex="-1"></a> 28419 <div class="slot-title-line"> 28420 <span class="slot-title" 28421 >Reinforcement Learning with an Abrupt Model 28422 Change</span 28423 > 28424 </div> 28425 <div class="slot-authors"> 28426 Wuxia Chen and Taposh Banerjee (University of 28427 Pittsburgh) and Jemin George and Carl Busart (US Army 28428 Research Lab) 28429 </div> 28430 <div class="slot-abstract"> 28431 <div> 28432 <a 28433 class="clickable no-decoration" 28434 id="vhsjs_view_636_1707793552_6649823" 28435 onclick="$('#vhsjs_view_636_1707793552_6649823').hide(); 28436 $('#vhsjs_hide_636_1707793552_6649823').show(); 28437 $('#635_1707793552_6649745').slideDown(function() { 28438 if (typeof Masonry === 'function') { 28439 $('.use_masonry').masonry(); 28440 }; 28441 28442 });" 28443 ><i class="fa fa-caret-right"></i> 28444 <span class="hover_link">Abstract</span></a 28445 ><a 28446 class="clickable no-decoration" 28447 id="vhsjs_hide_636_1707793552_6649823" 28448 onclick="$('#635_1707793552_6649745').hide(function() { 28449 if (typeof Masonry === 'function') { 28450 $('.use_masonry').masonry(); 28451 }; 28452 }); 28453 $('#vhsjs_hide_636_1707793552_6649823').hide(); 28454 $('#vhsjs_view_636_1707793552_6649823').show();" 28455 style="display: none" 28456 ><i class="fa fa-caret-down"></i> 28457 <span class="hover_link">Abstract</span></a 28458 > 28459 <div 28460 data-display-control="636_1707793552_6649823" 28461 id="635_1707793552_6649745" 28462 style="display: none" 28463 > 28464 <div class="arrow-slidedown"> 28465 <blockquote> 28466 The problem of reinforcement learning is 28467 considered where the environment or the model 28468 undergoes a change. An algorithm is proposed 28469 that an agent can apply in such a problem to 28470 achieve the optimal long-time discounted reward. 28471 The algorithm is model-free and learns the 28472 optimal policy by interacting with the 28473 environment. It is shown that the proposed 28474 algorithm has strong optimality properties. The 28475 effectiveness of the algorithm is also 28476 demonstrated using simulation results. The 28477 proposed algorithm exploits a fundamental 28478 reward-detection trade-off present in these 28479 problems and uses an algorithm for the quickest 28480 detection of the model change. Recommendations 28481 are provided for faster detection of model 28482 changes and for smart initialization strategies. 28483 </blockquote> 28484 </div> 28485 </div> 28486 </div> 28487 </div> 28488 <div class="slot-urls"></div> 28489 <a href="/wsc23papers/253.pdf" target="_blank">pdf</a 28490 ><br /> 28491 </div> 28492 <div class="slot-entry"> 28493 <a name="con216" tabindex="-1"></a> 28494 <div class="slot-title-line">
28495 <span class="slot-title" 28496 >Dynamic Scheduling of Gantry Robots using Simulation 28497 and Reinforcement Learning</span 28498 > 28499 </div> 28500 <div class="slot-authors"> 28501 Horst Zisgen and Robert Miltenberger (Hochschule 28502 Darmstadt) and Markus Hochhaus and Niklas Stöhr 28503 (SimPlan AG) 28504 </div> 28505 <div class="slot-abstract"> 28506 <div> 28507 <a 28508 class="clickable no-decoration" 28509 id="vhsjs_view_638_1707793552_6673741" 28510 onclick="$('#vhsjs_view_638_1707793552_6673741').hide(); 28511 $('#vhsjs_hide_638_1707793552_6673741').show(); 28512 $('#637_1707793552_667366').slideDown(function() { 28513 if (typeof Masonry === 'function') { 28514 $('.use_masonry').masonry(); 28515 }; 28516 28517 });" 28518 ><i class="fa fa-caret-right"></i> 28519 <span class="hover_link">Abstract</span></a 28520 ><a 28521 class="clickable no-decoration" 28522 id="vhsjs_hide_638_1707793552_6673741" 28523 onclick="$('#637_1707793552_667366').hide(function() { 28524 if (typeof Masonry === 'function') { 28525 $('.use_masonry').masonry(); 28526 }; 28527 }); 28528 $('#vhsjs_hide_638_1707793552_6673741').hide(); 28529 $('#vhsjs_view_638_1707793552_6673741').show();" 28530 style="display: none" 28531 ><i class="fa fa-caret-down"></i> 28532 <span class="hover_link">Abstract</span></a 28533 > 28534 <div 28535 data-display-control="638_1707793552_6673741" 28536 id="637_1707793552_667366" 28537 style="display: none" 28538 > 28539 <div class="arrow-slidedown"> 28540 <blockquote> 28541 Industry 4.0 induces an increasing demand of 28542 autonomous interaction between the units of 28543 production facilities, like work centers and 28544 transportation equipment. This has an impact on 28545 the requirements for production scheduling and 28546 control algorithms. These must be capable to 28547 adapt autonomously to changes on the shop floor. 28548 This paper presents a combination of 28549 Reinforcement Learning and discrete event 28550 simulation for controlling a flexible flow shop 28551 using a gantry robot system as transportation 28552 unit. In a gantry robot system parts are 28553 transported by carriages fitted with grippers 28554 that travel along rails from machine to machine. 28555 The presented agent learns autonomously the 28556 right control policy to move the carriages. It 28557 is shown that in cases the optimal policy can be 28558 determined the Reinforcement Learning based 28559 policy is optimal and in other cases the 28560 achieved throughput does slightly exceed the 28561 throughput gained by a heuristic priority rule 28562 for controlling the gantry robot. 28563 </blockquote> 28564 </div> 28565 </div> 28566 </div> 28567 </div> 28568 <div class="slot-urls"></div> 28569 <a href="/wsc23papers/254.pdf" target="_blank">pdf</a 28570 ><br /> 28571 </div> 28572 <div class="slot-entry"> 28573 <a name="con117" tabindex="-1"></a> 28574 <div class="slot-title-line">
28575 <span class="slot-title" 28576 >Learning Environment for the Air Domain (LEAD)</span 28577 > 28578 </div> 28579 <div class="slot-authors"> 28580 Andreas Strand, Patrick R Gorton, Martin Asprusten, and 28581 Karsten Brathen (FFI) 28582 </div> 28583 <div class="slot-abstract"> 28584 <div> 28585 <a 28586 class="clickable no-decoration" 28587 id="vhsjs_view_640_1707793552_6697035" 28588 onclick="$('#vhsjs_view_640_1707793552_6697035').hide(); 28589 $('#vhsjs_hide_640_1707793552_6697035').show(); 28590 $('#639_1707793552_6696954').slideDown(function() { 28591 if (typeof Masonry === 'function') { 28592 $('.use_masonry').masonry(); 28593 }; 28594 28595 });" 28596 ><i class="fa fa-caret-right"></i> 28597 <span class="hover_link">Abstract</span></a 28598 ><a 28599 class="clickable no-decoration" 28600 id="vhsjs_hide_640_1707793552_6697035" 28601 onclick="$('#639_1707793552_6696954').hide(function() { 28602 if (typeof Masonry === 'function') { 28603 $('.use_masonry').masonry(); 28604 }; 28605 }); 28606 $('#vhsjs_hide_640_1707793552_6697035').hide(); 28607 $('#vhsjs_view_640_1707793552_6697035').show();" 28608 style="display: none" 28609 ><i class="fa fa-caret-down"></i> 28610 <span class="hover_link">Abstract</span></a 28611 > 28612 <div 28613 data-display-control="640_1707793552_6697035" 28614 id="639_1707793552_6696954" 28615 style="display: none" 28616 > 28617 <div class="arrow-slidedown"> 28618 <blockquote> 28619 A substantial part of fighter pilot training is 28620 simulation-based and involves computer-generated 28621 forces controlled by predefined behavior models. 28622 The behavior models are typically manually 28623 created by eliciting knowledge from experienced 28624 pilots, which is a time-consuming process. 28625 Despite the work put in, the behavior models are 28626 often unsatisfactory due to their predictable 28627 nature and lack of adaptivity, forcing 28628 instructors to spend time manually monitoring 28629 and controlling them. Reinforcement and 28630 imitation learning pose as alternatives to 28631 handcrafted models. This paper presents the 28632 Learning Environment for the Air Domain (LEAD), 28633 a system for creating and integrating 28634 intelligent air combat behavior in military 28635 simulations. By incorporating the popular 28636 programming library and interface Gymnasium, 28637 LEAD allows users to apply readily available 28638 machine learning algorithms. Additionally, LEAD 28639 can communicate with third-party simulation 28640 software through distributed simulation 28641 protocols, which allows behavior models to be 28642 learned and employed using simulation systems of 28643 different fidelities. 28644 </blockquote> 28645 </div> 28646 </div> 28647 </div> 28648 </div> 28649 <div class="slot-urls"></div> 28650 <a href="/wsc23papers/255.pdf" target="_blank">pdf</a 28651 ><br /> 28652 </div> 28653 </div> 28654 <div class="session-entry">
28655 <span class="session-event-type">Technical Session</span 28656 ><span class="type-track-spacer"> · </span 28657 ><span class="program-track" 28658 >Simulation and Artificial Intelligence</span 28659 ><br /> 28660 <div class="session-title"> 28661 Artificial Intelligence in Manufacturing Applications 28662 </div> 28663 <div class="session-chair"> 28664 Chair: Andreas Strand (FFI)<br /> 28665 </div> 28666 <div class="slot-entry"> 28667 <a name="con137" tabindex="-1"></a> 28668 <div class="slot-title-line"> 28669 <span class="slot-title" 28670 >Dispatching in Real Frontend Fabs with Industrial 28671 Grade Discrete-Event Simulations by Deep Reinforcement 28672 Learning with Evolution Strategies</span 28673 > 28674 </div> 28675 <div class="slot-authors"> 28676 Patrick Stöckermann, Alessandro Immordino, and 28677 Thomas Altenmüller (Infineon Technologies AG); 28678 Georg Seidel (Infineon Technologies Austria); Martin 28679 Gebser and Pierre Tassel (University of Klagenfurt); and 28680 Chew Wye Chan and Feifei Zhang (D-SIMLAB Technologies 28681 Pte Ltd) 28682 </div> 28683 <div class="slot-abstract"> 28684 <div> 28685 <a 28686 class="clickable no-decoration" 28687 id="vhsjs_view_642_1707793552_6753645" 28688 onclick="$('#vhsjs_view_642_1707793552_6753645').hide(); 28689 $('#vhsjs_hide_642_1707793552_6753645').show(); 28690 $('#641_1707793552_6753564').slideDown(function() { 28691 if (typeof Masonry === 'function') { 28692 $('.use_masonry').masonry(); 28693 }; 28694 28695 });" 28696 ><i class="fa fa-caret-right"></i> 28697 <span class="hover_link">Abstract</span></a 28698 ><a 28699 class="clickable no-decoration" 28700 id="vhsjs_hide_642_1707793552_6753645" 28701 onclick="$('#641_1707793552_6753564').hide(function() { 28702 if (typeof Masonry === 'function') { 28703 $('.use_masonry').masonry(); 28704 }; 28705 }); 28706 $('#vhsjs_hide_642_1707793552_6753645').hide(); 28707 $('#vhsjs_view_642_1707793552_6753645').show();" 28708 style="display: none" 28709 ><i class="fa fa-caret-down"></i> 28710 <span class="hover_link">Abstract</span></a 28711 > 28712 <div 28713 data-display-control="642_1707793552_6753645" 28714 id="641_1707793552_6753564" 28715 style="display: none" 28716 > 28717 <div class="arrow-slidedown"> 28718 <blockquote> 28719 Scheduling is a fundamental task in each 28720 production facility with implications on the 28721 overall efficiency of the facility. While 28722 classic job-shop scheduling problems become 28723 intractable when the number of machines and jobs 28724 increase, the problem gets even more complex in 28725 the context of semiconductor manufacturing, 28726 where flexible production control and stochastic 28727 event handling are required. In this paper, we 28728 propose a Deep Reinforcement Learning approach 28729 for lot dispatching to minimize the Flow Factor 28730 (FF) of a digital twin of a real-world, 28731 stochastic, large-scale semiconductor 28732 manufacturing facility. We present the first 28733 application of Reinforcement Learning to an 28734 industrial grade semiconductor manufacturing 28735 scenario of that size. Our approach leverages
28736 self-attention mechanisms to learn an effective 28737 dispatching policy for the manufacturing 28738 facility and is able to reduce the global FF of 28739 the fab. 28740 </blockquote> 28741 </div> 28742 </div> 28743 </div> 28744 </div> 28745 <div class="slot-urls"></div> 28746 <a href="/wsc23papers/256.pdf" target="_blank">pdf</a 28747 ><br /> 28748 </div> 28749 <div class="slot-entry"> 28750 <a name="cea160" tabindex="-1"></a> 28751 <div class="slot-title-line"> 28752 <span class="slot-title" 28753 >Managing Bottlenecks in Systems with Product 28754 Recovery</span 28755 > 28756 </div> 28757 <div class="slot-authors"> 28758 Leila Talebi and Lin Guo (South Dakota School of Mines & 28759 Technology) 28760 </div> 28761 <div class="slot-abstract"> 28762 <div> 28763 <a 28764 class="clickable no-decoration" 28765 id="vhsjs_view_644_1707793552_6774805" 28766 onclick="$('#vhsjs_view_644_1707793552_6774805').hide(); 28767 $('#vhsjs_hide_644_1707793552_6774805').show(); 28768 $('#643_1707793552_6774724').slideDown(function() { 28769 if (typeof Masonry === 'function') { 28770 $('.use_masonry').masonry(); 28771 }; 28772 28773 });" 28774 ><i class="fa fa-caret-right"></i> 28775 <span class="hover_link">Abstract</span></a 28776 ><a 28777 class="clickable no-decoration" 28778 id="vhsjs_hide_644_1707793552_6774805" 28779 onclick="$('#643_1707793552_6774724').hide(function() { 28780 if (typeof Masonry === 'function') { 28781 $('.use_masonry').masonry(); 28782 }; 28783 }); 28784 $('#vhsjs_hide_644_1707793552_6774805').hide(); 28785 $('#vhsjs_view_644_1707793552_6774805').show();" 28786 style="display: none" 28787 ><i class="fa fa-caret-down"></i> 28788 <span class="hover_link">Abstract</span></a 28789 > 28790 <div 28791 data-display-control="644_1707793552_6774805" 28792 id="643_1707793552_6774724" 28793 style="display: none" 28794 > 28795 <div class="arrow-slidedown"> 28796 <blockquote>
28797 Effectively managing products at the end of 28798 their lifecycle is increasingly crucial as 28799 numerous systems adopt recovery strategies. 28800 However, many are limited to remanufacturing or 28801 recycling as the only recovery option. 28802 Effectively handling end-of-life products 28803 demands diverse approaches, including 28804 refurbishing and cannibalization. Sustainable 28805 recovery centers and manufacturers encounter 28806 challenges linked to uncertainties about the 28807 quantity and condition of returned products, 28808 which can disrupt operations and lead to 28809 bottlenecks. Our solution employs machine 28810 learning, specifically a CNN-LSTM model that 28811 combines Convolutional Neural Networks (CNN) and 28812 Long Short-Term Memory (LSTM), for predicting 28813 return product quantity and quality. 28814 Additionally, we utilize scenario-based 28815 simulations to proactively pre-identify and 28816 address bottlenecks within a short timeframe, 28817 especially within systems managing multiple 28818 recovery options or dealing with complex and 28819 hazardous materials. 28820 </blockquote> 28821 </div> 28822 </div> 28823 </div> 28824 </div> 28825 <div class="slot-urls"></div> 28826 <a href="/wsc23papers/cea160.pdf" target="_blank">pdf</a 28827 ><br /> 28828 </div> 28829 <div class="slot-entry"> 28830 <a name="cea105" tabindex="-1"></a> 28831 <div class="slot-title-line"> 28832 <span class="slot-title" 28833 >Simulation-Based Optimization for Enhanced CCS 28834 Schematic Arrangement Design</span 28835 > 28836 </div> 28837 <div class="slot-authors"> 28838 SookYoung Son (Seoul National University, HDKSOE); 28839 HyeonGoo Pyeon (HDKSOE); Jihee Kim (HDHHI); and Jong Hun 28840 Woo (Seoul National University, Research Institute of 28841 Marine Systems Engineering) 28842 </div> 28843 <div class="slot-abstract"> 28844 <div> 28845 <a 28846 class="clickable no-decoration" 28847 id="vhsjs_view_646_1707793552_67971" 28848 onclick="$('#vhsjs_view_646_1707793552_67971').hide(); 28849 $('#vhsjs_hide_646_1707793552_67971').show(); 28850 $('#645_1707793552_6797018').slideDown(function() { 28851 if (typeof Masonry === 'function') { 28852 $('.use_masonry').masonry(); 28853 }; 28854 28855 });" 28856 ><i class="fa fa-caret-right"></i> 28857 <span class="hover_link">Abstract</span></a 28858 ><a 28859 class="clickable no-decoration" 28860 id="vhsjs_hide_646_1707793552_67971" 28861 onclick="$('#645_1707793552_6797018').hide(function() { 28862 if (typeof Masonry === 'function') { 28863 $('.use_masonry').masonry(); 28864 }; 28865 }); 28866 $('#vhsjs_hide_646_1707793552_67971').hide(); 28867 $('#vhsjs_view_646_1707793552_67971').show();" 28868 style="display: none" 28869 ><i class="fa fa-caret-down"></i>
28870 <span class="hover_link">Abstract</span></a 28871 > 28872 <div 28873 data-display-control="646_1707793552_67971" 28874 id="645_1707793552_6797018" 28875 style="display: none" 28876 > 28877 <div class="arrow-slidedown"> 28878 <blockquote> 28879 An LNG cargo tank, referred to as the Cargo 28880 Containment System(CCS), encompasses several 28881 barriers intended for the storage of LNG at 28882 extremely low temperatures. In the case of the 28883 membrane-type CCS, each barrier is composed of 28884 insulation panels and membrane sheets. The CCS 28885 schematic arrangement endeavors to minimize the 28886 number of panels and sheets to enhance the 28887 manufacturing productivity. In this study, a 28888 combinatorial optimization approach is adopted 28889 to obtain the optimal CCS schematic arrangement. 28890 Then, a simulation environment is established to 28891 assess the arrangement results under diverse 28892 design conditions. By comparing the actual CCS 28893 design with the results of the proposed 28894 arrangement, the effectiveness of the proposed 28895 approach is valiated. 28896 </blockquote> 28897 </div> 28898 </div> 28899 </div> 28900 </div> 28901 <div class="slot-urls"></div> 28902 <a href="/wsc23papers/cea105.pdf" target="_blank">pdf</a 28903 ><br /> 28904 </div> 28905 </div> 28906 <div class="session-entry"> 28907 <span class="session-event-type">Technical Session</span 28908 ><span class="type-track-spacer"> · </span 28909 ><span class="program-track" 28910 >Simulation and Artificial Intelligence</span 28911 ><br /> 28912 <div class="session-title"> 28913 Artificial Intelligence and Optimization 28914 </div> 28915 <div class="session-chair"> 28916 Chair: Patrick Stöckermann (Infineon Technologies AG, 28917 University of Klagenfurt)<br /> 28918 </div> 28919 <div class="slot-entry"> 28920 <a name="con200" tabindex="-1"></a> 28921 <div class="slot-title-line"> 28922 <span class="slot-title" 28923 >Ensemble-Based Infill Search Simulation Optimization 28924 Framework</span 28925 > 28926 </div> 28927 <div class="slot-authors"> 28928 José Arnaldo Barra Montevechi, João Victor 28929 Soares do Amaral, Rafael de Carvalho Miranda, and Carlos 28930 Henrique dos Santos (Federal University of Itajubá) 28931 and Flávio de Oliveira Brito and Michael E. F. H. 28932 S. Machado (FlexSim Brazil, Inc.) 28933 </div> 28934 <div class="slot-abstract"> 28935 <div> 28936 <a 28937 class="clickable no-decoration" 28938 id="vhsjs_view_648_1707793552_6845188" 28939 onclick="$('#vhsjs_view_648_1707793552_6845188').hide(); 28940 $('#vhsjs_hide_648_1707793552_6845188').show(); 28941 $('#647_1707793552_684511').slideDown(function() { 28942 if (typeof Masonry === 'function') { 28943 $('.use_masonry').masonry(); 28944 }; 28945 28946 });" 28947 ><i class="fa fa-caret-right"></i>
28948 <span class="hover_link">Abstract</span></a 28949 ><a 28950 class="clickable no-decoration" 28951 id="vhsjs_hide_648_1707793552_6845188" 28952 onclick="$('#647_1707793552_684511').hide(function() { 28953 if (typeof Masonry === 'function') { 28954 $('.use_masonry').masonry(); 28955 }; 28956 }); 28957 $('#vhsjs_hide_648_1707793552_6845188').hide(); 28958 $('#vhsjs_view_648_1707793552_6845188').show();" 28959 style="display: none" 28960 ><i class="fa fa-caret-down"></i> 28961 <span class="hover_link">Abstract</span></a 28962 > 28963 <div 28964 data-display-control="648_1707793552_6845188" 28965 id="647_1707793552_684511" 28966 style="display: none" 28967 > 28968 <div class="arrow-slidedown"> 28969 <blockquote> 28970 Simulation is widely used in several areas of 28971 knowledge, from engineering to biology, 28972 including physics and finance. It allows the 28973 evaluation of the model’s results under 28974 different conditions, enabling performance 28975 analysis and more assertive decision-making. 28976 However, simulation can be computationally 28977 intensive, especially when we consider complex 28978 models. To deal with this problem, metamodeling 28979 has been increasingly used as a simulation 28980 optimization technique. In this article, we 28981 propose a new adaptive metamodeling method for 28982 simulation optimization, which aims to achieve 28983 better results using fewer experiments. This 28984 method combines machine learning and 28985 metaheuristic techniques, allowing the 28986 identification of the most important regions of 28987 the search space, which can be explored more 28988 efficiently to obtain optimal solutions. The 28989 results achieved in a manufacturing problem show 28990 that the proposed method presents a significant 28991 improvement in the achieved objective function 28992 value, in comparison with the conventional 28993 benchmark method, without compromising the 28994 simulation execution time. 28995 </blockquote> 28996 </div> 28997 </div> 28998 </div> 28999 </div> 29000 <div class="slot-urls"></div> 29001 <a href="/wsc23papers/257.pdf" target="_blank">pdf</a 29002 ><br /> 29003 </div> 29004 <div class="slot-entry"> 29005 <a name="con380" tabindex="-1"></a> 29006 <div class="slot-title-line"> 29007 <span class="slot-title" 29008 >Reusing Historical Observations in Natural Policy 29009 Gradient</span 29010 > 29011 </div> 29012 <div class="slot-authors"> 29013 Yifan Lin and Enlu Zhou (Georgia Institute of 29014 Technology) 29015 </div> 29016 <div class="slot-abstract"> 29017 <div> 29018 <a 29019 class="clickable no-decoration" 29020 id="vhsjs_view_650_1707793552_686849" 29021 onclick="$('#vhsjs_view_650_1707793552_686849').hide(); 29022 $('#vhsjs_hide_650_1707793552_686849').show(); 29023 $('#649_1707793552_686841').slideDown(function() { 29024 if (typeof Masonry === 'function') { 29025 $('.use_masonry').masonry(); 29026 }; 29027 29028 });" 29029 ><i class="fa fa-caret-right"></i>
29030 <span class="hover_link">Abstract</span></a 29031 ><a 29032 class="clickable no-decoration" 29033 id="vhsjs_hide_650_1707793552_686849" 29034 onclick="$('#649_1707793552_686841').hide(function() { 29035 if (typeof Masonry === 'function') { 29036 $('.use_masonry').masonry(); 29037 }; 29038 }); 29039 $('#vhsjs_hide_650_1707793552_686849').hide(); 29040 $('#vhsjs_view_650_1707793552_686849').show();" 29041 style="display: none" 29042 ><i class="fa fa-caret-down"></i> 29043 <span class="hover_link">Abstract</span></a 29044 > 29045 <div 29046 data-display-control="650_1707793552_686849" 29047 id="649_1707793552_686841" 29048 style="display: none" 29049 > 29050 <div class="arrow-slidedown"> 29051 <blockquote> 29052 Reinforcement learning provides a mathematical 29053 framework for learning-based control, whose 29054 success largely depends on the amount of data it 29055 can utilize. The efficient utilization of 29056 historical samples obtained from previous 29057 iterations is essential for expediting policy 29058 optimization. Empirical evidence has shown that 29059 offline variants of policy gradient methods 29060 based on importance sampling work well. However, 29061 existing literature often neglect the 29062 interdependence between observations from 29063 different iterations, and the good empirical 29064 performance lacks a rigorous theoretical 29065 justification. In this paper, we study an 29066 offline variant of the natural policy gradient 29067 method with reusing historical observations. We 29068 show that the biases of the proposed estimators 29069 of Fisher information matrix and gradient are 29070 asymptotically negligible and reduce the 29071 conditional variance of the gradient estimator. 29072 The proposed algorithm and convergence analysis 29073 could be further applied to popular policy 29074 optimization algorithms such as trust region 29075 policy optimization. Our theoretical results are 29076 verified on classical benchmarks. 29077 </blockquote> 29078 </div> 29079 </div> 29080 </div> 29081 </div> 29082 <div class="slot-urls"></div> 29083 <a href="/wsc23papers/258.pdf" target="_blank">pdf</a 29084 ><br /> 29085 </div> 29086 </div> 29087 </div> 29088 <div class="centered"> 29089 <div class="top-link"><a href="#top">Return to Top</a></div> 29090 </div> 29091 <hr /> 29092 </div> 29093 <div class="area-section"> 29094 <div class="centered"> 29095 <a name="ptrack120" tabindex="-1"></a> 29096 <div class="section-title">Simulation as Digital Twin</div> 29097 </div> 29098 <div class="centered track-chair"> 29099 <span class="track-chair-role" 29100 >Track Coordinator - Simulation as Digital Twin: </span 29101 ><span class="track-chair-names"
29102 >Andrea Matta (Via La Masa 1, Politecnico di Milano), Jie Xu 29103 (George Mason University)</span 29104 > 29105 </div> 29106 <div class="section-entry"> 29107 <div class="session-entry"> 29108 <span class="session-event-type">Technical Session</span 29109 ><span class="type-track-spacer"> · </span 29110 ><span class="program-track" 29111 >Simulation as Digital Twin</span 29112 ><br /> 29113 <div class="session-title"> 29114 Human Systems and Digital Twins 29115 </div> 29116 <div class="session-chair"> 29117 Chair: Jie Xu (George Mason University)<br /> 29118 </div> 29119 <div class="slot-entry"> 29120 <a name="inv127" tabindex="-1"></a> 29121 <div class="slot-title-line"> 29122 <span class="slot-title" 29123 >Leveraging Digital Twins to Support a Sustained Human 29124 Presence on the Lunar Surface</span 29125 > 29126 </div> 29127 <div class="slot-authors"> 29128 Edward Hua and Linda Boan (The MITRE Corporation) 29129 </div> 29130 <div class="slot-abstract"> 29131 <div> 29132 <a 29133 class="clickable no-decoration" 29134 id="vhsjs_view_652_1707793552_6955085" 29135 onclick="$('#vhsjs_view_652_1707793552_6955085').hide(); 29136 $('#vhsjs_hide_652_1707793552_6955085').show(); 29137 $('#651_1707793552_6955001').slideDown(function() { 29138 if (typeof Masonry === 'function') { 29139 $('.use_masonry').masonry(); 29140 }; 29141 29142 });" 29143 ><i class="fa fa-caret-right"></i> 29144 <span class="hover_link">Abstract</span></a 29145 ><a 29146 class="clickable no-decoration" 29147 id="vhsjs_hide_652_1707793552_6955085" 29148 onclick="$('#651_1707793552_6955001').hide(function() { 29149 if (typeof Masonry === 'function') { 29150 $('.use_masonry').masonry(); 29151 }; 29152 }); 29153 $('#vhsjs_hide_652_1707793552_6955085').hide(); 29154 $('#vhsjs_view_652_1707793552_6955085').show();" 29155 style="display: none" 29156 ><i class="fa fa-caret-down"></i> 29157 <span class="hover_link">Abstract</span></a 29158 > 29159 <div 29160 data-display-control="652_1707793552_6955085" 29161 id="651_1707793552_6955001" 29162 style="display: none" 29163 > 29164 <div class="arrow-slidedown"> 29165 <blockquote> 29166 Having a sustained human presence on the lunar 29167 surface is a central objective of the Artemis 29168 Program, as it represents a key pre-requisite in 29169 resource mining operations on the Moon as well 29170 as an important steppingstone for future Martian 29171 exploration and colonization. Despite its 29172 importance, this endeavor has little precedent 29173 to rely on to inform the many challenges it 29174 needs to address. Digital Twin (DT), in recent 29175 years, has been employed in a wide range of 29176 applications. In this paper, we explore its 29177 usefulness in establishing the Artemis Base 29178 Camp. DT can be applied to various stages of the 29179 lifecycle of the lunar base development. We also 29180 identify several open questions that need be 29181 addressed before the digital twin can be 29182 utilized effectively in this project. In fact, 29183 addressing these questions could facilitate 29184 deploying DTs in use cases in a wider spectrum 29185 of industries and sectors. 29186 </blockquote> 29187 </div> 29188 </div> 29189 </div> 29190 </div> 29191 <div class="slot-urls"></div> 29192 <a href="/wsc23papers/268.pdf" target="_blank">pdf</a 29193 ><br /> 29194 </div> 29195 <div class="slot-entry"> 29196 <a name="con357" tabindex="-1"></a> 29197 <div class="slot-title-line">
29198 <span class="slot-title" 29199 >A General Framework for Human-in-the-loop Cognitive 29200 Digital Twins</span 29201 > 29202 </div> 29203 <div class="slot-authors"> 29204 Parisa Niloofar (University of Southern Denmark); Sanja 29205 Lazarova-Molnar (Institute AIFB, Karlsruhe Institute of 29206 Technology); Olufemi A. Omitaomu and Haowen Xu (Oak 29207 Ridge National Laboratory); and Xueping Li (University 29208 of Tennessee) 29209 </div> 29210 <div class="slot-abstract"> 29211 <div> 29212 <a 29213 class="clickable no-decoration" 29214 id="vhsjs_view_654_1707793552_6978097" 29215 onclick="$('#vhsjs_view_654_1707793552_6978097').hide(); 29216 $('#vhsjs_hide_654_1707793552_6978097').show(); 29217 $('#653_1707793552_6978016').slideDown(function() { 29218 if (typeof Masonry === 'function') { 29219 $('.use_masonry').masonry(); 29220 }; 29221 29222 });" 29223 ><i class="fa fa-caret-right"></i> 29224 <span class="hover_link">Abstract</span></a 29225 ><a 29226 class="clickable no-decoration" 29227 id="vhsjs_hide_654_1707793552_6978097" 29228 onclick="$('#653_1707793552_6978016').hide(function() { 29229 if (typeof Masonry === 'function') { 29230 $('.use_masonry').masonry(); 29231 }; 29232 }); 29233 $('#vhsjs_hide_654_1707793552_6978097').hide(); 29234 $('#vhsjs_view_654_1707793552_6978097').show();" 29235 style="display: none" 29236 ><i class="fa fa-caret-down"></i> 29237 <span class="hover_link">Abstract</span></a 29238 > 29239 <div 29240 data-display-control="654_1707793552_6978097" 29241 id="653_1707793552_6978016" 29242 style="display: none" 29243 > 29244 <div class="arrow-slidedown"> 29245 <blockquote> 29246 Modelling and analysis of systems that are 29247 equipped with sensors and connected to the 29248 Internet are becoming more automated and less 29249 human-dependent. However, bringing expert 29250 knowledge into the loop along with data obtained 29251 from Internet of Thing (IoT) devices minimizes 29252 the risk of making poor and unexplainable 29253 decisions and helps to assess the impact of 29254 different strategies before applying them in 29255 reality. While Digital Twins are more of a 29256 data-driven simulation of the physical system, 29257 Cognitive Digital Twins bring the human 29258 dimension into the modelling and simulation. In 29259 this paper, we aim to emphasize the crucial role 29260 of explainability and the underlying rationale 29261 behind automated or interactive decision-making 29262 processes. Furthermore, we propose an initial 29263 framework that delineates the specific points 29264 within the feedback loop of a cognitive digital 29265 twin where human involvement can be 29266 incorporated. 29267 </blockquote> 29268 </div> 29269 </div> 29270 </div> 29271 </div> 29272 <div class="slot-urls"></div> 29273 <a href="/wsc23papers/269.pdf" target="_blank">pdf</a 29274 ><br /> 29275 </div> 29276 <div class="slot-entry"> 29277 <a name="cea116" tabindex="-1"></a> 29278 <div class="slot-title-line">
29279 <span class="slot-title" 29280 >A Behavior Simulation-Based Approach to Improve 29281 Retail Performance: A Comprehensive Framework</span 29282 > 29283 </div> 29284 <div class="slot-authors"> 29285 Siddhartha Sarkar, Suman Kumar, and Vivek Balaraman 29286 (Tata Consultancy Services Ltd) 29287 </div> 29288 <div class="slot-abstract"> 29289 <div> 29290 <a 29291 class="clickable no-decoration" 29292 id="vhsjs_view_656_1707793552_7000113" 29293 onclick="$('#vhsjs_view_656_1707793552_7000113').hide(); 29294 $('#vhsjs_hide_656_1707793552_7000113').show(); 29295 $('#655_1707793552_7000034').slideDown(function() { 29296 if (typeof Masonry === 'function') { 29297 $('.use_masonry').masonry(); 29298 }; 29299 29300 });" 29301 ><i class="fa fa-caret-right"></i> 29302 <span class="hover_link">Abstract</span></a 29303 ><a 29304 class="clickable no-decoration" 29305 id="vhsjs_hide_656_1707793552_7000113" 29306 onclick="$('#655_1707793552_7000034').hide(function() { 29307 if (typeof Masonry === 'function') { 29308 $('.use_masonry').masonry(); 29309 }; 29310 }); 29311 $('#vhsjs_hide_656_1707793552_7000113').hide(); 29312 $('#vhsjs_view_656_1707793552_7000113').show();" 29313 style="display: none" 29314 ><i class="fa fa-caret-down"></i> 29315 <span class="hover_link">Abstract</span></a 29316 > 29317 <div 29318 data-display-control="656_1707793552_7000113" 29319 id="655_1707793552_7000034" 29320 style="display: none" 29321 > 29322 <div class="arrow-slidedown"> 29323 <blockquote> 29324 The retail industry is undergoing a profou
29324nd 29325 transformation, driven by technological 29326 advancements including AI and evolving consumer 29327 behaviors. However, what retail decision making 29328 lacks at present is knowledge of and integration 29329 of ways to factor in customer behavioral drivers 29330 in purchase decisions. We show how this can be 29331 done through a four-step approach that will 29332 create a behavior simulation model for retail 29333 use cases. We use a real world problem as a 29334 guiding example to explain our approach. Our 29335 approach enables retailers to use behavioral 29336 drivers to nudge customers and better 29337 explainability of the decisions. 29338 </blockquote> 29339 </div> 29340 </div> 29341 </div> 29342 </div> 29343 <div class="slot-urls"></div> 29344 <a href="/wsc23papers/cea116.pdf" target="_blank">pdf</a 29345 ><br /> 29346 </div> 29347 </div> 29348 <div class="session-entry"> 29349 <span class="session-event-type">Technical Session</span 29350 ><span class="type-track-spacer"> · </span 29351 ><span class="program-track" 29352 >Simulation as Digital Twin</span 29353 ><br /> 29354 <div class="session-title"> 29355 Applications of Digital Twins 29356 </div> 29357 <div class="session-chair"> 29358 Chair: Giovanni Lugaresi (CentraleSupelec, Politecnico di 29359 Milano)<br /> 29360 </div> 29361 <div class="slot-entry"> 29362 <a name="cea113" tabindex="-1"></a> 29363 <div class="slot-title-line"> 29364 <span class="slot-title" 29365 >Designing a Digital Twin Prototype for Improving 29366 Vaccination Centers' Daily Operations</span 29367 > 29368 </div> 29369 <div class="slot-authors"> 29370 Mohamed Ali Wafdi, Yasmina Maïzi, and Ygal Bendavid 29371 (ESG UQAM) 29372 </div> 29373 <div class="slot-abstract"> 29374 <div> 29375 <a 29376 class="clickable no-decoration" 29377 id="vhsjs_view_658_1707793552_704641" 29378 onclick="$('#vhsjs_view_658_1707793552_704641').hide(); 29379 $('#vhsjs_hide_658_1707793552_704641').show(); 29380 $('#657_1707793552_7046328').slideDown(function() { 29381 if (typeof Masonry === 'function') { 29382 $('.use_masonry').masonry(); 29383 }; 29384 29385 });" 29386 ><i class="fa fa-caret-right"></i> 29387 <span class="hover_link">Abstract</span></a 29388 ><a 29389 class="clickable no-decoration" 29390 id="vhsjs_hide_658_1707793552_704641" 29391 onclick="$('#657_1707793552_7046328').hide(function() { 29392 if (typeof Masonry === 'function') { 29393 $('.use_masonry').masonry(); 29394 }; 29395 }); 29396 $('#vhsjs_hide_658_1707793552_704641').hide(); 29397 $('#vhsjs_view_658_1707793552_704641').show();" 29398 style="display: none" 29399 ><i class="fa fa-caret-down"></i> 29400 <span class="hover_link">Abstract</span></a 29401 > 29402 <div
29403 data-display-control="658_1707793552_704641" 29404 id="657_1707793552_7046328" 29405 style="display: none" 29406 > 29407 <div class="arrow-slidedown"> 29408 <blockquote> 29409 In this research paper, we propose a digital 29410 twin prototype to improve mass vaccination 29411 centers in the Montreal region. This research is 29412 important because is it always challenging to 29413 define an optimal layout/capacity for healthcare 29414 operations, especially in an emergency mode 29415 (e.g., pandemic mode). Indeed, in such stressful 29416 situations, all managers are more concerned 29417 about the effectiveness of daily operations, 29418 regardless of their efficiency. Following a 29419 "design science" research approach, we developed 29420 (i) an IoT prototype for real-time patient 29421 tracking, (ii) a simulation model, and (iii) 29422 integrated them to build our digital twin 29423 prototype. Our institution's IoT lab was used as 29424 a testbed research environment for developing 29425 the IoT infrastructure and simulating the 29426 vaccination center. While the prototype was 29427 developed for vaccination centers, the approach 29428 can be used in any other multi-patient/multi 29429 flow operational environment where real-time 29430 visibility and simulation are required 29431 </blockquote> 29432 </div> 29433 </div> 29434 </div> 29435 </div> 29436 <div class="slot-urls"></div> 29437 <a href="/wsc23papers/cea113.pdf" target="_blank">pdf</a 29438 ><br /> 29439 </div> 29440 <div class="slot-entry"> 29441 <a name="cea143" tabindex="-1"></a> 29442 <div class="slot-title-line"> 29443 <span class="slot-title" 29444 >Utilizing Simulation to Evalute the Design of a 29445 Greenfield Multi-story Parking Structure and Impacts 29446 to Surrounding Areas</span 29447 > 29448 </div> 29449 <div class="slot-authors"> 29450 Lourdes Murphy (National Institutes of Health (NIH)) and 29451 Yusuke Legard (MOSIMTEC) 29452 </div> 29453 <div class="slot-abstract"> 29454 <div> 29455 <a 29456 class="clickable no-decoration" 29457 id="vhsjs_view_660_1707793552_7068043" 29458 onclick="$('#vhsjs_view_660_1707793552_7068043').hide(); 29459 $('#vhsjs_hide_660_1707793552_7068043').show(); 29460 $('#659_1707793552_7067962').slideDown(function() { 29461 if (typeof Masonry === 'function') { 29462 $('.use_masonry').masonry(); 29463 }; 29464 29465 });" 29466 ><i class="fa fa-caret-right"></i> 29467 <span class="hover_link">Abstract</span></a 29468 ><a 29469 class="clickable no-decoration" 29470 id="vhsjs_hide_660_1707793552_7068043" 29471 onclick="$('#659_1707793552_7067962').hide(function() { 29472 if (typeof Masonry === 'function') { 29473 $('.use_masonry').masonry(); 29474 }; 29475 }); 29476 $('#vhsjs_hide_660_1707793552_7068043').hide(); 29477 $('#vhsjs_view_660_1707793552_7068043').show();" 29478 style="display: none" 29479 ><i class="fa fa-caret-down"></i>
29480 <span class="hover_link">Abstract</span></a 29481 > 29482 <div 29483 data-display-control="660_1707793552_7068043" 29484 id="659_1707793552_7067962" 29485 style="display: none" 29486 > 29487 <div class="arrow-slidedown"> 29488 <blockquote> 29489 The National Institutes of Health (NIH) main 29490 campus in Bethesda, Maryland currently contains 29491 30 parking structures. On any given day, 12,000 29492 vehicles enter the campus. NIH is planning for 29493 the south side of the campus to become the main 29494 parking areas for employees and visitors. 29495 Central to this vision is replacing a surface 29496 lot, which contains 241 parking spaces, with the 29497 construction of a greenfield six story parking 29498 structure that has a planned capacity of 1420 29499 parking spaces. NIH wanted to prioritize the 29500 employee experience and emphasize the safety of 29501 pedestrians and vehicles. MOSIMTEC utilized 29502 simulation modeling to provide NIH with insight 29503 on the impact of various entrance and exit 29504 combinations into the parking structure. This 29505 presentation will further describe the project, 29506 the system being modeled, the inputs and outputs 29507 of the simulation tool and the outcome upon the 29508 design of the greenfield parking structure. 29509 </blockquote> 29510 </div> 29511 </div> 29512 </div> 29513 </div> 29514 <div class="slot-urls"></div> 29515 <a href="/wsc23papers/cea143.pdf" target="_blank">pdf</a 29516 ><br /> 29517 </div> 29518 <div class="slot-entry"> 29519 <a name="cea134" tabindex="-1"></a> 29520 <div class="slot-title-line"> 29521 <span class="slot-title" 29522 >Increasing Efficiency of Fresh Meal Production Using 29523 Simulation</span 29524 > 29525 </div> 29526 <div class="slot-authors"> 29527 Kean Dequeant and Daniel Paddon (Gousto) and Stephane 29528 Dauzère-Pérès and Claude Yugma (Mines 29529 Saint-Étienne, Univ Clermont Auvergne) 29530 </div> 29531 <div class="slot-abstract"> 29532 <div> 29533 <a 29534 class="clickable no-decoration" 29535 id="vhsjs_view_662_1707793552_7089128" 29536 onclick="$('#vhsjs_view_662_1707793552_7089128').hide(); 29537 $('#vhsjs_hide_662_1707793552_7089128').show(); 29538 $('#661_1707793552_7089047').slideDown(function() { 29539 if (typeof Masonry === 'function') { 29540 $('.use_masonry').masonry(); 29541 }; 29542 29543 });" 29544 ><i class="fa fa-caret-right"></i> 29545 <span class="hover_link">Abstract</span></a 29546 ><a 29547 class="clickable no-decoration" 29548 id="vhsjs_hide_662_1707793552_7089128" 29549 onclick="$('#661_1707793552_7089047').hide(function() { 29550 if (typeof Masonry === 'function') { 29551 $('.use_masonry').masonry(); 29552 }; 29553 }); 29554 $('#vhsjs_hide_662_1707793552_7089128').hide(); 29555 $('#vhsjs_view_662_1707793552_7089128').show();" 29556 style="display: none" 29557 ><i class="fa fa-caret-down"></i>
29558 <span class="hover_link">Abstract</span></a 29559 > 29560 <div 29561 data-display-control="662_1707793552_7089128" 29562 id="661_1707793552_7089047" 29563 style="display: none" 29564 > 29565 <div class="arrow-slidedown"> 29566 <blockquote> 29567 The pandemic period has witnessed a rapid growth 29568 of online delivery services in various sectors, 29569 especially in the domain of fresh produce 29570 e-commerce. Gousto, for instance, provides a 29571 meal subscription service where customers select 29572 their meals for a week, and subsequently receive 29573 a box containing all the required ingredients 29574 along with step-by-step cooking instructions for 29575 the chosen recipes. In light of recent economic 29576 difficulties worldwide, Gousto is prioritising 29577 its efficiency to reduce cost and to continue 29578 providing affordable meals to its customers. One 29579 key aspect for Gousto was to improve its station 29580 utilisation, through better routing of boxes 29581 throughout the factory. The use of simulation as 29582 a digital twin has been a key factor in the 29583 development of a new routing algorithm, that has 29584 now been put in production and has increased 29585 station utilisation by 20%, in line with the 29586 simulation's predictions. 29587 </blockquote> 29588 </div> 29589 </div> 29590 </div> 29591 </div> 29592 <div class="slot-urls"></div> 29593 <a href="/wsc23papers/cea134.pdf" target="_blank">pdf</a 29594 ><br /> 29595 </div> 29596 </div> 29597 <div class="session-entry"> 29598 <span class="session-event-type">Technical Session</span 29599 ><span class="type-track-spacer"> · </span 29600 ><span class="program-track" 29601 >Simulation as Digital Twin</span 29602 ><br /> 29603 <div class="session-title"> 29604 Digital Twins and Energy Systems 29605 </div> 29606 <div class="session-chair"> 29607 Chair: Sanja Lazarova-Molnar (Karlsruhe Institute of 29608 Technology, University of Southern Denmark)<br /> 29609 </div> 29610 <div class="slot-entry"> 29611 <a name="con291" tabindex="-1"></a> 29612 <div class="slot-title-line"> 29613 <span class="slot-title" 29614 >Modeling and Real-time Simulation of Microgrid 29615 Components using SystemC-AMS</span 29616 > 29617 </div> 29618 <div class="slot-authors"> 29619 Rahul Bhadani (Vanderbilt University, The University of 29620 Alabama in Huntsville); Hao Tu and Srdjan Lukic (North 29621 Carolina State University); and Gabor Karsai (Vanderbilt 29622 University) 29623 </div> 29624 <div class="slot-abstract"> 29625 <div> 29626 <a 29627 class="clickable no-decoration" 29628 id="vhsjs_view_664_1707793552_7137" 29629 onclick="$('#vhsjs_view_664_1707793552_7137').hide(); 29630 $('#vhsjs_hide_664_1707793552_7137').show(); 29631 $('#663_1707793552_713692').slideDown(function() { 29632 if (typeof Masonry === 'function') { 29633 $('.use_masonry').masonry(); 29634 }; 29635 29636 });" 29637 ><i class="fa fa-caret-right"></i>
29638 <span class="hover_link">Abstract</span></a 29639 ><a 29640 class="clickable no-decoration" 29641 id="vhsjs_hide_664_1707793552_7137" 29642 onclick="$('#663_1707793552_713692').hide(function() { 29643 if (typeof Masonry === 'function') { 29644 $('.use_masonry').masonry(); 29645 }; 29646 }); 29647 $('#vhsjs_hide_664_1707793552_7137').hide(); 29648 $('#vhsjs_view_664_1707793552_7137').show();" 29649 style="display: none" 29650 ><i class="fa fa-caret-down"></i> 29651 <span class="hover_link">Abstract</span></a 29652 > 29653 <div 29654 data-display-control="664_1707793552_7137" 29655 id="663_1707793552_713692" 29656 style="display: none" 29657 > 29658 <div class="arrow-slidedown"> 29659 <blockquote> 29660 Microgrids are localized power systems that can 29661 function independently or alongside the main 29662 grid. They consist of interconnected generators, 29663 energy storage, and loads that can be managed 29664 locally. Using SystemC-AMS, we demonstrate how 29665 microgrid components, including solar panels and 29666 converters, can be accurately modeled and 29667 simulated, along with their interactions. 29668 Real-time simulations are crucial for 29669 understanding microgrid behavior and optimizing 29670 components. This approach facilitates seamless 29671 integration with hardware prototypes and 29672 automation systems, supporting various 29673 development stages. Our study presents a 29674 best-case scenario for real-time simulation, 29675 assuming each loop takes less time than the 29676 simulation time step, with fallback to the 29677 previous value if data isn't received in time. 29678 This article introduces the first known 29679 real-time simulation strategy using SystemC-AMS, 29680 enabling the real-time simulation of microgrid 29681 components and integration with external 29682 devices. The implementation adopts a model-based 29683 design approach, creating increasingly complex 29684 systems with grid components and controllers. 29685 </blockquote> 29686 </div> 29687 </div> 29688 </div> 29689 </div> 29690 <div class="slot-urls"></div> 29691 <a href="/wsc23papers/270.pdf" target="_blank">pdf</a 29692 ><br /> 29693 </div> 29694 <div class="slot-entry"> 29695 <a name="cea147" tabindex="-1"></a> 29696 <div class="slot-title-line"> 29697 <span class="slot-title" 29698 >Advancing Safety in Nuclear Applications with Reduced 29699 Order Modeling and Digital Twin</span 29700 > 29701 </div> 29702 <div class="slot-authors"> 29703 Justin Williams, Nicole Hatch, Jean Ragusa, and Jian Tao 29704 (Texas A&M University) 29705 </div> 29706 <div class="slot-abstract"> 29707 <div> 29708 <a 29709 class="clickable no-decoration" 29710 id="vhsjs_view_666_1707793552_7158077" 29711 onclick="$('#vhsjs_view_666_1707793552_7158077').hide(); 29712 $('#vhsjs_hide_666_1707793552_7158077').show(); 29713 $('#665_1707793552_7158').slideDown(function() { 29714 if (typeof Masonry === 'function') { 29715 $('.use_masonry').masonry(); 29716 }; 29717 29718 });" 29719 ><i class="fa fa-caret-right"></i>
29720 <span class="hover_link">Abstract</span></a 29721 ><a 29722 class="clickable no-decoration" 29723 id="vhsjs_hide_666_1707793552_7158077" 29724 onclick="$('#665_1707793552_7158').hide(function() { 29725 if (typeof Masonry === 'function') { 29726 $('.use_masonry').masonry(); 29727 }; 29728 }); 29729 $('#vhsjs_hide_666_1707793552_7158077').hide(); 29730 $('#vhsjs_view_666_1707793552_7158077').show();" 29731 style="display: none" 29732 ><i class="fa fa-caret-down"></i> 29733 <span class="hover_link">Abstract</span></a 29734 > 29735 <div 29736 data-display-control="666_1707793552_7158077" 29737 id="665_1707793552_7158" 29738 style="display: none" 29739 > 29740 <div class="arrow-slidedown"> 29741 <blockquote> 29742 Ionizing radiation refers to particles or 29743 photons that carry enough energy to remove 29744 electrons from atoms or molecules. Through 29745 ionizing interactions, radiation can have severe 29746 implications for human health and the 29747 environment, making it essential to develop 29748 effective strategies to manage the risks it 29749 poses. To display the potential benefits from 29750 the application of digital twin technologies to 29751 concerns regarding radioactive material in 29752 laboratory, university, and national defense 29753 settings, this paper presents the development of 29754 a digital twin framework, and potential use 29755 cases for the framework. The platform was 29756 demonstrated in two scenario studies. The first 29757 scenario involves a faux radiation-detecting 29758 glovebox used for lab safety education, while 29759 the second scenario addresses training for first 29760 responders in a nuclear defense and safety 29761 situation. 29762 </blockquote> 29763 </div> 29764 </div> 29765 </div> 29766 </div> 29767 <div class="slot-urls"></div> 29768 <a href="/wsc23papers/cea147.pdf" target="_blank">pdf</a 29769 ><br /> 29770 </div> 29771 <div class="slot-entry"> 29772 <a name="cea140" tabindex="-1"></a> 29773 <div class="slot-title-line"> 29774 <span class="slot-title" 29775 >Simulation as a Soft Digital Twin for Maintenance 29776 Reliability Operations</span 29777 > 29778 </div> 29779 <div class="slot-authors"> 29780 Xueping Li, Thomas Berg, Gerald Jones, and Kimon Swanson 29781 (University of Tennessee, Knoxville) and Vincent 29782 Lamberti, Luke Birt, and Pugazenthi Atchayagopal 29783 (Consolidated Nuclear Security, LLC) 29784 </div> 29785 <div class="slot-abstract"> 29786 <div> 29787 <a 29788 class="clickable no-decoration" 29789 id="vhsjs_view_668_1707793552_7180915" 29790 onclick="$('#vhsjs_view_668_1707793552_7180915').hide(); 29791 $('#vhsjs_hide_668_1707793552_7180915').show(); 29792 $('#667_1707793552_7180836').slideDown(function() { 29793 if (typeof Masonry === 'function') { 29794 $('.use_masonry').masonry(); 29795 }; 29796 29797 });" 29798 ><i class="fa fa-caret-right"></i>
29799 <span class="hover_link">Abstract</span></a 29800 ><a 29801 class="clickable no-decoration" 29802 id="vhsjs_hide_668_1707793552_7180915" 29803 onclick="$('#667_1707793552_7180836').hide(function() { 29804 if (typeof Masonry === 'function') { 29805 $('.use_masonry').masonry(); 29806 }; 29807 }); 29808 $('#vhsjs_hide_668_1707793552_7180915').hide(); 29809 $('#vhsjs_view_668_1707793552_7180915').show();" 29810 style="display: none" 29811 ><i class="fa fa-caret-down"></i> 29812 <span class="hover_link">Abstract</span></a 29813 > 29814 <div 29815 data-display-control="668_1707793552_7180915" 29816 id="667_1707793552_7180836" 29817 style="display: none" 29818 > 29819 <div class="arrow-slidedown"> 29820 <blockquote> 29821 A critical facility's reliability relies heavily 29822 on its maintenance process's effectiveness. This 29823 process involves numerous sub-processes, which 29824 can be challenging to model due to uncertainties 29825 and complexities. System managers often seek a 29826 predictive tool, and this work extends a 29827 previous study that developed a digital twin of 29828 a nuclear facility's maintenance task process 29829 using data-driven and stochastic modeling, along 29830 with expert input. The authors extended the 29831 project's previous iteration by enhancing the 29832 bootstrapping technique and improving the 29833 model's fidelity. 29834 </blockquote> 29835 </div> 29836 </div> 29837 </div> 29838 </div> 29839 <div class="slot-urls"></div> 29840 <a href="/wsc23papers/cea140.pdf" target="_blank">pdf</a 29841 ><br /> 29842 </div> 29843 </div> 29844 <div class="session-entry"> 29845 <span class="session-event-type">Technical Session</span 29846 ><span class="type-track-spacer"> · </span 29847 ><span class="program-track" 29848 >Simulation as Digital Twin</span 29849 ><br /> 29850 <div class="session-title"> 29851 Digital Twins and Warehouse Logistics 29852 </div> 29853 <div class="session-chair"> 29854 Chair: Edward Y. Hua (MITRE Corporation)<br /> 29855 </div> 29856 <div class="slot-entry"> 29857 <a name="inv139" tabindex="-1"></a> 29858 <div class="slot-title-line"> 29859 <span class="slot-title" 29860 >Renovation Logistics Park with Digital Twinning: A 29861 Simulation-Optimization-Powered Toolbox</span 29862 > 29863 </div> 29864 <div class="slot-authors"> 29865 Peixue Yuan (Northwestern Polytechnical University), Chi 29866 Zhang (Xi'an Jiaotong University), and Chenhao Zhou and 29867 Li Xue (Northwestern Polytechnical University) 29868 </div> 29869 <div class="slot-abstract"> 29870 <div> 29871 <a 29872 class="clickable no-decoration" 29873 id="vhsjs_view_670_1707793552_7224627" 29874 onclick="$('#vhsjs_view_670_1707793552_7224627').hide(); 29875 $('#vhsjs_hide_670_1707793552_7224627').show(); 29876 $('#669_1707793552_7224548').slideDown(function() { 29877 if (typeof Masonry === 'function') { 29878 $('.use_masonry').masonry(); 29879 }; 29880 29881 });" 29882 ><i class="fa fa-caret-right"></i>
29883 <span class="hover_link">Abstract</span></a 29884 ><a 29885 class="clickable no-decoration" 29886 id="vhsjs_hide_670_1707793552_7224627" 29887 onclick="$('#669_1707793552_7224548').hide(function() { 29888 if (typeof Masonry === 'function') { 29889 $('.use_masonry').masonry(); 29890 }; 29891 }); 29892 $('#vhsjs_hide_670_1707793552_7224627').hide(); 29893 $('#vhsjs_view_670_1707793552_7224627').show();" 29894 style="display: none" 29895 ><i class="fa fa-caret-down"></i> 29896 <span class="hover_link">Abstract</span></a 29897 > 29898 <div 29899 data-display-control="670_1707793552_7224627" 29900 id="669_1707793552_7224548" 29901 style="display: none" 29902 > 29903 <div class="arrow-slidedown"> 29904 <blockquote> 29905 Taking into account the crucial node of the 29906 logistics network, this paper concentrates on 29907 the layout design problem of logistics parks 29908 considering numerous uncertain factors during 29909 operations. To provide comprehensive support for 29910 park planners and managers, a 29911 simulation-optimization-powered toolbox is 29912 developed for decision-making, with core 29913 functions such as park layout design, 29914 construction quantity calculations, and 29915 performance evaluations. A case study 29916 demonstrates the toolbox's effectiveness in 29917 assisting users to achieve their desired layout 29918 designs, and the result shows that the optimized 29919 layout generated by the toolbox can lead to 29920 improvements of approximately 13%. 29921 </blockquote> 29922 </div> 29923 </div> 29924 </div> 29925 </div> 29926 <div class="slot-urls"></div> 29927 <a href="/wsc23papers/271.pdf" target="_blank">pdf</a 29928 ><br /> 29929 </div> 29930 <div class="slot-entry"> 29931 <a name="con111" tabindex="-1"></a> 29932 <div class="slot-title-line"> 29933 <span class="slot-title" 29934 >A Simulation Optimization Method for Scheduling 29935 Automated Guided Vehicles in a Stochastic Warehouse 29936 Management System</span 29937 > 29938 </div> 29939 <div class="slot-authors"> 29940 Gongbo Zhang, Xiaotian Liu, and Yijie Peng (Peking 29941 University) 29942 </div> 29943 <div class="slot-abstract"> 29944 <div> 29945 <a 29946 class="clickable no-decoration" 29947 id="vhsjs_view_672_1707793552_7247078" 29948 onclick="$('#vhsjs_view_672_1707793552_7247078').hide(); 29949 $('#vhsjs_hide_672_1707793552_7247078').show(); 29950 $('#671_1707793552_7246997').slideDown(function() { 29951 if (typeof Masonry === 'function') { 29952 $('.use_masonry').masonry(); 29953 }; 29954 29955 });" 29956 ><i class="fa fa-caret-right"></i> 29957 <span class="hover_link">Abstract</span></a 29958 ><a 29959 class="clickable no-decoration" 29960 id="vhsjs_hide_672_1707793552_7247078" 29961 onclick="$('#671_1707793552_7246997').hide(function() { 29962 if (typeof Masonry === 'function') { 29963 $('.use_masonry').masonry(); 29964 }; 29965 }); 29966 $('#vhsjs_hide_672_1707793552_7247078').hide(); 29967 $('#vhsjs_view_672_1707793552_7247078').show();" 29968 style="display: none" 29969 ><i class="fa fa-caret-down"></i>
29970 <span class="hover_link">Abstract</span></a 29971 > 29972 <div 29973 data-display-control="672_1707793552_7247078" 29974 id="671_1707793552_7246997" 29975 style="display: none" 29976 > 29977 <div class="arrow-slidedown"> 29978 <blockquote> 29979 We consider the problem of scheduling automated 29980 guided vehicles (AGVs) in a stochastic warehouse 29981 management system. This problem was studied in 29982 the Case Study Competition of the 2022 Winter 29983 Simulation Conference. We propose a simulation 29984 optimization method that simultaneously 29985 optimizes dispatching and route planning for 29986 AGVs to enhance the system performance. 29987 Experimental results on two warehouse system 29988 simulation scenarios demonstrate that the 29989 proposed method outperforms the default method. 29990 </blockquote> 29991 </div> 29992 </div> 29993 </div> 29994 </div> 29995 <div class="slot-urls"></div> 29996 <a href="/wsc23papers/272.pdf" target="_blank">pdf</a 29997 ><br /> 29998 </div> 29999 <div class="slot-entry"> 30000 <a name="con256" tabindex="-1"></a> 30001 <div class="slot-title-line"> 30002 <span class="slot-title" 30003 >Emulation and Digital Twin Framework for the 30004 Validation of Material Handling Equipment in Warehouse 30005 Environments</span 30006 > 30007 </div> 30008 <div class="slot-authors"> 30009 Ankit Pandey, Rachael Flam, Raashid Mohammed, and Achuta 30010 Kalidindi (Amazon) 30011 </div> 30012 <div class="slot-abstract"> 30013 <div> 30014 <a 30015 class="clickable no-decoration" 30016 id="vhsjs_view_674_1707793552_7269747" 30017 onclick="$('#vhsjs_view_674_1707793552_7269747').hide(); 30018 $('#vhsjs_hide_674_1707793552_7269747').show(); 30019 $('#673_1707793552_7269666').slideDown(function() { 30020 if (typeof Masonry === 'function') { 30021 $('.use_masonry').masonry(); 30022 }; 30023 30024 });" 30025 ><i class="fa fa-caret-right"></i> 30026 <span class="hover_link">Abstract</span></a 30027 ><a 30028 class="clickable no-decoration" 30029 id="vhsjs_hide_674_1707793552_7269747" 30030 onclick="$('#673_1707793552_7269666').hide(function() { 30031 if (typeof Masonry === 'function') { 30032 $('.use_masonry').masonry(); 30033 }; 30034 }); 30035 $('#vhsjs_hide_674_1707793552_7269747').hide(); 30036 $('#vhsjs_view_674_1707793552_7269747').show();" 30037 style="display: none" 30038 ><i class="fa fa-caret-down"></i> 30039 <span class="hover_link">Abstract</span></a 30040 > 30041 <div 30042 data-display-control="674_1707793552_7269747" 30043 id="673_1707793552_7269666" 30044 style="display: none" 30045 > 30046 <div class="arrow-slidedown"> 30047 <blockquote> 30048 With modern warehouses becoming more automated, 30049 there is a growing opportunity to test and 30050 validate material handling concepts throughout 30051 the project life cycle. Emulation and digital 30052 twin pose a capability for material handling 30053 system validation from the ideation stage 30054 through post-implementation. An emulation model 30055 is a virtual replica of a physical system, and 30056 digital twin is a transformation of an emulation 30057 model via connection to a virtual or physical 30058 controller. They can test factors such as design 30059 mechanics and layouts, calculate throughput, 30060 test controls logic, and perform product flow 30061 analysis. Evaluation of these factors can 30062 provide a relatively accurate metric for system 30063 performance and lead to a more comprehensive 30064 return on investment (ROI) analysis. This paper 30065 discusses how incorporation of emulation and 30066 digital twin into all stages of the project life 30067 cycle of material handling systems can improve 30068 system efficiency and prevent live system 30069 commissioning risk. 30070 </blockquote> 30071 </div> 30072 </div> 30073 </div> 30074 </div> 30075 <div class="slot-urls"></div> 30076 <a href="/wsc23papers/273.pdf" target="_blank">pdf</a 30077 ><br /> 30078 </div> 30079 </div> 30080 <div class="session-entry">
30081 <span class="session-event-type">Technical Session</span 30082 ><span class="type-track-spacer"> · </span 30083 ><span class="program-track" 30084 >Simulation as Digital Twin</span 30085 ><br /> 30086 <div class="session-title"> 30087 Digital Twins and Manufacturing 30088 </div> 30089 <div class="session-chair"> 30090 Chair: Cathal Heavey (University of Limerick)<br /> 30091 </div> 30092 <div class="slot-entry"> 30093 <a name="inv147" tabindex="-1"></a> 30094 <div class="slot-title-line"> 30095 <span class="slot-title" 30096 >Simulation Based High Fidelity Digital Twins of 30097 Manufacturing Systems: An Application Model and 30098 Industrial Use Case</span 30099 > 30100 </div> 30101 <div class="slot-authors"> 30102 Ali Ahmad Malik (Oakland University) 30103 </div> 30104 <div class="slot-abstract"> 30105 <div> 30106 <a 30107 class="clickable no-decoration" 30108 id="vhsjs_view_676_1707793552_7312381" 30109 onclick="$('#vhsjs_view_676_1707793552_7312381').hide(); 30110 $('#vhsjs_hide_676_1707793552_7312381').show(); 30111 $('#675_1707793552_7312295').slideDown(function() { 30112 if (typeof Masonry === 'function') { 30113 $('.use_masonry').masonry(); 30114 }; 30115 30116 });" 30117 ><i class="fa fa-caret-right"></i> 30118 <span class="hover_link">Abstract</span></a 30119 ><a 30120 class="clickable no-decoration" 30121 id="vhsjs_hide_676_1707793552_7312381" 30122 onclick="$('#675_1707793552_7312295').hide(function() { 30123 if (typeof Masonry === 'function') { 30124 $('.use_masonry').masonry(); 30125 }; 30126 }); 30127 $('#vhsjs_hide_676_1707793552_7312381').hide(); 30128 $('#vhsjs_view_676_1707793552_7312381').show();" 30129 style="display: none" 30130 ><i class="fa fa-caret-down"></i> 30131 <span class="hover_link">Abstract</span></a 30132 > 30133 <div 30134 data-display-control="676_1707793552_7312381" 30135 id="675_1707793552_7312295" 30136 style="display: none" 30137 > 30138 <div class="arrow-slidedown"> 30139 <blockquote> 30140 Modern manufacturing systems are required to be 30141 developed, commissioned, and reconfigured faster 30142 than ever before. Conventional methods for the 30143 development of manufacturing systems are 30144 time-consuming due to their sequential nature. A 30145 digital twin is an emerging technology that can 30146 offer a high-fidelity simulation of a real 30147 manufacturing system including its kinematics, 30148 automation program, behavior, user interface, 30149 and production parameters. Such a unified 30150 digital twin can be used as a support tool for 30151 verification and validation of complex behavior 30152 of modern-day manufacturing systems during 30153 design, commissioning, reconfiguration, 30154 maintenance, and for end-of-life. The resulting 30155 benefits are to speed up the development and 30156 reconfiguration phases and improve system 30157 reliability. This article presents a framework 30158 to develop and use a digital twin for the 30159 development of complex machines. An industrial 30160 case from a large automation company is 30161 presented. 30162 </blockquote> 30163 </div> 30164 </div> 30165 </div> 30166 </div> 30167 <div class="slot-urls"></div> 30168 <a href="/wsc23papers/274.pdf" target="_blank">pdf</a 30169 ><br /> 30170 </div> 30171 <div class="slot-entry"> 30172 <a name="inv173" tabindex="-1"></a> 30173 <div class="slot-title-line">
30174 <span class="slot-title" 30175 >Data Requirements for a Digital Twin of a Robot 30176 Workcell</span 30177 > 30178 </div> 30179 <div class="slot-authors"> 30180 Deogratias Kibira (National Institute of Standards and 30181 Technology, University of Maryland - College Park) and 30182 Guodong Shao (National Institute of Standards and 30183 Technology) 30184 </div> 30185 <div class="slot-abstract"> 30186 <div> 30187 <a 30188 class="clickable no-decoration" 30189 id="vhsjs_view_678_1707793552_7334347" 30190 onclick="$('#vhsjs_view_678_1707793552_7334347').hide(); 30191 $('#vhsjs_hide_678_1707793552_7334347').show(); 30192 $('#677_1707793552_733427').slideDown(function() { 30193 if (typeof Masonry === 'function') { 30194 $('.use_masonry').masonry(); 30195 }; 30196 30197 });" 30198 ><i class="fa fa-caret-right"></i> 30199 <span class="hover_link">Abstract</span></a 30200 ><a 30201 class="clickable no-decoration" 30202 id="vhsjs_hide_678_1707793552_7334347" 30203 onclick="$('#677_1707793552_733427').hide(function() { 30204 if (typeof Masonry === 'function') { 30205 $('.use_masonry').masonry(); 30206 }; 30207 }); 30208 $('#vhsjs_hide_678_1707793552_7334347').hide(); 30209 $('#vhsjs_view_678_1707793552_7334347').show();" 30210 style="display: none" 30211 ><i class="fa fa-caret-down"></i> 30212 <span class="hover_link">Abstract</span></a 30213 > 30214 <div 30215 data-display-control="678_1707793552_7334347" 30216 id="677_1707793552_733427" 30217 style="display: none" 30218 > 30219 <div class="arrow-slidedown"> 30220 <blockquote> 30221 The applications of digital twins continue to 30222 grow with the volume and variety of data 30223 collected. These data support the modeling of 30224 function, behavior, and structure of a physical 30225 element. However, successfully building a 30226 digital twin requires data identification, data 30227 fusion, and data management. Thus, despite the 30228 increase in data availability, there are still 30229 challenges of data usage, especially data 30230 scoping and scaling to implement a digital twin 30231 for a specific purpose. The objective of this 30232 paper is to identify data requirements for 30233 various types of digital twins for a robot 30234 workcell. The identification includes data 30235 description, source, method of collection, and 30236 data formats. The digital twin types include 30237 descriptive digital twins, diagnostics and 30238 prognostics digital twins, prescriptive digital 30239 twins, and intelligent digital twins. The 30240 outcome of this data requirements identification 30241 can be used as a guide for developing and 30242 validating digital twins for a robot workcell 30243 lifecycle. 30244 </blockquote> 30245 </div> 30246 </div> 30247 </div> 30248 </div> 30249 <div class="slot-urls"></div> 30250 <a href="/wsc23papers/275.pdf" target="_blank">pdf</a 30251 ><br /> 30252 </div> 30253 <div class="slot-entry"> 30254 <a name="con255" tabindex="-1"></a> 30255 <div class="slot-title-line">
30256 <span class="slot-title" 30257 >A Digital Twin for Production Control Based on 30258 Remaining Cycle Time Prediction</span 30259 > 30260 </div> 30261 <div class="slot-authors"> 30262 Giovanni Lugaresi (KU Leuven); Pedro Luis Bacelar Dos 30263 Santos, Alex Chalissery Lona, and Monica Rossi 30264 (Politecnico di Milano); Eduardo Zancul (University of 30265 Sao Paulo); and Andrea Matta (Politecnico di Milano) 30266 </div> 30267 <div class="slot-abstract"> 30268 <div> 30269 <a 30270 class="clickable no-decoration" 30271 id="vhsjs_view_680_1707793552_735861" 30272 onclick="$('#vhsjs_view_680_1707793552_735861').hide(); 30273 $('#vhsjs_hide_680_1707793552_735861').show(); 30274 $('#679_1707793552_735853').slideDown(function() { 30275 if (typeof Masonry === 'function') { 30276 $('.use_masonry').masonry(); 30277 }; 30278 30279 });" 30280 ><i class="fa fa-caret-right"></i> 30281 <span class="hover_link">Abstract</span></a 30282 ><a 30283 class="clickable no-decoration" 30284 id="vhsjs_hide_680_1707793552_735861" 30285 onclick="$('#679_1707793552_735853').hide(function() { 30286 if (typeof Masonry === 'function') { 30287 $('.use_masonry').masonry(); 30288 }; 30289 }); 30290 $('#vhsjs_hide_680_1707793552_735861').hide(); 30291 $('#vhsjs_view_680_1707793552_735861').show();" 30292 style="display: none" 30293 ><i class="fa fa-caret-down"></i> 30294 <span class="hover_link">Abstract</span></a 30295 > 30296 <div 30297 data-display-control="680_1707793552_735861" 30298 id="679_1707793552_735853" 30299 style="display: none" 30300 > 30301 <div class="arrow-slidedown"> 30302 <blockquote> 30303 The recent industrial context pushed 30304 manufacturers to invest heavily in digitization 30305 for a more efficient use of their equipment and 30306 scarce resources. The digitization of industrial 30307 environments allows the establishment of digital 30308 decision-support tools such as digital twins, to 30309 exploit the shop-floor data for making more 30310 accurate decisions considering the real system 30311 state. Existing literature focuses on the 30312 development of specific digital twin components 30313 as well as methods that are typically developed 30314 and tested without an integration within a 30315 digital twin architecture. This paper proposes a 30316 complete digital twin framework with the purpose 30317 of aiding production planning and control 30318 operations. The focus is on the design of a 30319 production control service that manages the 30320 material flow in the real system using 30321 simulation-based predictions of the remaining 30322 cycle time. Preliminary experiments are done by 30323 applying the digital twin architecture on a 30324 lab-scale model, demonstrating the applicability 30325 of the proposed approach. 30326 </blockquote> 30327 </div> 30328 </div> 30329 </div> 30330 </div> 30331 <div class="slot-urls"></div> 30332 <a href="/wsc23papers/276.pdf" target="_blank">pdf</a 30333 ><br /> 30334 </div> 30335 </div> 30336 <div class="session-entry">
30337 <span class="session-event-type">Technical Session</span 30338 ><span class="type-track-spacer"> · </span 30339 ><span class="program-track" 30340 >Simulation as Digital Twin</span 30341 ><br /> 30342 <div class="session-title"> 30343 Panel: Enhancing Digital Twins with Advances in Simulation 30344 and Artificial Intelligence: Opportunities and Challenges 30345 </div> 30346 <div class="session-chair"> 30347 Chair: Barry L. Nelson (Northwestern University)<br /> 30348 </div> 30349 <div class="slot-entry"> 30350 <a name="inv205" tabindex="-1"></a> 30351 <div class="slot-title-line"> 30352 <span class="slot-title" 30353 >Enhancing Digital Twins with Advances in Simulation 30354 and Artificial Intelligence: Opportunities and 30355 Challenges</span 30356 > 30357 </div> 30358 <div class="slot-authors"> 30359 Simon J. E. Taylor (Brunel University London), Charles 30360 Macal (Argonne National Laboratory), Andrea Matta 30361 (Politecnico di Milano), Markus Rabe (TU Dortmund 30362 University), Susan Sanchez (Naval Postgraduate School), 30363 and Guodong Shao (National Institute of Standards and 30364 Technology) 30365 </div> 30366 <div class="slot-abstract"> 30367 <div> 30368 <a 30369 class="clickable no-decoration" 30370 id="vhsjs_view_682_1707793552_7402744" 30371 onclick="$('#vhsjs_view_682_1707793552_7402744').hide(); 30372 $('#vhsjs_hide_682_1707793552_7402744').show(); 30373 $('#681_1707793552_7402658').slideDown(function() { 30374 if (typeof Masonry === 'function') { 30375 $('.use_masonry').masonry(); 30376 }; 30377 30378 });" 30379 ><i class="fa fa-caret-right"></i> 30380 <span class="hover_link">Abstract</span></a 30381 ><a 30382 class="clickable no-decoration" 30383 id="vhsjs_hide_682_1707793552_7402744" 30384 onclick="$('#681_1707793552_7402658').hide(function() { 30385 if (typeof Masonry === 'function') { 30386 $('.use_masonry').masonry(); 30387 }; 30388 }); 30389 $('#vhsjs_hide_682_1707793552_7402744').hide(); 30390 $('#vhsjs_view_682_1707793552_7402744').show();" 30391 style="display: none" 30392 ><i class="fa fa-caret-down"></i> 30393 <span class="hover_link">Abstract</span></a 30394 > 30395 <div 30396 data-display-control="682_1707793552_7402744" 30397 id="681_1707793552_7402658" 30398 style="display: none" 30399 > 30400 <div class="arrow-slidedown"> 30401 <blockquote> 30402 Simulations are used to investigate physical 30403 systems. A digital twin goes beyond this by 30404 connecting a simulation with the physical system 30405 with the purpose of analyzing and controlling 30406 that system in real-time. In the past 5 years 30407 there has been a substantial increase in 30408 research into Simulation and Artificial 30409 Intelligence (AI). The combination of Simulation 30410 with AI presents many possible innovations. 30411 Similarly, combining AI with Simulation presents 30412 further possibilities including approaches to 30413 developing trustworthy and explainable AI 30414 methods, solutions to problems arising from 30415 sparce or no data and better methods for time
30416 series analysis. Given the progress that has 30417 been made in Digital Twins and Simulation and 30418 AI, what opportunities are there from combining 30419 these two exciting research areas? What 30420 challenges need to be overcome to achieve these? 30421 This article discusses these from the 30422 perspectives of six leading members of the 30423 Modeling & Simulation community. 30424 </blockquote> 30425 </div> 30426 </div> 30427 </div> 30428 </div> 30429 <div class="slot-urls"></div> 30430 <a href="/wsc23papers/277.pdf" target="_blank">pdf</a 30431 ><br /> 30432 </div> 30433 </div> 30434 </div> 30435 <div class="centered"> 30436 <div class="top-link"><a href="#top">Return to Top</a></div> 30437 </div> 30438 <hr /> 30439 </div> 30440 <div class="area-section"> 30441 <div class="centered"> 30442 <a name="ptrack106" tabindex="-1"></a> 30443 <div class="section-title">Simulation in Education</div> 30444 </div> 30445 <div class="centered track-chair"> 30446 <span class="track-chair-role" 30447 >Track Coordinator - Simulation in Education: </span 30448 ><span class="track-chair-names" 30449 >Omar Ashour (Penn State University), Christopher Lynch (Old 30450 Dominion University)</span 30451 > 30452 </div> 30453 <div class="section-entry"> 30454 <div class="session-entry"> 30455 <span class="session-event-type">Technical Session</span 30456 ><span class="type-track-spacer"> · </span 30457 ><span class="program-track">Simulation in Education</span 30458 ><br /> 30459 <div class="session-title"> 30460 Tools and Technologies in Simulation Education 30461 </div> 30462 <div class="session-chair"> 30463 Chair: Manuel D. Rossetti (University of Arkansas)<br /> 30464 </div> 30465 <div class="slot-entry"> 30466 <a name="con152" tabindex="-1"></a> 30467 <div class="slot-title-line"> 30468 <span class="slot-title" 30469 >Introducing the Kotlin Simulation Library (KSL)</span 30470 > 30471 </div> 30472 <div class="slot-authors"> 30473 Manuel D. Rossetti (University of Arkansas) 30474 </div> 30475 <div class="slot-abstract"> 30476 <div> 30477 <a 30478 class="clickable no-decoration" 30479 id="vhsjs_view_684_1707793552_7472665" 30480 onclick="$('#vhsjs_view_684_1707793552_7472665').hide(); 30481 $('#vhsjs_hide_684_1707793552_7472665').show(); 30482 $('#683_1707793552_7472582').slideDown(function() { 30483 if (typeof Masonry === 'function') { 30484 $('.use_masonry').masonry(); 30485 }; 30486 30487 });" 30488 ><i class="fa fa-caret-right"></i> 30489 <span class="hover_link">Abstract</span></a 30490 ><a 30491 class="clickable no-decoration" 30492 id="vhsjs_hide_684_1707793552_7472665" 30493 onclick="$('#683_1707793552_7472582').hide(function() { 30494 if (typeof Masonry === 'function') { 30495 $('.use_masonry').masonry(); 30496 }; 30497 }); 30498 $('#vhsjs_hide_684_1707793552_7472665').hide(); 30499 $('#vhsjs_view_684_1707793552_7472665').show();" 30500 style="display: none" 30501 ><i class="fa fa-caret-down"></i>
30502 <span class="hover_link">Abstract</span></a 30503 > 30504 <div 30505 data-display-control="684_1707793552_7472665" 30506 id="683_1707793552_7472582" 30507 style="display: none" 30508 > 30509 <div class="arrow-slidedown"> 30510 <blockquote> 30511 This paper introduces a Monte Carlo and 30512 discrete-event simulation library for the Kotlin 30513 programming language. The Kotlin Simulation 30514 Library (KSL) provides functionality to perform 30515 simulation experiments involving the generation 30516 of random processes, the execution of 30517 discrete-event simulation via the event and 30518 process views, and the analysis of the 30519 statistical quantities generated by simulation 30520 models. The architecture of the library 30521 leverages the object-oriented and functional 30522 programming capabilities of the widely used 30523 Kotlin programming language. The library 30524 provides functionality that is similar to 30525 proprietary software, while being open-source 30526 and readily extensible. This paper provides an 30527 overview of the architecture of the library. The 30528 functionality of the library is illustrated 30529 through several examples. 30530 </blockquote> 30531 </div> 30532 </div> 30533 </div> 30534 </div> 30535 <div class="slot-urls"></div> 30536 <a href="/wsc23papers/278.pdf" target="_blank">pdf</a 30537 ><br /> 30538 </div> 30539 <div class="slot-entry"> 30540 <a name="con298" tabindex="-1"></a> 30541 <div class="slot-title-line"> 30542 <span class="slot-title" 30543 >Teaching Discrete Event Simulation Software Design in 30544 the Context of Computer Engineering</span 30545 > 30546 </div> 30547 <div class="slot-authors"> 30548 James Frederick Leathrum (Old Dominion University) 30549 </div> 30550 <div class="slot-abstract"> 30551 <div> 30552 <a 30553 class="clickable no-decoration" 30554 id="vhsjs_view_686_1707793552_7494211" 30555 onclick="$('#vhsjs_view_686_1707793552_7494211').hide(); 30556 $('#vhsjs_hide_686_1707793552_7494211').show(); 30557 $('#685_1707793552_749413').slideDown(function() { 30558 if (typeof Masonry === 'function') { 30559 $('.use_masonry').masonry(); 30560 }; 30561 30562 });" 30563 ><i class="fa fa-caret-right"></i> 30564 <span class="hover_link">Abstract</span></a 30565 ><a 30566 class="clickable no-decoration" 30567 id="vhsjs_hide_686_1707793552_7494211" 30568 onclick="$('#685_1707793552_749413').hide(function() { 30569 if (typeof Masonry === 'function') { 30570 $('.use_masonry').masonry(); 30571 }; 30572 }); 30573 $('#vhsjs_hide_686_1707793552_7494211').hide(); 30574 $('#vhsjs_view_686_1707793552_7494211').show();" 30575 style="display: none" 30576 ><i class="fa fa-caret-down"></i>
30577 <span class="hover_link">Abstract</span></a 30578 > 30579 <div 30580 data-display-control="686_1707793552_7494211" 30581 id="685_1707793552_749413" 30582 style="display: none" 30583 > 30584 <div class="arrow-slidedown"> 30585 <blockquote> 30586 Recent events resulted in the consolidation of a 30587 degree program in Modeling & Simulation 30588 Engineering with a degree in Computer 30589 Engineering, though with a major in Modeling & 30590 Simulation Engineering. The resulting major 30591 strongly highlights the computational aspects of 30592 M&S. However, the needs of discrete event 30593 simulation in computer engineering have somewhat 30594 of a different focus. For instance, the 30595 management of simultaneous events is crucial in 30596 digital circuit simulation. This paper looks at 30597 refocusing a course on discrete event simulation 30598 software design to meet the needs of a computer 30599 engineering degree while maintaining 30600 applicability to the more general community. It 30601 discusses modifications in the treatment of 30602 models and then mapping those models to 30603 software. 30604 </blockquote> 30605 </div> 30606 </div> 30607 </div> 30608 </div> 30609 <div class="slot-urls"></div> 30610 <a href="/wsc23papers/279.pdf" target="_blank">pdf</a 30611 ><br /> 30612 </div> 30613 </div> 30614 <div class="session-entry"> 30615 <span class="session-event-type">Technical Session</span 30616 ><span class="type-track-spacer"> · </span 30617 ><span class="program-track">Simulation in Education</span 30618 ><br /> 30619 <div class="session-title"> 30620 Panel: ChatGPT in M&S Education: Opportunities and 30621 Challenges 30622 </div> 30623 <div class="session-chair"> 30624 Chair: Andreas Tolk (The MITRE Corporation)<br /> 30625 </div> 30626 <div class="slot-entry"> 30627 <a name="inv105" tabindex="-1"></a> 30628 <div class="slot-title-line"> 30629 <span class="slot-title" 30630 >Chances and Challenges of ChatGPT and Similar Models 30631 for Education in M&S</span 30632 > 30633 </div> 30634 <div class="slot-authors"> 30635 Andreas Tolk (The MITRE Corporation), Philip Barry 30636 (L3Harris Corporation), Margaret Loper (Georgia Tech 30637 Research Institute), Ghaith Rabadi (University of 30638 Central Florida), William Scherer (University of 30639 Virginia), and Levent Yilmaz (Auburn University) 30640 </div> 30641 <div class="slot-abstract"> 30642 <div> 30643 <a 30644 class="clickable no-decoration" 30645 id="vhsjs_view_688_1707793552_7554483" 30646 onclick="$('#vhsjs_view_688_1707793552_7554483').hide(); 30647 $('#vhsjs_hide_688_1707793552_7554483').show(); 30648 $('#687_1707793552_7554402').slideDown(function() { 30649 if (typeof Masonry === 'function') { 30650 $('.use_masonry').masonry(); 30651 }; 30652 30653 });" 30654 ><i class="fa fa-caret-right"></i>
30655 <span class="hover_link">Abstract</span></a 30656 ><a 30657 class="clickable no-decoration" 30658 id="vhsjs_hide_688_1707793552_7554483" 30659 onclick="$('#687_1707793552_7554402').hide(function() { 30660 if (typeof Masonry === 'function') { 30661 $('.use_masonry').masonry(); 30662 }; 30663 }); 30664 $('#vhsjs_hide_688_1707793552_7554483').hide(); 30665 $('#vhsjs_view_688_1707793552_7554483').show();" 30666 style="display: none" 30667 ><i class="fa fa-caret-down"></i> 30668 <span class="hover_link">Abstract</span></a 30669 > 30670 <div 30671 data-display-control="688_1707793552_7554483" 30672 id="687_1707793552_7554402" 30673 style="display: none" 30674 > 30675 <div class="arrow-slidedown"> 30676 <blockquote> 30677 This position paper summarizes the inputs of a 30678 group of experts from academia and industry 30679 presenting their view on chances and challenges 30680 of using ChatGPT within Modeling and Simulation 30681 education. The experts also address the need to 30682 evaluate continuous education as well as 30683 education of faculty members to address 30684 scholastic challenges and opportunities while 30685 meeting the expectation of industry. Generally, 30686 the use of ChatGPT is encouraged, but it needs 30687 to be embedded into an updated curriculum with 30688 more emphasis on validity constraints, systems 30689 thinking, and ethics. 30690 </blockquote> 30691 </div> 30692 </div> 30693 </div> 30694 </div> 30695 <div class="slot-urls"></div> 30696 <a href="/wsc23papers/280.pdf" target="_blank">pdf</a 30697 ><br /> 30698 </div> 30699 </div> 30700 <div class="session-entry"> 30701 <span class="session-event-type">Technical Session</span 30702 ><span class="type-track-spacer"> · </span 30703 ><span class="program-track">Simulation in Education</span 30704 ><br /> 30705 <div class="session-title"> 30706 Behavioral and Entrepreneurial Aspects in Simulation 30707 </div> 30708 <div class="session-chair"> 30709 Chair: Canan Gunes Corlu (Boston University)<br /> 30710 </div> 30711 <div class="slot-entry"> 30712 <a name="con345" tabindex="-1"></a> 30713 <div class="slot-title-line"> 30714 <span class="slot-title" 30715 >Entrepreneurial Mindset Learning (EML) in Simulation 30716 Education</span 30717 > 30718 </div> 30719 <div class="slot-authors"> 30720 Michael E. Kuhl (Rochester Institute of Technology) 30721 </div> 30722 <div class="slot-abstract"> 30723 <div> 30724 <a 30725 class="clickable no-decoration" 30726 id="vhsjs_view_690_1707793552_7613363" 30727 onclick="$('#vhsjs_view_690_1707793552_7613363').hide(); 30728 $('#vhsjs_hide_690_1707793552_7613363').show(); 30729 $('#689_1707793552_7613275').slideDown(function() { 30730 if (typeof Masonry === 'function') { 30731 $('.use_masonry').masonry(); 30732 }; 30733 30734 });" 30735 ><i class="fa fa-caret-right"></i>
30736 <span class="hover_link">Abstract</span></a 30737 ><a 30738 class="clickable no-decoration" 30739 id="vhsjs_hide_690_1707793552_7613363" 30740 onclick="$('#689_1707793552_7613275').hide(function() { 30741 if (typeof Masonry === 'function') { 30742 $('.use_masonry').masonry(); 30743 }; 30744 }); 30745 $('#vhsjs_hide_690_1707793552_7613363').hide(); 30746 $('#vhsjs_view_690_1707793552_7613363').show();" 30747 style="display: none" 30748 ><i class="fa fa-caret-down"></i> 30749 <span class="hover_link">Abstract</span></a 30750 > 30751 <div 30752 data-display-control="690_1707793552_7613363" 30753 id="689_1707793552_7613275" 30754 style="display: none" 30755 > 30756 <div class="arrow-slidedown"> 30757 <blockquote> 30758 An entrepreneurial mindset is associated with 30759 recognizing and seeking opportunity that can 30760 result in societal benefits. Entrepreneurial 30761 minded learning (EML) is a pedagogy that has 30762 gained increasing attention in science, 30763 technology, engineering, and math education. In 30764 this paper, we present as set of examples to 30765 illustrate how EML methods can be applied in 30766 simulation courses to foster the development of 30767 the entrepreneurial mindset of students. In 30768 addition, we discuss some of the opportunities 30769 and challenges for adoption of EML in simulation 30770 education. 30771 </blockquote> 30772 </div> 30773 </div> 30774 </div> 30775 </div> 30776 <div class="slot-urls"></div> 30777 <a href="/wsc23papers/281.pdf" target="_blank">pdf</a 30778 ><br /> 30779 </div> 30780 <div class="slot-entry"> 30781 <a name="con240" tabindex="-1"></a> 30782 <div class="slot-title-line"> 30783 <span class="slot-title" 30784 >Can Gambling Ads Affect Customer Risk Behavior? A 30785 Simulation Study to the “888” Case</span 30786 > 30787 </div> 30788 <div class="slot-authors"> 30789 David Lopez-Lopez (ESADE business school), Giovanni 30790 Giusti (Tecnocampus - Pompeu Fabra University), Angel A. 30791 Juan (Universitat Politecnica de Valenci), and Canan 30792 Gunes Corlu (Boston University) 30793 </div> 30794 <div class="slot-abstract"> 30795 <div> 30796 <a 30797 class="clickable no-decoration" 30798 id="vhsjs_view_692_1707793552_7637403" 30799 onclick="$('#vhsjs_view_692_1707793552_7637403').hide(); 30800 $('#vhsjs_hide_692_1707793552_7637403').show(); 30801 $('#691_1707793552_7637324').slideDown(function() { 30802 if (typeof Masonry === 'function') { 30803 $('.use_masonry').masonry(); 30804 }; 30805 30806 });" 30807 ><i class="fa fa-caret-right"></i> 30808 <span class="hover_link">Abstract</span></a 30809 ><a 30810 class="clickable no-decoration" 30811 id="vhsjs_hide_692_1707793552_7637403" 30812 onclick="$('#691_1707793552_7637324').hide(function() { 30813 if (typeof Masonry === 'function') { 30814 $('.use_masonry').masonry(); 30815 }; 30816 }); 30817 $('#vhsjs_hide_692_1707793552_7637403').hide(); 30818 $('#vhsjs_view_692_1707793552_7637403').show();" 30819 style="display: none" 30820 ><i class="fa fa-caret-down"></i>
30821 <span class="hover_link">Abstract</span></a 30822 > 30823 <div 30824 data-display-control="692_1707793552_7637403" 30825 id="691_1707793552_7637324" 30826 style="display: none" 30827 > 30828 <div class="arrow-slidedown"> 30829 <blockquote> 30830 The aim of this research paper is to investigate 30831 the connection between advertising and consumer 30832 behavior in the gambling industry, which heavily 30833 relies on advertising. Specifically, it examines 30834 the impact of advertising on risky behavior 30835 among consumers, using the well-known Spanish 30836 gambling brand “888 Poker” as a case 30837 study. The experimental design involves a 30838 simulated asset market approach with 92 30839 participants, and the data collected is analyzed 30840 to draw conclusions regarding the relationship 30841 between advertising and risky behavior in the 30842 context of the gambling industry. 30843 </blockquote> 30844 </div> 30845 </div> 30846 </div> 30847 </div> 30848 <div class="slot-urls"></div> 30849 <a href="/wsc23papers/282.pdf" target="_blank">pdf</a 30850 ><br /> 30851 </div> 30852 </div> 30853 </div> 30854 <div class="centered"> 30855 <div class="top-link"><a href="#top">Return to Top</a></div> 30856 </div> 30857 <hr /> 30858 </div> 30859 <div class="area-section"> 30860 <div class="centered"> 30861 <a name="ptrack121" tabindex="-1"></a> 30862 <div class="section-title">Simulation Optimization</div> 30863 </div> 30864 <div class="centered track-chair"> 30865 <span class="track-chair-role" 30866 >Track Coordinator - Simulation Optimization: </span 30867 ><span class="track-chair-names" 30868 >David J. Eckman (Texas A&M University), Siyang Gao (City 30869 University of Hong Kong)</span 30870 > 30871 </div> 30872 <div class="section-entry"> 30873 <div class="session-entry"> 30874 <span class="session-event-type">Technical Session</span 30875 ><span class="type-track-spacer"> · </span 30876 ><span class="program-track">Simulation Optimization</span 30877 ><br /> 30878 <div class="session-title">Ranking and Selection I</div> 30879 <div class="session-chair"> 30880 Chair: Travis Goodwin (MITRE Corporation)<br /> 30881 </div> 30882 <div class="slot-entry"> 30883 <a name="con279" tabindex="-1"></a> 30884 <div class="slot-title-line"> 30885 <span class="slot-title" 30886 >Risk-Sensitive Ordinal Optimization</span 30887 > 30888 </div> 30889 <div class="slot-authors"> 30890 Dohyun Ahn (The Chinese University of Hong Kong) and 30891 Taeho Kim (Texas A&M University) 30892 </div> 30893 <div class="slot-abstract"> 30894 <div> 30895 <a 30896 class="clickable no-decoration" 30897 id="vhsjs_view_694_1707793552_7709208" 30898 onclick="$('#vhsjs_view_694_1707793552_7709208').hide(); 30899 $('#vhsjs_hide_694_1707793552_7709208').show(); 30900 $('#693_1707793552_770913').slideDown(function() { 30901 if (typeof Masonry === 'function') { 30902 $('.use_masonry').masonry(); 30903 }; 30904 30905 });" 30906 ><i class="fa fa-caret-right"></i>
30907 <span class="hover_link">Abstract</span></a 30908 ><a 30909 class="clickable no-decoration" 30910 id="vhsjs_hide_694_1707793552_7709208" 30911 onclick="$('#693_1707793552_770913').hide(function() { 30912 if (typeof Masonry === 'function') { 30913 $('.use_masonry').masonry(); 30914 }; 30915 }); 30916 $('#vhsjs_hide_694_1707793552_7709208').hide(); 30917 $('#vhsjs_view_694_1707793552_7709208').show();" 30918 style="display: none" 30919 ><i class="fa fa-caret-down"></i> 30920 <span class="hover_link">Abstract</span></a 30921 > 30922 <div 30923 data-display-control="694_1707793552_7709208" 30924 id="693_1707793552_770913" 30925 style="display: none" 30926 > 30927 <div class="arrow-slidedown"> 30928 <blockquote> 30929 We consider the problem of risk-sensitive 30930 ordinal optimization, which aims to identify the 30931 "least risky'' system among a finite number of 30932 stochastic systems. Each system's riskiness is 30933 assumed to be measured by the probability that 30934 the system's loss exceeds a common threshold. 30935 Since the crude Monte Carlo estimator is highly 30936 inefficient in estimating rare-event 30937 probabilities, conventional ordinal optimization 30938 approaches coupled with that estimator show 30939 significant performance degradation in this 30940 problem, particularly for sufficiently large 30941 loss thresholds. To detour this issue, assuming 30942 that the parametric form of the underlying 30943 distribution is known, we propose to use the 30944 tail parameter, a function of distributional 30945 parameters, as a surrogate for the loss 30946 probability in comparing and ranking systems, 30947 which is shown to work well for many well-known 30948 distributions. Building upon this observation, 30949 we find the optimal computing budget allocation 30950 scheme that maximizes the likelihood of 30951 identifying the least risky system. 30952 </blockquote> 30953 </div> 30954 </div> 30955 </div> 30956 </div> 30957 <div class="slot-urls"></div> 30958 <a href="/wsc23papers/283.pdf" target="_blank">pdf</a 30959 ><br /> 30960 </div> 30961 <div class="slot-entry"> 30962 <a name="con104" tabindex="-1"></a> 30963 <div class="slot-title-line"> 30964 <span class="slot-title" 30965 >Data-Driven Optimal Allocation for Ranking and 30966 Selection under Unknown Sampling Distributions</span 30967 > 30968 </div> 30969 <div class="slot-authors"> 30970 Ye Chen (Virginia Commonwealth University) 30971 </div> 30972 <div class="slot-abstract"> 30973 <div> 30974 <a 30975 class="clickable no-decoration" 30976 id="vhsjs_view_696_1707793552_7730947" 30977 onclick="$('#vhsjs_view_696_1707793552_7730947').hide(); 30978 $('#vhsjs_hide_696_1707793552_7730947').show(); 30979 $('#695_1707793552_7730865').slideDown(function() { 30980 if (typeof Masonry === 'function') { 30981 $('.use_masonry').masonry(); 30982 }; 30983 30984 });" 30985 ><i class="fa fa-caret-right"></i>
30986 <span class="hover_link">Abstract</span></a 30987 ><a 30988 class="clickable no-decoration" 30989 id="vhsjs_hide_696_1707793552_7730947" 30990 onclick="$('#695_1707793552_7730865').hide(function() { 30991 if (typeof Masonry === 'function') { 30992 $('.use_masonry').masonry(); 30993 }; 30994 }); 30995 $('#vhsjs_hide_696_1707793552_7730947').hide(); 30996 $('#vhsjs_view_696_1707793552_7730947').show();" 30997 style="display: none" 30998 ><i class="fa fa-caret-down"></i> 30999 <span class="hover_link">Abstract</span></a 31000 > 31001 <div 31002 data-display-control="696_1707793552_7730947" 31003 id="695_1707793552_7730865" 31004 style="display: none" 31005 > 31006 <div class="arrow-slidedown"> 31007 <blockquote> 31008 Ranking and selection (R&S) is the problem of 31009 identifying the optimal alternative from 31010 multiple alternatives through sampling them. In 31011 the existing R&S literature, sampling 31012 distributions of the observations are usually 31013 assumed to be from some known parametric 31014 distribution families, even in works that 31015 consider input uncertainty. By contrast, this 31016 paper considers R&S under completely unknown 31017 sampling distributions. We for the first time 31018 propose a computationally-tractable 31019 nonparametric tuning-free sequential budget 31020 allocation strategy that can asymptotically 31021 achieve the optimal allocation specified by 31022 large deviation analysis. Especially, we propose 31023 a new point estimation approach for estimating 31024 the optimal large deviation rates directly, 31025 which efficiently solves the challenge of 31026 estimating large deviation rate functions for 31027 lack of known sampling distributions. 31028 </blockquote> 31029 </div> 31030 </div> 31031 </div> 31032 </div> 31033 <div class="slot-urls"></div> 31034 <a href="/wsc23papers/284.pdf" target="_blank">pdf</a 31035 ><br /> 31036 </div> 31037 <div class="slot-entry"> 31038 <a name="con148" tabindex="-1"></a> 31039 <div class="slot-title-line"> 31040 <span class="slot-title" 31041 >POMDP-based Ranking and Selection</span 31042 > 31043 </div> 31044 <div class="slot-authors"> 31045 Ruihan Zhou and Yijie Peng (China) 31046 </div> 31047 <div class="slot-abstract"> 31048 <div> 31049 <a 31050 class="clickable no-decoration" 31051 id="vhsjs_view_698_1707793552_7751992" 31052 onclick="$('#vhsjs_view_698_1707793552_7751992').hide(); 31053 $('#vhsjs_hide_698_1707793552_7751992').show(); 31054 $('#697_1707793552_775191').slideDown(function() { 31055 if (typeof Masonry === 'function') { 31056 $('.use_masonry').masonry(); 31057 }; 31058 31059 });" 31060 ><i class="fa fa-caret-right"></i>
31061 <span class="hover_link">Abstract</span></a 31062 ><a 31063 class="clickable no-decoration" 31064 id="vhsjs_hide_698_1707793552_7751992" 31065 onclick="$('#697_1707793552_775191').hide(function() { 31066 if (typeof Masonry === 'function') { 31067 $('.use_masonry').masonry(); 31068 }; 31069 }); 31070 $('#vhsjs_hide_698_1707793552_7751992').hide(); 31071 $('#vhsjs_view_698_1707793552_7751992').show();" 31072 style="display: none" 31073 ><i class="fa fa-caret-down"></i> 31074 <span class="hover_link">Abstract</span></a 31075 > 31076 <div 31077 data-display-control="698_1707793552_7751992" 31078 id="697_1707793552_775191" 31079 style="display: none" 31080 > 31081 <div class="arrow-slidedown"> 31082 <blockquote> 31083 In this paper, we formulate the ranking and 31084 selection (R&S) problem as a stochastic control 31085 problem under the Bayesian framework. We propose 31086 to use particle filter to approximate the 31087 posterior distribution of states under the 31088 general Bayesian framework. The learning and 31089 decision are treated under the umbrella of a 31090 partially observable Markov decision process and 31091 a rollout policy based on Monte Carlo simulation 31092 is proposed. This policy can use one or more 31093 classic R&S approaches as base policies to 31094 efficiently learn the value function by rolling 31095 out simulation trajectories. We present 31096 numerical examples to demonstrate the 31097 effectiveness of the rollout policy and the 31098 performance of our policy is significantly 31099 improved relatively to the base policies. 31100 </blockquote> 31101 </div> 31102 </div> 31103 </div> 31104 </div> 31105 <div class="slot-urls"></div> 31106 <a href="/wsc23papers/285.pdf" target="_blank">pdf</a 31107 ><br /> 31108 </div> 31109 </div> 31110 <div class="session-entry"> 31111 <span class="session-event-type">Technical Session</span 31112 ><span class="type-track-spacer"> · </span 31113 ><span class="program-track">Simulation Optimization</span 31114 ><br /> 31115 <div class="session-title">Ranking and Selection II</div> 31116 <div class="session-chair"> 31117 Chair: Ye Chen (Virginia Commonwealth University)<br /> 31118 </div> 31119 <div class="slot-entry"> 31120 <a name="con143" tabindex="-1"></a> 31121 <div class="slot-title-line"> 31122 <span class="slot-title" 31123 >Top-Two Thompson Sampling for Selecting 31124 Context-Dependent Best Designs</span 31125 > 31126 </div> 31127 <div class="slot-authors"> 31128 Xinbo Shi, Yijie Peng, and Gongbo Zhang (Guanghua School 31129 of Management, Peking University) 31130 </div> 31131 <div class="slot-abstract"> 31132 <div> 31133 <a 31134 class="clickable no-decoration" 31135 id="vhsjs_view_700_1707793552_7798982" 31136 onclick="$('#vhsjs_view_700_1707793552_7798982').hide(); 31137 $('#vhsjs_hide_700_1707793552_7798982').show(); 31138 $('#699_1707793552_7798896').slideDown(function() { 31139 if (typeof Masonry === 'function') { 31140 $('.use_masonry').masonry(); 31141 }; 31142 31143 });" 31144 ><i class="fa fa-caret-right"></i>
31145 <span class="hover_link">Abstract</span></a 31146 ><a 31147 class="clickable no-decoration" 31148 id="vhsjs_hide_700_1707793552_7798982" 31149 onclick="$('#699_1707793552_7798896').hide(function() { 31150 if (typeof Masonry === 'function') { 31151 $('.use_masonry').masonry(); 31152 }; 31153 }); 31154 $('#vhsjs_hide_700_1707793552_7798982').hide(); 31155 $('#vhsjs_view_700_1707793552_7798982').show();" 31156 style="display: none" 31157 ><i class="fa fa-caret-down"></i> 31158 <span class="hover_link">Abstract</span></a 31159 > 31160 <div 31161 data-display-control="700_1707793552_7798982" 31162 id="699_1707793552_7798896" 31163 style="display: none" 31164 > 31165 <div class="arrow-slidedown"> 31166 <blockquote> 31167 We consider a contextual ranking and selection 31168 problem which aims to identify the 31169 best-performing alternative for each context. 31170 The performance is measured by an arbitrary 31171 identifiable statistical characteristic. Under a 31172 Bayesian framework, we establish the posterior 31173 large deviation ratios for general adaptive 31174 sampling policies. We propose an efficient 31175 sampling policy based on top-two Thompson 31176 sampling, which is proven to be consistent. 31177 Numerical experiments demonstrate that the 31178 proposed algorithm outperforms existing 31179 algorithms under both Gaussian and non-Gaussian 31180 settings. 31181 </blockquote> 31182 </div> 31183 </div> 31184 </div> 31185 </div> 31186 <div class="slot-urls"></div> 31187 <a href="/wsc23papers/286.pdf" target="_blank">pdf</a 31188 ><br /> 31189 </div> 31190 <div class="slot-entry"> 31191 <a name="con116" tabindex="-1"></a> 31192 <div class="slot-title-line"> 31193 <span class="slot-title">Epsilon Optimal Sampling</span> 31194 </div> 31195 <div class="slot-authors"> 31196 Travis Goodwin (MITRE Corporation), Jie Xu (George Mason 31197 University), Nurcin Celik (University of Miami), and 31198 Chun-Hung Chen (George Mason University) 31199 </div> 31200 <div class="slot-abstract"> 31201 <div> 31202 <a 31203 class="clickable no-decoration" 31204 id="vhsjs_view_702_1707793552_7822347" 31205 onclick="$('#vhsjs_view_702_1707793552_7822347').hide(); 31206 $('#vhsjs_hide_702_1707793552_7822347').show(); 31207 $('#701_1707793552_782226').slideDown(function() { 31208 if (typeof Masonry === 'function') { 31209 $('.use_masonry').masonry(); 31210 }; 31211 31212 });" 31213 ><i class="fa fa-caret-right"></i> 31214 <span class="hover_link">Abstract</span></a 31215 ><a 31216 class="clickable no-decoration" 31217 id="vhsjs_hide_702_1707793552_7822347" 31218 onclick="$('#701_1707793552_782226').hide(function() { 31219 if (typeof Masonry === 'function') { 31220 $('.use_masonry').masonry(); 31221 }; 31222 }); 31223 $('#vhsjs_hide_702_1707793552_7822347').hide(); 31224 $('#vhsjs_view_702_1707793552_7822347').show();" 31225 style="display: none" 31226 ><i class="fa fa-caret-down"></i>
31227 <span class="hover_link">Abstract</span></a 31228 > 31229 <div 31230 data-display-control="702_1707793552_7822347" 31231 id="701_1707793552_782226" 31232 style="display: none" 31233 > 31234 <div class="arrow-slidedown"> 31235 <blockquote> 31236 Epsilon Optimal Sampling (EOS) is a novel 31237 algorithm that seeks to reduce the computational 31238 complexity of selecting the best design using 31239 stochastic simulation. EOS is an Optimal 31240 Computing Budget Allocation (OCBA) type 31241 algorithm that reduces computational complexity 31242 by integrating machine learning (ML) models into 31243 the simulation optimization algorithm. EOS 31244 avoids the pitfall of trading computational 31245 overhead in simulation execution for 31246 computational overhead in ML model training by 31247 using a concept we call policy stability. In 31248 this paper, we present the concept of policy 31249 stability, how it can be used to improve dynamic 31250 sampling techniques, and how low-fidelity ML 31251 estimates can be integrated into the process. 31252 Numerical results are presented to provide 31253 evidence as to the improvement in computational 31254 efficiency that can be achieved when using EOS 31255 in conjunction with ML models over the standard 31256 OCBA algorithm. 31257 </blockquote> 31258 </div> 31259 </div> 31260 </div> 31261 </div> 31262 <div class="slot-urls"></div> 31263 <a href="/wsc23papers/287.pdf" target="_blank">pdf</a 31264 ><br /> 31265 </div> 31266 <div class="slot-entry"> 31267 <a name="con354" tabindex="-1"></a> 31268 <div class="slot-title-line"> 31269 <span class="slot-title" 31270 >Adaptive Ranking and Selection Based Genetic 31271 Algorithms for Data-driven Problems</span 31272 > 31273 </div> 31274 <div class="slot-authors"> 31275 Kimia Vahdat and Sara Shashaani (North Carolina State 31276 University) 31277 </div> 31278 <div class="slot-abstract"> 31279 <div> 31280 <a 31281 class="clickable no-decoration" 31282 id="vhsjs_view_704_1707793552_7844226" 31283 onclick="$('#vhsjs_view_704_1707793552_7844226').hide(); 31284 $('#vhsjs_hide_704_1707793552_7844226').show(); 31285 $('#703_1707793552_7844145').slideDown(function() { 31286 if (typeof Masonry === 'function') { 31287 $('.use_masonry').masonry(); 31288 }; 31289 31290 });" 31291 ><i class="fa fa-caret-right"></i> 31292 <span class="hover_link">Abstract</span></a 31293 ><a 31294 class="clickable no-decoration" 31295 id="vhsjs_hide_704_1707793552_7844226" 31296 onclick="$('#703_1707793552_7844145').hide(function() { 31297 if (typeof Masonry === 'function') { 31298 $('.use_masonry').masonry(); 31299 }; 31300 }); 31301 $('#vhsjs_hide_704_1707793552_7844226').hide(); 31302 $('#vhsjs_view_704_1707793552_7844226').show();" 31303 style="display: none" 31304 ><i class="fa fa-caret-down"></i>
31305 <span class="hover_link">Abstract</span></a 31306 > 31307 <div 31308 data-display-control="704_1707793552_7844226" 31309 id="703_1707793552_7844145" 31310 style="display: none" 31311 > 31312 <div class="arrow-slidedown"> 31313 <blockquote> 31314 We present ARGA–Adaptive Robust Genetic 31315 Algorithm–to optimize zero-one simulation 31316 problems by incorporating input uncertainty. In 31317 ARGA, a surviving population of solutions 31318 evolves as more information about the 31319 high-dimensional problem affected by 31320 stochasticity becomes available. A ranking and 31321 selection operation in each iteration is 31322 enhanced with a debiasing mechanism of fitness 31323 values using fast iterated bootstraps and 31324 control variates. Debiasing reduces the model 31325 risk from input uncertainty bias, obtaining a 31326 more accurate ranking of the current surviving 31327 solutions. Given the double loop of function 31328 evaluations, we adaptively increase budget only 31329 if the current population’s proximity to 31330 optimality signals the need for a smaller 31331 standard error. In that case, we allocate 31332 additional replications to the input model of a 31333 current surviving solution that is most 31334 responsible for risk. The empirical results with 31335 a fixed optimization budget demonstrate that 31336 ARGA obtains significantly better solutions in a 31337 feature selection problem on various datasets. 31338 </blockquote> 31339 </div> 31340 </div> 31341 </div> 31342 </div> 31343 <div class="slot-urls"></div> 31344 <a href="/wsc23papers/288.pdf" target="_blank">pdf</a 31345 ><br /> 31346 </div> 31347 </div> 31348 <div class="session-entry"> 31349 <span class="session-event-type">Technical Session</span 31350 ><span class="type-track-spacer"> · </span 31351 ><span class="program-track">Simulation Optimization</span 31352 ><br /> 31353 <div class="session-title">Sampling in Optimization</div> 31354 <div class="session-chair"> 31355 Chair: Yunsoo Ha (North Carolina State University)<br /> 31356 </div> 31357 <div class="slot-entry"> 31358 <a name="con327" tabindex="-1"></a> 31359 <div class="slot-title-line"> 31360 <span class="slot-title" 31361 >Parameter Optimization with Conscious Allocation 31362 (POCA)</span 31363 > 31364 </div> 31365 <div class="slot-authors"> 31366 Joshua Inman, Tanmay Khandait, Giulia Pedrielli, and 31367 Lalitha Sankar (Arizona State University) 31368 </div> 31369 <div class="slot-abstract"> 31370 <div> 31371 <a 31372 class="clickable no-decoration" 31373 id="vhsjs_view_706_1707793552_790435" 31374 onclick="$('#vhsjs_view_706_1707793552_790435').hide(); 31375 $('#vhsjs_hide_706_1707793552_790435').show(); 31376 $('#705_1707793552_790427').slideDown(function() { 31377 if (typeof Masonry === 'function') { 31378 $('.use_masonry').masonry(); 31379 }; 31380 31381 });" 31382 ><i class="fa fa-caret-right"></i>
31383 <span class="hover_link">Abstract</span></a 31384 ><a 31385 class="clickable no-decoration" 31386 id="vhsjs_hide_706_1707793552_790435" 31387 onclick="$('#705_1707793552_790427').hide(function() { 31388 if (typeof Masonry === 'function') { 31389 $('.use_masonry').masonry(); 31390 }; 31391 }); 31392 $('#vhsjs_hide_706_1707793552_790435').hide(); 31393 $('#vhsjs_view_706_1707793552_790435').show();" 31394 style="display: none" 31395 ><i class="fa fa-caret-down"></i> 31396 <span class="hover_link">Abstract</span></a 31397 > 31398 <div 31399 data-display-control="706_1707793552_790435" 31400 id="705_1707793552_790427" 31401 style="display: none" 31402 > 31403 <div class="arrow-slidedown"> 31404 <blockquote> 31405 The performance of modern machine learning 31406 algorithms depends upon the selection of a set 31407 of hyperparameters. Common examples of 31408 hyperparameters are learning rate and the number 31409 of layers in a dense neural network. Auto-ML is 31410 a branch of optimization that has produced 31411 important contributions in this area. Within 31412 Auto-ML, hyperband-based approaches, which 31413 eliminate poorly-performing configurations after 31414 evaluating them at low budgets, are among the 31415 most effective. However, the performance of 31416 these algorithms strongly depends on how 31417 effectively they allocate the computational 31418 budget to various hyperparameter configurations. 31419 We present the new Parameter Optimization with 31420 Conscious Allocation (POCA), a hyperband-based 31421 algorithm that adaptively allocates the inputted 31422 budget to the hyperparameter configurations it 31423 generates following a Bayesian sampling scheme. 31424 We compare POCA to its nearest competitor at 31425 optimizing the hyperparameters of an artificial 31426 toy function and a deep neural network and find 31427 that POCA finds strong configurations faster in 31428 both settings. 31429 </blockquote> 31430 </div> 31431 </div> 31432 </div> 31433 </div> 31434 <div class="slot-urls"></div> 31435 <a href="/wsc23papers/289.pdf" target="_blank">pdf</a 31436 ><br /> 31437 </div> 31438 <div class="slot-entry"> 31439 <a name="con129" tabindex="-1"></a> 31440 <div class="slot-title-line"> 31441 <span class="slot-title" 31442 >Cluster-based Sampling Allocation for Multi-fidelity 31443 Simulation Optimization</span 31444 > 31445 </div> 31446 <div class="slot-authors"> 31447 Zirui Cao (National University of Singapore); Haowei 31448 Wang (Rice-Rick Digitalization PTE. Ltd.); and Haobin 31449 Li, Ek Peng Chew, and Kok Choon Tan (National University 31450 of Singapore) 31451 </div> 31452 <div class="slot-abstract"> 31453 <div> 31454 <a 31455 class="clickable no-decoration" 31456 id="vhsjs_view_708_1707793552_7927647" 31457 onclick="$('#vhsjs_view_708_1707793552_7927647').hide(); 31458 $('#vhsjs_hide_708_1707793552_7927647').show(); 31459 $('#707_1707793552_7927566').slideDown(function() { 31460 if (typeof Masonry === 'function') { 31461 $('.use_masonry').masonry(); 31462 }; 31463 31464 });" 31465 ><i class="fa fa-caret-right"></i>
31466 <span class="hover_link">Abstract</span></a 31467 ><a 31468 class="clickable no-decoration" 31469 id="vhsjs_hide_708_1707793552_7927647" 31470 onclick="$('#707_1707793552_7927566').hide(function() { 31471 if (typeof Masonry === 'function') { 31472 $('.use_masonry').masonry(); 31473 }; 31474 }); 31475 $('#vhsjs_hide_708_1707793552_7927647').hide(); 31476 $('#vhsjs_view_708_1707793552_7927647').show();" 31477 style="display: none" 31478 ><i class="fa fa-caret-down"></i> 31479 <span class="hover_link">Abstract</span></a 31480 > 31481 <div 31482 data-display-control="708_1707793552_7927647" 31483 id="707_1707793552_7927566" 31484 style="display: none" 31485 > 31486 <div class="arrow-slidedown"> 31487 <blockquote> 31488 Simulation optimization is widely used to 31489 optimize complex systems. High-fidelity 31490 simulation can be expensive, especially when the 31491 number of designs is large. In practice, fast 31492 but less accurate low-fidelity simulation is 31493 often available and can provide valuable 31494 information. In this paper, we propose a 31495 sampling algorithm that utilizes information 31496 from multiple fidelity simulation models to 31497 improve the efficiency of searching for the best 31498 design. A k-means algorithm is introduced to 31499 help capture the performance clustering 31500 phenomenon among designs, and a cluster validity 31501 index is proposed to determine the optimal 31502 number of clusters. The proposed sampling 31503 algorithm can incorporate the information of 31504 performance clusters and approximately minimize 31505 the expected opportunity cost of the selected 31506 best design. Numerical results substantiate the 31507 superior performance of the proposed algorithm. 31508 </blockquote> 31509 </div> 31510 </div> 31511 </div> 31512 </div> 31513 <div class="slot-urls"></div> 31514 <a href="/wsc23papers/290.pdf" target="_blank">pdf</a 31515 ><br /> 31516 </div> 31517 <div class="slot-entry"> 31518 <a name="inv186" tabindex="-1"></a> 31519 <div class="slot-title-line"> 31520 <span class="slot-title" 31521 >Dynamic Stratification and Post-stratified Adaptive 31522 Sampling for Simulation Optimization</span 31523 > 31524 </div> 31525 <div class="slot-authors"> 31526 Pranav Jain and Sara Shashaani (North Carolina State 31527 University) 31528 </div> 31529 <div class="slot-abstract"> 31530 <div> 31531 <a 31532 class="clickable no-decoration" 31533 id="vhsjs_view_710_1707793552_7948525" 31534 onclick="$('#vhsjs_view_710_1707793552_7948525').hide(); 31535 $('#vhsjs_hide_710_1707793552_7948525').show(); 31536 $('#709_1707793552_7948446').slideDown(function() { 31537 if (typeof Masonry === 'function') { 31538 $('.use_masonry').masonry(); 31539 }; 31540 31541 });" 31542 ><i class="fa fa-caret-right"></i>
31543 <span class="hover_link">Abstract</span></a 31544 ><a 31545 class="clickable no-decoration" 31546 id="vhsjs_hide_710_1707793552_7948525" 31547 onclick="$('#709_1707793552_7948446').hide(function() { 31548 if (typeof Masonry === 'function') { 31549 $('.use_masonry').masonry(); 31550 }; 31551 }); 31552 $('#vhsjs_hide_710_1707793552_7948525').hide(); 31553 $('#vhsjs_view_710_1707793552_7948525').show();" 31554 style="display: none" 31555 ><i class="fa fa-caret-down"></i> 31556 <span class="hover_link">Abstract</span></a 31557 > 31558 <div 31559 data-display-control="710_1707793552_7948525" 31560 id="709_1707793552_7948446" 31561 style="display: none" 31562 > 31563 <div class="arrow-slidedown"> 31564 <blockquote> 31565 Post-stratification is a variance reduction 31566 technique that groups samples in respective 31567 strata only after collecting the samples 31568 randomly. We incorporate this technique within 31569 an adaptive sampling procedure in simulation 31570 optimization. We use concomitant variables to 31571 increase the accuracy of our proposed 31572 post-stratified adaptive sampling. Concomitant 31573 variables are auxiliary variables in simulation 31574 that approximate the boundaries of the optimal 31575 strata at each visited solution during the 31576 optimization procedure. A linear relationship 31577 between the concomitant variable and the output 31578 is desirable but not necessary for the 31579 effectiveness of the proposed methodology. In 31580 numerical experiments, we observe that 31581 performing post-stratified adaptive sampling 31582 with dynamically updated strata boundaries 31583 robustifies the algorithm in the sense that it 31584 reduces the algorithm's sensitivity to the 31585 initial solution and solver input parameters. 31586 </blockquote> 31587 </div> 31588 </div> 31589 </div> 31590 </div> 31591 <div class="slot-urls"></div> 31592 <a href="/wsc23papers/291.pdf" target="_blank">pdf</a 31593 ><br /> 31594 </div> 31595 </div> 31596 <div class="session-entry"> 31597 <span class="session-event-type">Technical Session</span 31598 ><span class="type-track-spacer"> · </span 31599 ><span class="program-track">Simulation Optimization</span 31600 ><br /> 31601 <div class="session-title">Gaussian Process Surrogates</div> 31602 <div class="session-chair"> 31603 Chair: Zirui Cao (National University of Singapore)<br /> 31604 </div> 31605 <div class="slot-entry"> 31606 <a name="con184" tabindex="-1"></a> 31607 <div class="slot-title-line"> 31608 <span class="slot-title" 31609 >Simulation Optimization with Multiple Attempts</span 31610 > 31611 </div> 31612 <div class="slot-authors"> 31613 Jingjun Men and Zhihao Liu (Southern University of 31614 Science and Technology), Haowei Wang (Rice-Rick 31615 Digitalization PTE. Ltd.), and Songhao Wang (Southern 31616 University of Science and Technology) 31617 </div> 31618 <div class="slot-abstract"> 31619 <div> 31620 <a 31621 class="clickable no-decoration" 31622 id="vhsjs_view_712_1707793552_800329" 31623 onclick="$('#vhsjs_view_712_1707793552_800329').hide(); 31624 $('#vhsjs_hide_712_1707793552_800329').show(); 31625 $('#711_1707793552_8003206').slideDown(function() { 31626 if (typeof Masonry === 'function') { 31627 $('.use_masonry').masonry(); 31628 }; 31629 31630 });" 31631 ><i class="fa fa-caret-right"></i>
31632 <span class="hover_link">Abstract</span></a 31633 ><a 31634 class="clickable no-decoration" 31635 id="vhsjs_hide_712_1707793552_800329" 31636 onclick="$('#711_1707793552_8003206').hide(function() { 31637 if (typeof Masonry === 'function') { 31638 $('.use_masonry').masonry(); 31639 }; 31640 }); 31641 $('#vhsjs_hide_712_1707793552_800329').hide(); 31642 $('#vhsjs_view_712_1707793552_800329').show();" 31643 style="display: none" 31644 ><i class="fa fa-caret-down"></i> 31645 <span class="hover_link">Abstract</span></a 31646 > 31647 <div 31648 data-display-control="712_1707793552_800329" 31649 id="711_1707793552_8003206" 31650 style="display: none" 31651 > 31652 <div class="arrow-slidedown"> 31653 <blockquote> 31654 Simulation optimization is a widely utilized 31655 approach that allows decision-makers to test 31656 various decision variable settings in simulators 31657 before implementing a final recommended action 31658 on the real systems. In some real-world 31659 scenarios, the recommended action can be 31660 executed multiple times and the performance is 31661 evaluated as the best one among these multiple 31662 attempts. In this paper, we introduce such 31663 simulation optimization problem with multiple 31664 attempts and provide insights of the problem 31665 through comparison to risk-averse decision 31666 making problem. We propose a surrogate-assisted 31667 algorithm based on the Gaussian process model 31668 and the upper confidence bound criterion for 31669 efficiently solving such problems. We 31670 demonstrate the efficiency and effectiveness of 31671 the proposed approach with several numerical 31672 examples. 31673 </blockquote> 31674 </div> 31675 </div> 31676 </div> 31677 </div> 31678 <div class="slot-urls"></div> 31679 <a href="/wsc23papers/292.pdf" target="_blank">pdf</a 31680 ><br /> 31681 </div> 31682 <div class="slot-entry"> 31683 <a name="inv153" tabindex="-1"></a> 31684 <div class="slot-title-line"> 31685 <span class="slot-title" 31686 >Hyperparameter Adaptive Search for Surrogate 31687 Optimization: A Self-Adjusting Approach</span 31688 > 31689 </div> 31690 <div class="slot-authors"> 31691 Nazanin Nezami and Hadis Anahideh (University of 31692 Illinois Chicago) 31693 </div> 31694 <div class="slot-abstract"> 31695 <div> 31696 <a 31697 class="clickable no-decoration" 31698 id="vhsjs_view_714_1707793552_8026462" 31699 onclick="$('#vhsjs_view_714_1707793552_8026462').hide(); 31700 $('#vhsjs_hide_714_1707793552_8026462').show(); 31701 $('#713_1707793552_8026383').slideDown(function() { 31702 if (typeof Masonry === 'function') { 31703 $('.use_masonry').masonry(); 31704 }; 31705 31706 });" 31707 ><i class="fa fa-caret-right"></i>
31708 <span class="hover_link">Abstract</span></a 31709 ><a 31710 class="clickable no-decoration" 31711 id="vhsjs_hide_714_1707793552_8026462" 31712 onclick="$('#713_1707793552_8026383').hide(function() { 31713 if (typeof Masonry === 'function') { 31714 $('.use_masonry').masonry(); 31715 }; 31716 }); 31717 $('#vhsjs_hide_714_1707793552_8026462').hide(); 31718 $('#vhsjs_view_714_1707793552_8026462').show();" 31719 style="display: none" 31720 ><i class="fa fa-caret-down"></i> 31721 <span class="hover_link">Abstract</span></a 31722 > 31723 <div 31724 data-display-control="714_1707793552_8026462" 31725 id="713_1707793552_8026383" 31726 style="display: none" 31727 > 31728 <div class="arrow-slidedown"> 31729 <blockquote> 31730 Surrogate Optimization (SO) algorithms have 31731 shown promise for optimizing expensive black-box 31732 functions. However, their performance is heavily 31733 influenced by hyperparameters related to 31734 sampling and surrogate fitting, which poses a 31735 challenge to their widespread adoption. We 31736 investigate the impact of hyperparameters on 31737 various SO algorithms and propose a 31738 Hyperparameter Adaptive Search for SO (HASSO) 31739 approach. HASSO is not a hyperparameter tuning 31740 algorithm, but a generic self-adjusting SO 31741 algorithm that dynamically tunes its own 31742 hyperparameters while concurrently optimizing 31743 the primary objective function, without 31744 requiring additional evaluations. The aim is to 31745 improve the accessibility, effectiveness, and 31746 convergence speed of SO algorithms for 31747 practitioners. Our approach identifies and 31748 modifies the most influential hyperparameters 31749 specific to each problem and SO approach, 31750 reducing the need for manual tuning without 31751 significantly increasing the computational 31752 burden. Experimental results demonstrate the 31753 effectiveness of HASSO in enhancing the 31754 performance of various SO algorithms across 31755 different global optimization test problems. 31756 </blockquote> 31757 </div> 31758 </div> 31759 </div> 31760 </div> 31761 <div class="slot-urls"></div> 31762 <a href="/wsc23papers/293.pdf" target="_blank">pdf</a 31763 ><br /> 31764 </div> 31765 <div class="slot-entry"> 31766 <a name="inv144" tabindex="-1"></a> 31767 <div class="slot-title-line"> 31768 <span class="slot-title" 31769 >Approximate Gaussian Process Regression with Pairwise 31770 Comparison Data</span 31771 > 31772 </div> 31773 <div class="slot-authors"> 31774 Efe Sertkaya and Ilya Ryzhov (University of Maryland) 31775 </div> 31776 <div class="slot-abstract"> 31777 <div> 31778 <a 31779 class="clickable no-decoration" 31780 id="vhsjs_view_716_1707793552_8047354" 31781 onclick="$('#vhsjs_view_716_1707793552_8047354').hide(); 31782 $('#vhsjs_hide_716_1707793552_8047354').show(); 31783 $('#715_1707793552_8047273').slideDown(function() { 31784 if (typeof Masonry === 'function') { 31785 $('.use_masonry').masonry(); 31786 }; 31787 31788 });" 31789 ><i class="fa fa-caret-right"></i>
31790 <span class="hover_link">Abstract</span></a 31791 ><a 31792 class="clickable no-decoration" 31793 id="vhsjs_hide_716_1707793552_8047354" 31794 onclick="$('#715_1707793552_8047273').hide(function() { 31795 if (typeof Masonry === 'function') { 31796 $('.use_masonry').masonry(); 31797 }; 31798 }); 31799 $('#vhsjs_hide_716_1707793552_8047354').hide(); 31800 $('#vhsjs_view_716_1707793552_8047354').show();" 31801 style="display: none" 31802 ><i class="fa fa-caret-down"></i> 31803 <span class="hover_link">Abstract</span></a 31804 > 31805 <div 31806 data-display-control="716_1707793552_8047354" 31807 id="715_1707793552_8047273" 31808 style="display: none" 31809 > 31810 <div class="arrow-slidedown"> 31811 <blockquote> 31812 We use approximate Bayesian inference, together 31813 with Gaussian process regression, to create a 31814 new estimator for an unknown function in a 31815 situation where we can only observe pairwise 31816 comparisons of function values at different 31817 inputs. Preliminary experimental results suggest 31818 that, although information is heavily censored 31819 in this setting, it may still be possible to 31820 learn the local and global minima of the 31821 underlying function. We discuss possible 31822 sampling criteria, and explore the performance 31823 of the "probability of improvement" strategy 31824 numerically. 31825 </blockquote> 31826 </div> 31827 </div> 31828 </div> 31829 </div> 31830 <div class="slot-urls"></div> 31831 <a href="/wsc23papers/294.pdf" target="_blank">pdf</a 31832 ><br /> 31833 </div> 31834 </div> 31835 <div class="session-entry"> 31836 <span class="session-event-type">Technical Session</span 31837 ><span class="type-track-spacer"> · </span 31838 ><span class="program-track">Simulation Optimization</span 31839 ><br /> 31840 <div class="session-title">Continuous Optimization</div> 31841 <div class="session-chair"> 31842 Chair: Meichen Song (Stony Brook University)<br /> 31843 </div> 31844 <div class="slot-entry"> 31845 <a name="con311" tabindex="-1"></a> 31846 <div class="slot-title-line"> 31847 <span class="slot-title" 31848 >Towards Greener Stochastic Derivative-Free 31849 Optimization with Trust Regions and Adaptive 31850 Sampling</span 31851 > 31852 </div> 31853 <div class="slot-authors"> 31854 Yunsoo Ha and Sara Shashaani (North Carolina State 31855 University) 31856 </div> 31857 <div class="slot-abstract"> 31858 <div> 31859 <a 31860 class="clickable no-decoration" 31861 id="vhsjs_view_718_1707793552_8092048" 31862 onclick="$('#vhsjs_view_718_1707793552_8092048').hide(); 31863 $('#vhsjs_hide_718_1707793552_8092048').show(); 31864 $('#717_1707793552_8091962').slideDown(function() { 31865 if (typeof Masonry === 'function') { 31866 $('.use_masonry').masonry(); 31867 }; 31868 31869 });" 31870 ><i class="fa fa-caret-right"></i>
31871 <span class="hover_link">Abstract</span></a 31872 ><a 31873 class="clickable no-decoration" 31874 id="vhsjs_hide_718_1707793552_8092048" 31875 onclick="$('#717_1707793552_8091962').hide(function() { 31876 if (typeof Masonry === 'function') { 31877 $('.use_masonry').masonry(); 31878 }; 31879 }); 31880 $('#vhsjs_hide_718_1707793552_8092048').hide(); 31881 $('#vhsjs_view_718_1707793552_8092048').show();" 31882 style="display: none" 31883 ><i class="fa fa-caret-down"></i> 31884 <span class="hover_link">Abstract</span></a 31885 > 31886 <div 31887 data-display-control="718_1707793552_8092048" 31888 id="717_1707793552_8091962" 31889 style="display: none" 31890 > 31891 <div class="arrow-slidedown"> 31892 <blockquote> 31893 Adaptive sampling-based trust-region 31894 optimization has emerged as an efficient solver 31895 for nonlinear and nonconvex problems in noisy 31896 derivative-free environments. This class of 31897 algorithms proceeds by iteratively constructing 31898 local models on objective function estimates 31899 that use a carefully chosen number of calls to 31900 the stochastic oracle. In this paper, we 31901 introduce a refined version of this class of 31902 algorithms that reuse the information from 31903 previous iterations. The advantage of this 31904 approach is reducing computational burden 31905 without sacrificing consistency or work 31906 complexity to attain the same level of 31907 optimality, which we demonstrate through 31908 numerical results using the SimOpt library. 31909 </blockquote> 31910 </div> 31911 </div> 31912 </div> 31913 </div> 31914 <div class="slot-urls"></div> 31915 <a href="/wsc23papers/295.pdf" target="_blank">pdf</a 31916 ><br /> 31917 </div> 31918 <div class="slot-entry"> 31919 <a name="inv128" tabindex="-1"></a> 31920 <div class="slot-title-line"> 31921 <span class="slot-title" 31922 >Stochastic Adaptive Regularization Method with 31923 Cubics: A High Probability Complexity Bound</span 31924 > 31925 </div> 31926 <div class="slot-authors"> 31927 Katya Scheinberg and Miaolan Xie (Cornell University) 31928 </div> 31929 <div class="slot-abstract"> 31930 <div> 31931 <a 31932 class="clickable no-decoration" 31933 id="vhsjs_view_720_1707793552_8114111" 31934 onclick="$('#vhsjs_view_720_1707793552_8114111').hide(); 31935 $('#vhsjs_hide_720_1707793552_8114111').show(); 31936 $('#719_1707793552_8114033').slideDown(function() { 31937 if (typeof Masonry === 'function') { 31938 $('.use_masonry').masonry(); 31939 }; 31940 31941 });" 31942 ><i class="fa fa-caret-right"></i> 31943 <span class="hover_link">Abstract</span></a 31944 ><a 31945 class="clickable no-decoration" 31946 id="vhsjs_hide_720_1707793552_8114111" 31947 onclick="$('#719_1707793552_8114033').hide(function() { 31948 if (typeof Masonry === 'function') { 31949 $('.use_masonry').masonry(); 31950 }; 31951 }); 31952 $('#vhsjs_hide_720_1707793552_8114111').hide(); 31953 $('#vhsjs_view_720_1707793552_8114111').show();" 31954 style="display: none" 31955 ><i class="fa fa-caret-down"></i>
31956 <span class="hover_link">Abstract</span></a 31957 > 31958 <div 31959 data-display-control="720_1707793552_8114111" 31960 id="719_1707793552_8114033" 31961 style="display: none" 31962 > 31963 <div class="arrow-slidedown"> 31964 <blockquote> 31965 We present a high probability complexity bound 31966 for a stochastic adaptive regularization method 31967 with cubics, also known as regularized Newton 31968 method. The method makes use of stochastic 31969 zeroth-, first- and second-order oracles that 31970 satisfy certain accuracy and reliability 31971 assumptions. Such oracles have been used in the 31972 literature by other stochastic adaptive methods, 31973 such as trust region and line search. These 31974 oracles capture many settings, such as expected 31975 risk minimization, stochastic zeroth-order 31976 optimization, and others. In this paper, we give 31977 the first high probability iteration bound for 31978 stochastic cubic regularization, and show that 31979 just as in the deterministic case, it is 31980 superior to other stochastic adaptive methods. 31981 </blockquote> 31982 </div> 31983 </div> 31984 </div> 31985 </div> 31986 <div class="slot-urls"></div> 31987 <a href="/wsc23papers/296.pdf" target="_blank">pdf</a 31988 ><br /> 31989 </div> 31990 <div class="slot-entry"> 31991 <a name="inv135" tabindex="-1"></a> 31992 <div class="slot-title-line"> 31993 <span class="slot-title" 31994 >A Projection-Based Algorithm for Solving Stochastic 31995 Inverse Variational Inequality Problems</span 31996 > 31997 </div> 31998 <div class="slot-authors"> 31999 Zeinab Alizadeh, Felipe Parra Polanco, and Afrooz 32000 Jalilzadeh (The University of Arizona) 32001 </div> 32002 <div class="slot-abstract"> 32003 <div> 32004 <a 32005 class="clickable no-decoration" 32006 id="vhsjs_view_722_1707793552_8135428" 32007 onclick="$('#vhsjs_view_722_1707793552_8135428').hide(); 32008 $('#vhsjs_hide_722_1707793552_8135428').show(); 32009 $('#721_1707793552_813535').slideDown(function() { 32010 if (typeof Masonry === 'function') { 32011 $('.use_masonry').masonry(); 32012 }; 32013 32014 });" 32015 ><i class="fa fa-caret-right"></i> 32016 <span class="hover_link">Abstract</span></a 32017 ><a 32018 class="clickable no-decoration" 32019 id="vhsjs_hide_722_1707793552_8135428" 32020 onclick="$('#721_1707793552_813535').hide(function() { 32021 if (typeof Masonry === 'function') { 32022 $('.use_masonry').masonry(); 32023 }; 32024 }); 32025 $('#vhsjs_hide_722_1707793552_8135428').hide(); 32026 $('#vhsjs_view_722_1707793552_8135428').show();" 32027 style="display: none" 32028 ><i class="fa fa-caret-down"></i> 32029 <span class="hover_link">Abstract</span></a 32030 > 32031 <div
32032 data-display-control="722_1707793552_8135428" 32033 id="721_1707793552_813535" 32034 style="display: none" 32035 > 32036 <div class="arrow-slidedown"> 32037 <blockquote> 32038 We consider a stochastic Inverse Variational 32039 Inequality (IVI) problem defined by a continuous 32040 and co-coercive map over a closed and convex 32041 set. Motivated by the absence of performance 32042 guarantees for stochastic IVI, we present a 32043 variance-reduced projection-based gradient 32044 method. Our proposed method ensures an almost 32045 sure convergence of the generated iterates to 32046 the solution, and we establish a convergence 32047 rate guarantee. To verify our results, we apply 32048 the proposed algorithm to a network equilibrium 32049 control problem. 32050 </blockquote> 32051 </div> 32052 </div> 32053 </div> 32054 </div> 32055 <div class="slot-urls"></div> 32056 <a href="/wsc23papers/297.pdf" target="_blank">pdf</a 32057 ><br /> 32058 </div> 32059 </div> 32060 <div class="session-entry"> 32061 <span class="session-event-type">Technical Session</span 32062 ><span class="type-track-spacer"> · </span 32063 ><span class="program-track">Simulation Optimization</span 32064 ><br /> 32065 <div class="session-title">Learning for Optimization</div> 32066 <div class="session-chair"> 32067 Chair: Peter J Haas (University of Massachusetts 32068 Amherst)<br /> 32069 </div> 32070 <div class="slot-entry"> 32071 <a name="inv169" tabindex="-1"></a> 32072 <div class="slot-title-line"> 32073 <span class="slot-title" 32074 >Efficient Hybrid Simulation Optimization via Graph 32075 Neural Network Metamodeling</span 32076 > 32077 </div> 32078 <div class="slot-authors"> 32079 Wang Cen and Peter Haas (University of Massachusetts 32080 Amherst) 32081 </div> 32082 <div class="slot-abstract"> 32083 <div> 32084 <a 32085 class="clickable no-decoration" 32086 id="vhsjs_view_724_1707793552_8181288" 32087 onclick="$('#vhsjs_view_724_1707793552_8181288').hide(); 32088 $('#vhsjs_hide_724_1707793552_8181288').show(); 32089 $('#723_1707793552_8181207').slideDown(function() { 32090 if (typeof Masonry === 'function') { 32091 $('.use_masonry').masonry(); 32092 }; 32093 32094 });" 32095 ><i class="fa fa-caret-right"></i> 32096 <span class="hover_link">Abstract</span></a 32097 ><a 32098 class="clickable no-decoration" 32099 id="vhsjs_hide_724_1707793552_8181288" 32100 onclick="$('#723_1707793552_8181207').hide(function() { 32101 if (typeof Masonry === 'function') { 32102 $('.use_masonry').masonry(); 32103 }; 32104 }); 32105 $('#vhsjs_hide_724_1707793552_8181288').hide(); 32106 $('#vhsjs_view_724_1707793552_8181288').show();" 32107 style="display: none" 32108 ><i class="fa fa-caret-down"></i>
32109 <span class="hover_link">Abstract</span></a 32110 > 32111 <div 32112 data-display-control="724_1707793552_8181288" 32113 id="723_1707793552_8181207" 32114 style="display: none" 32115 > 32116 <div class="arrow-slidedown"> 32117 <blockquote> 32118 Simulation metamodeling is essential for 32119 speeding up optimization via simulation to 32120 support rapid decision making. During 32121 optimization, the metamodel, rather than 32122 expensive simulation, is used to compute 32123 objective values. We recently developed 32124 graphical neural metamodels (GMMs) that use 32125 graph neural networks to allow the graphical 32126 structure of a simulation model to be treated as 32127 a metamodel input parameter that can be varied 32128 along with scalar inputs. In this paper we 32129 provide novel methods for using GMMs to solve 32130 hybrid optimization problems where both 32131 real-valued input parameters and graphical 32132 structure are jointly optimized. The key ideas 32133 are to modify Monte Carlo tree search to 32134 incorporate both discrete and continuous 32135 optimization and to leverage the automatic 32136 differentiation infrastructure used for neural 32137 network training to quickly compute gradients of 32138 the objective function during stochastic 32139 gradient descent. Experiments on stoch
32139astic 32140 activity network and warehouse models 32141 demonstrate the potential of our method. 32142 </blockquote> 32143 </div> 32144 </div> 32145 </div> 32146 </div> 32147 <div class="slot-urls"></div> 32148 <a href="/wsc23papers/298.pdf" target="_blank">pdf</a 32149 ><br /> 32150 </div> 32151 <div class="slot-entry"> 32152 <a name="inv180" tabindex="-1"></a> 32153 <div class="slot-title-line"> 32154 <span class="slot-title" 32155 >Policy-Augmented Bayesian Network Optimization with 32156 Global Convergence</span 32157 > 32158 </div> 32159 <div class="slot-authors"> 32160 Junkai Zhao (Shanghai Jiao Tong University), Wei Xie 32161 (Northeastern University), and Jun Luo (Shanghai Jiao 32162 Tong University) 32163 </div> 32164 <div class="slot-abstract"> 32165 <div> 32166 <a 32167 class="clickable no-decoration" 32168 id="vhsjs_view_726_1707793552_8203835" 32169 onclick="$('#vhsjs_view_726_1707793552_8203835').hide(); 32170 $('#vhsjs_hide_726_1707793552_8203835').show(); 32171 $('#725_1707793552_8203752').slideDown(function() { 32172 if (typeof Masonry === 'function') { 32173 $('.use_masonry').masonry(); 32174 }; 32175 32176 });" 32177 ><i class="fa fa-caret-right"></i> 32178 <span class="hover_link">Abstract</span></a 32179 ><a 32180 class="clickable no-decoration" 32181 id="vhsjs_hide_726_1707793552_8203835" 32182 onclick="$('#725_1707793552_8203752').hide(function() { 32183 if (typeof Masonry === 'function') { 32184 $('.use_masonry').masonry(); 32185 }; 32186 }); 32187 $('#vhsjs_hide_726_1707793552_8203835').hide(); 32188 $('#vhsjs_view_726_1707793552_8203835').show();" 32189 style="display: none" 32190 ><i class="fa fa-caret-down"></i> 32191 <span class="hover_link">Abstract</span></a 32192 > 32193 <div 32194 data-display-control="726_1707793552_8203835" 32195 id="725_1707793552_8203752" 32196 style="display: none" 32197 > 32198 <div class="arrow-slidedown"> 32199 <blockquote> 32200 Driven by critical challenges in 32201 biomanufacturing, including high complexity and 32202 high uncertainty, we propose global optimization 32203 methods on the policy-augmented Bayesian network 32204 (PABN), characterizing risk- and science-based 32205 understanding of underlying bioprocess 32206 mechanisms, to guide the optimal control. We 32207 first develop a sequential optimization 32208 algorithm based on deep kernel learning (DKL) 32209 for PABN with general state transition dynamics, 32210 which can learn the spatial dependence of mean 32211 response through a deep neural network. In 32212 addition, to improve the interpretability and 32213 computational efficiency of policy optimization, 32214 a global metamodel is introduced to guide linear 32215 Gaussian PABN optimization, which explicitly 32216 accounts for the correlation of input-to-output 32217 pathways obtained under different candidate 32218 policies. Our empirical study provides the 32219 ablation analysis and the interpretation 32220 analysis of the DKL, and also shows that both 32221 proposed approaches demonstrate promising 32222 performance compared to the standard Bayesian 32223 optimization with Gaussian process. 32224 </blockquote> 32225 </div> 32226 </div> 32227 </div> 32228 </div> 32229 <div class="slot-urls"></div> 32230 <a href="/wsc23papers/299.pdf" target="_blank">pdf</a 32231 ><br /> 32232 </div> 32233 <div class="slot-entry"> 32234 <a name="con173" tabindex="-1"></a> 32235 <div class="slot-title-line">
32236 <span class="slot-title" 32237 >Simultaneous Perturbation-Based Stochastic 32238 Approximation for Quantile Optimization</span 32239 > 32240 </div> 32241 <div> 32242 <span class="BTP award" 32243 >Best Contributed Theoretical Paper - Finalist</span 32244 > 32245 </div> 32246 <div class="slot-authors"> 32247 Meichen Song and Jiaqiao Hu (Stony Brook University) and 32248 Michael C. Fu (University of Maryland, College Park) 32249 </div> 32250 <div class="slot-abstract"> 32251 <div> 32252 <a 32253 class="clickable no-decoration" 32254 id="vhsjs_view_728_1707793552_8226922" 32255 onclick="$('#vhsjs_view_728_1707793552_8226922').hide(); 32256 $('#vhsjs_hide_728_1707793552_8226922').show(); 32257 $('#727_1707793552_822684').slideDown(function() { 32258 if (typeof Masonry === 'function') { 32259 $('.use_masonry').masonry(); 32260 }; 32261 32262 });" 32263 ><i class="fa fa-caret-right"></i> 32264 <span class="hover_link">Abstract</span></a 32265 ><a 32266 class="clickable no-decoration" 32267 id="vhsjs_hide_728_1707793552_8226922" 32268 onclick="$('#727_1707793552_822684').hide(function() { 32269 if (typeof Masonry === 'function') { 32270 $('.use_masonry').masonry(); 32271 }; 32272 }); 32273 $('#vhsjs_hide_728_1707793552_8226922').hide(); 32274 $('#vhsjs_view_728_1707793552_8226922').show();" 32275 style="display: none" 32276 ><i class="fa fa-caret-down"></i> 32277 <span class="hover_link">Abstract</span></a 32278 > 32279 <div 32280 data-display-control="728_1707793552_8226922" 32281 id="727_1707793552_822684" 32282 style="display: none" 32283 > 32284 <div class="arrow-slidedown"> 32285 <blockquote> 32286 We study a gradient-based algorithm for solving 32287 differentiable quantile optimization problems 32288 under a black-box scenario. The algorithm finds 32289 improved solutions along the descent direction 32290 of the quantile objective function, which is 32291 approximated at each step using a simultaneous 32292 perturbation technique that involves the 32293 difference quotient of the output random 32294 variables. Compared to existing quantile 32295 optimization methods, our algorithm has a 32296 two-timescale stochastic approximation structure 32297 and uses only three observations of the output 32298 random variable per iteration without requiring 32299 knowledge of the underlying system model. We 32300 show the local convergence of the algorithm and 32301 establish a finite-time bound on the convergence 32302 rate of the algorithm. Numerical results are 32303 also presented to illustrate the algorithm. 32304 </blockquote> 32305 </div> 32306 </div> 32307 </div> 32308 </div> 32309 <div class="slot-urls"></div> 32310 <a href="/wsc23papers/300.pdf" target="_blank">pdf</a 32311 ><br /> 32312 </div> 32313 </div> 32314 <div class="session-entry">
32315 <span class="session-event-type">Technical Session</span 32316 ><span class="type-track-spacer"> · </span 32317 ><span class="program-track">Simulation Optimization</span 32318 ><br /> 32319 <div class="session-title"> 32320 Performance Indicators and Matrix Approximation 32321 </div> 32322 <div class="session-chair"> 32323 Chair: Sara Shashaani (North Carolina State University)<br /> 32324 </div> 32325 <div class="slot-entry"> 32326 <a name="inv191" tabindex="-1"></a> 32327 <div class="slot-title-line"> 32328 <span class="slot-title" 32329 >Properties of Several Performance Indicators for 32330 Global Multi-Objective Simulation Optimization</span 32331 > 32332 </div> 32333 <div class="slot-authors"> 32334 Susan R. Hunter and Burla E. Ondes (Purdue University) 32335 </div> 32336 <div class="slot-abstract"> 32337 <div> 32338 <a 32339 class="clickable no-decoration" 32340 id="vhsjs_view_730_1707793552_8271694" 32341 onclick="$('#vhsjs_view_730_1707793552_8271694').hide(); 32342 $('#vhsjs_hide_730_1707793552_8271694').show(); 32343 $('#729_1707793552_827161').slideDown(function() { 32344 if (typeof Masonry === 'function') { 32345 $('.use_masonry').masonry(); 32346 }; 32347 32348 });" 32349 ><i class="fa fa-caret-right"></i> 32350 <span class="hover_link">Abstract</span></a 32351 ><a 32352 class="clickable no-decoration" 32353 id="vhsjs_hide_730_1707793552_8271694" 32354 onclick="$('#729_1707793552_827161').hide(function() { 32355 if (typeof Masonry === 'function') { 32356 $('.use_masonry').masonry(); 32357 }; 32358 }); 32359 $('#vhsjs_hide_730_1707793552_8271694').hide(); 32360 $('#vhsjs_view_730_1707793552_8271694').show();" 32361 style="display: none" 32362 ><i class="fa fa-caret-down"></i> 32363 <span class="hover_link">Abstract</span></a 32364 > 32365 <div 32366 data-display-control="730_1707793552_8271694" 32367 id="729_1707793552_827161" 32368 style="display: none" 32369 > 32370 <div class="arrow-slidedown"> 32371 <blockquote> 32372 We discuss the challenges in constructing and 32373 analyzing performance indicators for 32374 multi-objective simulation optimization (MOSO), 32375 and we examine properties of several performance 32376 indicators for assessing algorithms designed to 32377 solve MOSO problems to global optimality. Our 32378 main contribution lies in the definition and 32379 analysis of a modified coverage error; the 32380 modification to the coverage error enables us to 32381 obtain an upper bound that is the sum of 32382 deterministic and stochastic error terms. Then, 32383 we analyze each error term separately to obtain 32384 an overall upper bound on the modified coverage 32385 error that is a function of the dispersion of 32386 the visited points in the compact feasible set 32387 and the sampling error of the objective function 32388 values at the visited points. The upper bound
32389 provides a foundation for future mathematical 32390 analyses that characterize the rate of decay of 32391 the modified coverage error. 32392 </blockquote> 32393 </div> 32394 </div> 32395 </div> 32396 </div> 32397 <div class="slot-urls"></div> 32398 <a href="/wsc23papers/301.pdf" target="_blank">pdf</a 32399 ><br /> 32400 </div> 32401 <div class="slot-entry"> 32402 <a name="inv158" tabindex="-1"></a> 32403 <div class="slot-title-line"> 32404 <span class="slot-title" 32405 >Stochastic Constraints: How Feasible is 32406 Feasible?</span 32407 > 32408 </div> 32409 <div class="slot-authors"> 32410 David Eckman (Texas A&M University), Shane Henderson 32411 (Cornell University), and Sara Shashaani (North Carolina 32412 State University) 32413 </div> 32414 <div class="slot-abstract"> 32415 <div> 32416 <a 32417 class="clickable no-decoration" 32418 id="vhsjs_view_732_1707793552_8302562" 32419 onclick="$('#vhsjs_view_732_1707793552_8302562').hide(); 32420 $('#vhsjs_hide_732_1707793552_8302562').show(); 32421 $('#731_1707793552_8302476').slideDown(function() { 32422 if (typeof Masonry === 'function') { 32423 $('.use_masonry').masonry(); 32424 }; 32425 32426 });" 32427 ><i class="fa fa-caret-right"></i> 32428 <span class="hover_link">Abstract</span></a 32429 ><a 32430 class="clickable no-decoration" 32431 id="vhsjs_hide_732_1707793552_8302562" 32432 onclick="$('#731_1707793552_8302476').hide(function() { 32433 if (typeof Masonry === 'function') { 32434 $('.use_masonry').masonry(); 32435 }; 32436 }); 32437 $('#vhsjs_hide_732_1707793552_8302562').hide(); 32438 $('#vhsjs_view_732_1707793552_8302562').show();" 32439 style="display: none" 32440 ><i class="fa fa-caret-down"></i> 32441 <span class="hover_link">Abstract</span></a 32442 > 32443 <div 32444 data-display-control="732_1707793552_8302562" 32445 id="731_1707793552_8302476" 32446 style="display: none" 32447 > 32448 <div class="arrow-slidedown"> 32449 <blockquote> 32450 Stochastic constraints, which constrain an 32451 expectation in the context of simulation 32452 optimization, can be hard to conceptualize and 32453 harder still to assess. As with a deterministic 32454 constraint, a solution is considered either 32455 feasible or infeasible with respect to a 32456 stochastic constraint. This perspective belies 32457 the subjective nature of stochastic constraints, 32458 which often arise when attempting to avoid 32459 alternative optimization formulations with 32460 multiple objectives or an aggregate objective 32461 with weights. Moreover, a solution's feasibility 32462 with respect to a stochastic constraint cannot, 32463 in general, be ascertained based on only a 32464 finite number of simulation replications. We 32465 introduce different means of estimating how 32466 "close" the expected performance of a given 32467 solution is to being feasible with respect to 32468 one or more stochastic constraints. We explore 32469 how these metrics and their bootstrapped error 32470 estimates can be incorporated into plots showing 32471 a solver's progress over time when solving a 32472 stochastically constrained problem. 32473 </blockquote> 32474 </div> 32475 </div> 32476 </div> 32477 </div> 32478 <div class="slot-urls"></div> 32479 <a href="/wsc23papers/302.pdf" target="_blank">pdf</a 32480 ><br /> 32481 </div> 32482 <div class="slot-entry"> 32483 <a name="con180" tabindex="-1"></a> 32484 <div class="slot-title-line">
32485 <span class="slot-title" 32486 >Column Subset Selection and Nyström 32487 Approximation via Continuous Optimization</span 32488 > 32489 </div> 32490 <div class="slot-authors"> 32491 Anant Mathur, Sarat Moka, and Zdravko Botev (UNSW) 32492 </div> 32493 <div class="slot-abstract"> 32494 <div> 32495 <a 32496 class="clickable no-decoration" 32497 id="vhsjs_view_734_1707793552_8324573" 32498 onclick="$('#vhsjs_view_734_1707793552_8324573').hide(); 32499 $('#vhsjs_hide_734_1707793552_8324573').show(); 32500 $('#733_1707793552_8324487').slideDown(function() { 32501 if (typeof Masonry === 'function') { 32502 $('.use_masonry').masonry(); 32503 }; 32504 32505 });" 32506 ><i class="fa fa-caret-right"></i> 32507 <span class="hover_link">Abstract</span></a 32508 ><a 32509 class="clickable no-decoration" 32510 id="vhsjs_hide_734_1707793552_8324573" 32511 onclick="$('#733_1707793552_8324487').hide(function() { 32512 if (typeof Masonry === 'function') { 32513 $('.use_masonry').masonry(); 32514 }; 32515 }); 32516 $('#vhsjs_hide_734_1707793552_8324573').hide(); 32517 $('#vhsjs_view_734_1707793552_8324573').show();" 32518 style="display: none" 32519 ><i class="fa fa-caret-down"></i> 32520 <span class="hover_link">Abstract</span></a 32521 > 32522 <div 32523 data-display-control="734_1707793552_8324573" 32524 id="733_1707793552_8324487" 32525 style="display: none" 32526 > 32527 <div class="arrow-slidedown"> 32528 <blockquote> 32529 We propose a continuous optimization algorithm 32530 for the Column Subset Selection Problem (CSSP) 32531 and Nyström approximation. The CSSP and 32532 Nyström method construct low-rank 32533 approximations of matrices based on a 32534 predetermined subset of columns. It is well 32535 known that choosing the best column subset of 32536 size k is a difficult combinatorial problem. In 32537 this work, we show how one can approximate the 32538 optimal solution by defining a penalized 32539 continuous loss function that is minimized via 32540 stochastic gradient descent. We show that the 32541 gradients of this loss function can be estimated 32542 efficiently using matrix-vector products with a 32543 data matrix X in the case of the CSSP or a 32544 kernel matrix K in the case of the Nyström 32545 approximation. We provide numerical results for 32546 a number of real datasets showing that this 32547 continuous optimization is competitive against 32548 existing methods. 32549 </blockquote> 32550 </div> 32551 </div> 32552 </div> 32553 </div> 32554 <div class="slot-urls"></div> 32555 <a href="/wsc23papers/303.pdf" target="_blank">pdf</a 32556 ><br /> 32557 </div> 32558 </div> 32559 <div class="session-entry">
32560 <span class="session-event-type">Technical Session</span 32561 ><span class="type-track-spacer"> · </span 32562 ><span class="program-track">Simulation Optimization</span 32563 ><br /> 32564 <div class="session-title"> 32565 Queueing Systems and Experiment Design 32566 </div> 32567 <div class="session-chair"> 32568 Chair: David J. Eckman (Texas A&M University)<br /> 32569 </div> 32570 <div class="slot-entry"> 32571 <a name="inv183" tabindex="-1"></a> 32572 <div class="slot-title-line"> 32573 <span class="slot-title" 32574 >Sequential Simulation Optimization with Censoring: An 32575 Application to Bike Sharing Systems</span 32576 > 32577 </div> 32578 <div class="slot-authors"> 32579 Cedric Gibbons (Chilean Navy), James Grant (Lancaster 32580 University), and Roberto Szechtman (Naval Postgraduate 32581 School) 32582 </div> 32583 <div class="slot-abstract"> 32584 <div> 32585 <a 32586 class="clickable no-decoration" 32587 id="vhsjs_view_736_1707793552_8366442" 32588 onclick="$('#vhsjs_view_736_1707793552_8366442').hide(); 32589 $('#vhsjs_hide_736_1707793552_8366442').show(); 32590 $('#735_1707793552_8366356').slideDown(function() { 32591 if (typeof Masonry === 'function') { 32592 $('.use_masonry').masonry(); 32593 }; 32594 32595 });" 32596 ><i class="fa fa-caret-right"></i> 32597 <span class="hover_link">Abstract</span></a 32598 ><a 32599 class="clickable no-decoration" 32600 id="vhsjs_hide_736_1707793552_8366442" 32601 onclick="$('#735_1707793552_8366356').hide(function() { 32602 if (typeof Masonry === 'function') { 32603 $('.use_masonry').masonry(); 32604 }; 32605 }); 32606 $('#vhsjs_hide_736_1707793552_8366442').hide(); 32607 $('#vhsjs_view_736_1707793552_8366442').show();" 32608 style="display: none" 32609 ><i class="fa fa-caret-down"></i> 32610 <span class="hover_link">Abstract</span></a 32611 > 32612 <div 32613 data-display-control="736_1707793552_8366442" 32614 id="735_1707793552_8366356" 32615 style="display: none" 32616 > 32617 <div class="arrow-slidedown"> 32618 <blockquote> 32619 Sequential Simulation Optimization is an online 32620 optimization framework where an operator 32621 iterates periodically between collecting data 32622 from a real-world system, using stochastic 32623 simulation to approximate the optimal values of 32624 some operational variables, and setting some 32625 choice of variables in the system for the next 32626 period. The aim is to converge to an optimum 32627 efficiently, as uncertainty due to finite data 32628 and finitely many simulations eventually 32629 reduces. Using Bike Sharing Systems (BSS) as a 32630 motivating example, we analyze a variant where 32631 data from the real-world system is subject to 32632 censoring, whose nature depends on the system 32633 variables selected by the operator. In the BSS 32634 setting, censoring is of customer demand, or 32635 slots in which to drop bikes off in. We show 32636 that a method built upon Sample Average 32637 Approximation attains asymptotically vanishing 32638 error in its parameter estimates and 32639 specification of the optimal operational 32640 variables. 32641 </blockquote> 32642 </div> 32643 </div> 32644 </div> 32645 </div> 32646 <div class="slot-urls"></div> 32647 <a href="/wsc23papers/305.pdf" target="_blank">pdf</a 32648 ><br /> 32649 </div> 32650 <div class="slot-entry"> 32651 <a name="con297" tabindex="-1"></a> 32652 <div class="slot-title-line">
32653 <span class="slot-title" 32654 >SF-SFD: Stochastic Optimization of Fourier 32655 Coefficients to Generate Space-Filling Designs</span 32656 > 32657 </div> 32658 <div class="slot-authors"> 32659 Manisha Garg (University of Illinois Urbana-Champaign, 32660 Argonne National Laboratory) and Tyler H. Chang and 32661 Krishnan Raghavan (Argonne National Laboratory) 32662 </div> 32663 <div class="slot-abstract"> 32664 <div> 32665 <a 32666 class="clickable no-decoration" 32667 id="vhsjs_view_738_1707793552_8388536" 32668 onclick="$('#vhsjs_view_738_1707793552_8388536').hide(); 32669 $('#vhsjs_hide_738_1707793552_8388536').show(); 32670 $('#737_1707793552_8388457').slideDown(function() { 32671 if (typeof Masonry === 'function') { 32672 $('.use_masonry').masonry(); 32673 }; 32674 32675 });" 32676 ><i class="fa fa-caret-right"></i> 32677 <span class="hover_link">Abstract</span></a 32678 ><a 32679 class="clickable no-decoration" 32680 id="vhsjs_hide_738_1707793552_8388536" 32681 onclick="$('#737_1707793552_8388457').hide(function() { 32682 if (typeof Masonry === 'function') { 32683 $('.use_masonry').masonry(); 32684 }; 32685 }); 32686 $('#vhsjs_hide_738_1707793552_8388536').hide(); 32687 $('#vhsjs_view_738_1707793552_8388536').show();" 32688 style="display: none" 32689 ><i class="fa fa-caret-down"></i> 32690 <span class="hover_link">Abstract</span></a 32691 > 32692 <div 32693 data-display-control="738_1707793552_8388536" 32694 id="737_1707793552_8388457" 32695 style="display: none" 32696 > 32697 <div class="arrow-slidedown"> 32698 <blockquote> 32699 Due to the curse of dimensionality, it is often 32700 prohibitively expensive to generate 32701 deterministic space-filling designs. On the 32702 other hand, when using naive uniform random 32703 sampling to generate designs cheaply, design 32704 points tend to concentrate in a small region of 32705 the design space. Although, it is preferable in 32706 these cases to utilize quasi-random techniques 32707 such as Sobol sequences and Latin hypercube 32708 designs over uniform random sampling in many 32709 settings, these methods have their own caveats 32710 especially in high-dimensional spaces. In this 32711 paper, we propose a technique that addresses the 32712 fundamental issue of measure concentration by 32713 updating high-dimensional distribution functions 32714 to produce better space-filling designs. Then, 32715 we show that our technique can outperform Latin 32716 hypercube sampling and Sobol sequences by the 32717 discrepancy metric while generating 32718 moderately-sized space-filling samples for 32719 high-dimensional problems. 32720 </blockquote> 32721 </div> 32722 </div> 32723 </div> 32724 </div> 32725 <div class="slot-urls"></div> 32726 <a href="/wsc23papers/306.pdf" target="_blank">pdf</a 32727 ><br /> 32728 </div> 32729 </div> 32730 </div> 32731 <div class="centered"> 32732 <div class="top-link"><a href="#top">Return to Top</a></div> 32733 </div> 32734 <hr /> 32735 </div> 32736 <div class="area-section"> 32737 <div class="centered"> 32738 <a name="ptrack122" tabindex="-1"></a> 32739 <div class="section-title"> 32740 Uncertainty Quantification and Robust Simulation 32741 </div> 32742 </div> 32743 <div class="centered track-chair">
32744 <span class="track-chair-role" 32745 >Track Coordinator - Uncertainty Quantification and Robust 32746 Simulation: </span 32747 ><span class="track-chair-names" 32748 >Xi Chen (Virginia Tech), Wei Xie (Northeastern 32749 University)</span 32750 > 32751 </div> 32752 <div class="section-entry"> 32753 <div class="session-entry"> 32754 <span class="session-event-type">Technical Session</span 32755 ><span class="type-track-spacer"> · </span 32756 ><span class="program-track" 32757 >Uncertainty Quantification and Robust Simulation</span 32758 ><br /> 32759 <div class="session-title"> 32760 Optimization under Input Uncertainty and Model Calibration 32761 </div> 32762 <div class="session-chair"> 32763 Chair: Guangwu Liu (City University of Hong Kong)<br /> 32764 </div> 32765 <div class="slot-entry"> 32766 <a name="inv136" tabindex="-1"></a> 32767 <div class="slot-title-line"> 32768 <span class="slot-title" 32769 >Upper-Confidence-Bound Procedure for Robust Selection 32770 of the Best</span 32771 > 32772 </div> 32773 <div class="slot-authors"> 32774 Yuchen Wan (Fudan University); Weiwei Fan (Tongji 32775 University); and L. Jeff Hong (Fudan University, School 32776 of Management) 32777 </div> 32778 <div class="slot-abstract"> 32779 <div> 32780 <a 32781 class="clickable no-decoration" 32782 id="vhsjs_view_740_1707793552_8467917" 32783 onclick="$('#vhsjs_view_740_1707793552_8467917').hide(); 32784 $('#vhsjs_hide_740_1707793552_8467917').show(); 32785 $('#739_1707793552_8467832').slideDown(function() { 32786 if (typeof Masonry === 'function') { 32787 $('.use_masonry').masonry(); 32788 }; 32789 32790 });" 32791 ><i class="fa fa-caret-right"></i> 32792 <span class="hover_link">Abstract</span></a 32793 ><a 32794 class="clickable no-decoration" 32795 id="vhsjs_hide_740_1707793552_8467917" 32796 onclick="$('#739_1707793552_8467832').hide(function() { 32797 if (typeof Masonry === 'function') { 32798 $('.use_masonry').masonry(); 32799 }; 32800 }); 32801 $('#vhsjs_hide_740_1707793552_8467917').hide(); 32802 $('#vhsjs_view_740_1707793552_8467917').show();" 32803 style="display: none" 32804 ><i class="fa fa-caret-down"></i> 32805 <span class="hover_link">Abstract</span></a 32806 > 32807 <div 32808 data-display-control="740_1707793552_8467917" 32809 id="739_1707793552_8467832" 32810 style="display: none" 32811 > 32812 <div class="arrow-slidedown"> 32813 <blockquote> 32814 Robust selection of the best (RSB) is an 32815 important problem in the simulation area, when 32816 there exists input uncertainty in the underlying 32817 simulation model. RSB models this input 32818 uncertainty by a discrete ambiguity set and then 32819 proposes a two-layer framework under which the 32820 best alternative is defined to have the best 32821 worst-case mean performance over the ambiguity 32822 set. In this paper, we adopt a fixed-budget 32823 framework to address the RSB problem. 32824 Specifically, in contrast with existing 32825 procedures, we develop a new robust 32826 upper-confidence-bound (UCB) procedure, named as 32827 R-UCB. We can show that, the R-UCB procedure 32828 successfully inherits the simplicity and 32829 convergence guarantee of the traditional UCB 32830 procedure. Furthermore, simulation experiments 32831 demonstrate that the R-UCB procedure numerically 32832 outperforms the existing RSB procedures. 32833 </blockquote> 32834 </div> 32835 </div> 32836 </div> 32837 </div> 32838 <div class="slot-urls"></div> 32839 <a href="/wsc23papers/307.pdf" target="_blank">pdf</a 32840 ><br /> 32841 </div> 32842 <div class="slot-entry"> 32843 <a name="con119" tabindex="-1"></a> 32844 <div class="slot-title-line">
32845 <span class="slot-title" 32846 >Input Data Collection versus Simulation: Simultaneous 32847 Resource Allocation</span 32848 > 32849 </div> 32850 <div class="slot-authors"> 32851 Yuhao Wang and Enlu Zhou (Georgia Institute of 32852 Technology) 32853 </div> 32854 <div class="slot-abstract"> 32855 <div> 32856 <a 32857 class="clickable no-decoration" 32858 id="vhsjs_view_742_1707793552_8489797" 32859 onclick="$('#vhsjs_view_742_1707793552_8489797').hide(); 32860 $('#vhsjs_hide_742_1707793552_8489797').show(); 32861 $('#741_1707793552_8489716').slideDown(function() { 32862 if (typeof Masonry === 'function') { 32863 $('.use_masonry').masonry(); 32864 }; 32865 32866 });" 32867 ><i class="fa fa-caret-right"></i> 32868 <span class="hover_link">Abstract</span></a 32869 ><a 32870 class="clickable no-decoration" 32871 id="vhsjs_hide_742_1707793552_8489797" 32872 onclick="$('#741_1707793552_8489716').hide(function() { 32873 if (typeof Masonry === 'function') { 32874 $('.use_masonry').masonry(); 32875 }; 32876 }); 32877 $('#vhsjs_hide_742_1707793552_8489797').hide(); 32878 $('#vhsjs_view_742_1707793552_8489797').show();" 32879 style="display: none" 32880 ><i class="fa fa-caret-down"></i> 32881 <span class="hover_link">Abstract</span></a 32882 > 32883 <div 32884 data-display-control="742_1707793552_8489797" 32885 id="741_1707793552_8489716" 32886 style="display: none" 32887 > 32888 <div class="arrow-slidedown"> 32889 <blockquote> 32890 This paper investigates the problem of ranking 32891 and selection under input uncertainty with 32892 simultaneous resource allocation. In this 32893 problem, two types of resources are sequentially 32894 allocated at the same time to collect input data 32895 to reduce input uncertainty and run simulations 32896 to reduce stochastic uncertainty. We formulate 32897 the simultaneous resource allocation problem as 32898 a concave optimization problem that aims to 32899 maximize the asymptotic probability of correct 32900 selection (PCS) through the allocation policy 32901 for both input data collection and simulation, 32902 based on a moving-average estimator for 32903 aggregation of simulation outputs and its 32904 asymptotic normality. The two optimal policies 32905 are interdependent since they jointly affect the 32906 PCS. We derive the optimality equations to 32907 characterize the optimal policies and develop a 32908 fully sequential algorithm that demonstrates 32909 high efficiency through numerical experiments. 32910 </blockquote> 32911 </div> 32912 </div> 32913 </div> 32914 </div> 32915 <div class="slot-urls"></div> 32916 <a href="/wsc23papers/308.pdf" target="_blank">pdf</a 32917 ><br /> 32918 </div> 32919 <div class="slot-entry"> 32920 <a name="con211" tabindex="-1"></a> 32921 <div class="slot-title-line">
32922 <span class="slot-title" 32923 >Representative Calibration Using Black-box 32924 Optimization and Clustering</span 32925 > 32926 </div> 32927 <div class="slot-authors"> 32928 Serin Lee, Pariyakorn Maneekul, and Zelda B. Zabinsky 32929 (University of Washington) 32930 </div> 32931 <div class="slot-abstract"> 32932 <div> 32933 <a 32934 class="clickable no-decoration" 32935 id="vhsjs_view_744_1707793552_8511744" 32936 onclick="$('#vhsjs_view_744_1707793552_8511744').hide(); 32937 $('#vhsjs_hide_744_1707793552_8511744').show(); 32938 $('#743_1707793552_851166').slideDown(function() { 32939 if (typeof Masonry === 'function') { 32940 $('.use_masonry').masonry(); 32941 }; 32942 32943 });" 32944 ><i class="fa fa-caret-right"></i> 32945 <span class="hover_link">Abstract</span></a 32946 ><a 32947 class="clickable no-decoration" 32948 id="vhsjs_hide_744_1707793552_8511744" 32949 onclick="$('#743_1707793552_851166').hide(function() { 32950 if (typeof Masonry === 'function') { 32951 $('.use_masonry').masonry(); 32952 }; 32953 }); 32954 $('#vhsjs_hide_744_1707793552_8511744').hide(); 32955 $('#vhsjs_view_744_1707793552_8511744').show();" 32956 style="display: none" 32957 ><i class="fa fa-caret-down"></i> 32958 <span class="hover_link">Abstract</span></a 32959 > 32960 <div 32961 data-display-control="744_1707793552_8511744" 32962 id="743_1707793552_851166" 32963 style="display: none" 32964 > 32965 <div class="arrow-slidedown"> 32966 <blockquote> 32967 Calibration is a crucial step for model 32968 validity, yet its representation is often 32969 disregarded. This paper proposes a two-stage 32970 approach to calibrate a model that represents 32971 target data by identifying multiple diverse 32972 parameter sets while remaining computationally 32973 efficient. The first stage employs a black-box 32974 optimization algorithm to generate near-optimal 32975 parameter sets, the second stage clusters the 32976 generated parameter sets. Five black-box 32977 optimization algorithms, namely, Latin Hypercube 32978 Sampling (LHS), Sequential Model-based Algorithm 32979 Configuration (SMAC), Optuna, Simulated 32980 Annealing (SA), and Genetic Algorithm (GA), are 32981 tested and compared using a disease-opinion 32982 compartmental model with predicted health 32983 outcomes. Results show that LHS and Optuna allow 32984 more exploration and capture more variety in 32985 possible future health outcomes. SMAC, SA, and 32986 GA, are better at finding the best parameter set 32987 but their sampling approach generates less 32988 diverse model outcomes. This two-stage approach 32989 can reduce computation time while producing 32990 robust and representative calibration. 32991 </blockquote> 32992 </div> 32993 </div> 32994 </div> 32995 </div> 32996 <div class="slot-urls"></div> 32997 <a href="/wsc23papers/309.pdf" target="_blank">pdf</a 32998 ><br /> 32999 </div> 33000 </div> 33001 <div class="session-entry">
33002 <span class="session-event-type">Technical Session</span 33003 ><span class="type-track-spacer"> · </span 33004 ><span class="program-track" 33005 >Uncertainty Quantification and Robust Simulation</span 33006 ><br /> 33007 <div class="session-title">Uncertainty Quantification</div> 33008 <div class="session-chair"> 33009 Chair: Hong Wan (North Carolina State University)<br /> 33010 </div> 33011 <div class="slot-entry"> 33012 <a name="inv168" tabindex="-1"></a> 33013 <div class="slot-title-line"> 33014 <span class="slot-title" 33015 >Resampling Stochastic Gradient Descent Cheaply</span 33016 > 33017 </div> 33018 <div class="slot-authors"> 33019 Henry Lam and Zitong Wang (Columbia University) 33020 </div> 33021 <div class="slot-abstract"> 33022 <div> 33023 <a 33024 class="clickable no-decoration" 33025 id="vhsjs_view_746_1707793552_8555696" 33026 onclick="$('#vhsjs_view_746_1707793552_8555696').hide(); 33027 $('#vhsjs_hide_746_1707793552_8555696').show(); 33028 $('#745_1707793552_8555615').slideDown(function() { 33029 if (typeof Masonry === 'function') { 33030 $('.use_masonry').masonry(); 33031 }; 33032 33033 });" 33034 ><i class="fa fa-caret-right"></i> 33035 <span class="hover_link">Abstract</span></a 33036 ><a 33037 class="clickable no-decoration" 33038 id="vhsjs_hide_746_1707793552_8555696" 33039 onclick="$('#745_1707793552_8555615').hide(function() { 33040 if (typeof Masonry === 'function') { 33041 $('.use_masonry').masonry(); 33042 }; 33043 }); 33044 $('#vhsjs_hide_746_1707793552_8555696').hide(); 33045 $('#vhsjs_view_746_1707793552_8555696').show();" 33046 style="display: none" 33047 ><i class="fa fa-caret-down"></i> 33048 <span class="hover_link">Abstract</span></a 33049 > 33050 <div 33051 data-display-control="746_1707793552_8555696" 33052 id="745_1707793552_8555615" 33053 style="display: none" 33054 > 33055 <div class="arrow-slidedown"> 33056 <blockquote> 33057 Stochastic gradient descent (SGD) or stochastic 33058 approximation has been widely used in model 33059 training and stochastic optimization. While 33060 there is a huge literature on analyzing its 33061 convergence, inference on the obtained solutions 33062 from SGD has only been recently studied, yet is 33063 important due to the growing need for 33064 uncertainty quantification. We investigate two 33065 easily implementable resampling-based methods to 33066 construct confidence intervals for SGD 33067 solutions. One uses multiple, but few, SGDs in 33068 parallel via resampling with replacement from 33069 the data, and another operates this in an online 33070 fashion. Our methods can be regarded as 33071 enhancements of established bootstrap schemes to 33072 substantially reduce the computation effort in 33073 terms of resampling requirements, while at the 33074 same time bypasses the intricate mi
33074xing 33075 conditions in existing batching methods. We 33076 achieve these via a recent cheap bootstrap idea 33077 and Berry-Esseen-type bound for SGD. 33078 </blockquote> 33079 </div> 33080 </div> 33081 </div> 33082 </div> 33083 <div class="slot-urls"></div> 33084 <a href="/wsc23papers/310.pdf" target="_blank">pdf</a 33085 ><br /> 33086 </div> 33087 <div class="slot-entry"> 33088 <a name="con187" tabindex="-1"></a> 33089 <div class="slot-title-line"> 33090 <span class="slot-title" 33091 >Input Uncertainty Quantification Via Simulation 33092 Bootstrapping</span 33093 > 33094 </div> 33095 <div class="slot-authors"> 33096 Manjing Zhang (Guangdong Laboratory of Artificial 33097 Intelligence and Digital Economy (SZ)), Guangwu Liu 33098 (City University of Hong Kong), Shan Dai (Shenzhen 33099 Research Institute of Big Data), and Yulin He (Guangdong 33100 Laboratory of Artificial Intelligence and Digital 33101 Economy (SZ)) 33102 </div> 33103 <div class="slot-abstract"> 33104 <div> 33105 <a 33106 class="clickable no-decoration" 33107 id="vhsjs_view_748_1707793552_857972" 33108 onclick="$('#vhsjs_view_748_1707793552_857972').hide(); 33109 $('#vhsjs_hide_748_1707793552_857972').show(); 33110 $('#747_1707793552_8579638').slideDown(function() { 33111 if (typeof Masonry === 'function') { 33112 $('.use_masonry').masonry(); 33113 }; 33114 33115 });" 33116 ><i class="fa fa-caret-right"></i> 33117 <span class="hover_link">Abstract</span></a 33118 ><a 33119 class="clickable no-decoration" 33120 id="vhsjs_hide_748_1707793552_857972" 33121 onclick="$('#747_1707793552_8579638').hide(function() { 33122 if (typeof Masonry === 'function') { 33123 $('.use_masonry').masonry(); 33124 }; 33125 }); 33126 $('#vhsjs_hide_748_1707793552_857972').hide(); 33127 $('#vhsjs_view_748_1707793552_857972').show();" 33128 style="display: none" 33129 ><i class="fa fa-caret-down"></i> 33130 <span class="hover_link">Abstract</span></a 33131 > 33132 <div 33133 data-display-control="748_1707793552_857972" 33134 id="747_1707793552_8579638" 33135 style="display: none" 33136 > 33137 <div class="arrow-slidedown"> 33138 <blockquote> 33139 Input uncertainty, which refers to the output 33140 variability arising from statistical noise in 33141 specifying the input models, has been 33142 intensively studied recently. Ignoring input 33143 uncertainty often leads to poor estimates of 33144 system performance. In the non-parametric 33145 setting, input uncertainty is commonly estimated 33146 via bootstrap, but the performance by 33147 traditional bootstrap resampling is compromised 33148 when input uncertainty is also associated with 33149 simulation uncertainty. Nested simulation is 33150 studied to improve the performance by taking 33151 variance estimation into account, but suffers 33152 from a substantial burden on required simulation 33153 effort. To tackle the above problems, this paper 33154 introduces a non-nested method to build 33155 asymptotically valid confidence intervals for 33156 input uncertainty quantification. The 33157 convergence properties are studied, which 33158 establish statistical guarantees for the 33159 proposed estimators related to real-data size 33160 and bootstrap budget. An easy-implemented 33161 algorithm is also provided. Numerical examples 33162 show that the estimated confidence intervals 33163 perform satisfactorily under given confidence 33164 levels. 33165 </blockquote> 33166 </div> 33167 </div> 33168 </div> 33169 </div> 33170 <div class="slot-urls"></div> 33171 <a href="/wsc23papers/311.pdf" target="_blank">pdf</a 33172 ><br /> 33173 </div> 33174 <div class="slot-entry"> 33175 <a name="con262" tabindex="-1"></a> 33176 <div class="slot-title-line">
33177 <span class="slot-title" 33178 >Asymptotic Normality of Joint Metamodel-Based Sobol' 33179 Index Estimators</span 33180 > 33181 </div> 33182 <div class="slot-authors"> 33183 Jingtao Zhang, Xi Chen, and Ruochen Wang (Virginia Tech) 33184 </div> 33185 <div class="slot-abstract"> 33186 <div> 33187 <a 33188 class="clickable no-decoration" 33189 id="vhsjs_view_750_1707793552_860098" 33190 onclick="$('#vhsjs_view_750_1707793552_860098').hide(); 33191 $('#vhsjs_hide_750_1707793552_860098').show(); 33192 $('#749_1707793552_8600898').slideDown(function() { 33193 if (typeof Masonry === 'function') { 33194 $('.use_masonry').masonry(); 33195 }; 33196 33197 });" 33198 ><i class="fa fa-caret-right"></i> 33199 <span class="hover_link">Abstract</span></a 33200 ><a 33201 class="clickable no-decoration" 33202 id="vhsjs_hide_750_1707793552_860098" 33203 onclick="$('#749_1707793552_8600898').hide(function() { 33204 if (typeof Masonry === 'function') { 33205 $('.use_masonry').masonry(); 33206 }; 33207 }); 33208 $('#vhsjs_hide_750_1707793552_860098').hide(); 33209 $('#vhsjs_view_750_1707793552_860098').show();" 33210 style="display: none" 33211 ><i class="fa fa-caret-down"></i> 33212 <span class="hover_link">Abstract</span></a 33213 > 33214 <div 33215 data-display-control="750_1707793552_860098" 33216 id="749_1707793552_8600898" 33217 style="display: none" 33218 > 33219 <div class="arrow-slidedown"> 33220 <blockquote> 33221 This paper proposes two joint metamodel-based 33222 Sobol' index estimators and investigates their 33223 asymptotic properties. The numerical evaluation 33224 corroborates the theoretical results and 33225 highlights the impact of the combination of 33226 training sample size and Monte Carlo sample size 33227 on the estimators' performance. 33228 </blockquote> 33229 </div> 33230 </div> 33231 </div> 33232 </div> 33233 <div class="slot-urls"></div> 33234 <a href="/wsc23papers/312.pdf" target="_blank">pdf</a 33235 ><br /> 33236 </div> 33237 </div> 33238 <div class="session-entry"> 33239 <span class="session-event-type">Technical Session</span 33240 ><span class="type-track-spacer"> · </span 33241 ><span class="program-track" 33242 >Uncertainty Quantification and Robust Simulation</span 33243 ><br /> 33244 <div class="session-title"> 33245 Input Modeling and Optimization via Machine Learning 33246 </div> 33247 <div class="session-chair"> 33248 Chair: Jingtao Zhang (Virginia Tech)<br /> 33249 </div> 33250 <div class="slot-entry"> 33251 <a name="con326" tabindex="-1"></a> 33252 <div class="slot-title-line"> 33253 <span class="slot-title" 33254 >An Intelligent Framework to Maximize Individual 33255 Driver Income</span 33256 > 33257 </div> 33258 <div class="slot-authors"> 33259 Fang Chen and Hua Cai (Purdue University) and Hong Wan 33260 (North Carolina State University) 33261 </div> 33262 <div class="slot-abstract"> 33263 <div> 33264 <a 33265 class="clickable no-decoration" 33266 id="vhsjs_view_752_1707793552_8647037" 33267 onclick="$('#vhsjs_view_752_1707793552_8647037').hide(); 33268 $('#vhsjs_hide_752_1707793552_8647037').show();
33269 $('#751_1707793552_8646955').slideDown(function() { 33270 if (typeof Masonry === 'function') { 33271 $('.use_masonry').masonry(); 33272 }; 33273 33274 });" 33275 ><i class="fa fa-caret-right"></i> 33276 <span class="hover_link">Abstract</span></a 33277 ><a 33278 class="clickable no-decoration" 33279 id="vhsjs_hide_752_1707793552_8647037" 33280 onclick="$('#751_1707793552_8646955').hide(function() { 33281 if (typeof Masonry === 'function') { 33282 $('.use_masonry').masonry(); 33283 }; 33284 }); 33285 $('#vhsjs_hide_752_1707793552_8647037').hide(); 33286 $('#vhsjs_view_752_1707793552_8647037').show();" 33287 style="display: none" 33288 ><i class="fa fa-caret-down"></i> 33289 <span class="hover_link">Abstract</span></a 33290 > 33291 <div 33292 data-display-control="752_1707793552_8647037" 33293 id="751_1707793552_8646955" 33294 style="display: none" 33295 > 33296 <div class="arrow-slidedown"> 33297 <blockquote> 33298 The ridesharing platform has significantly 33299 changed how taxis operate in recent years. Most 33300 previous works focus on improving the user 33301 experience and maximizing the revenue from the 33302 platform or system level. The individual driver 33303 benefits are rarely addressed. In this work, we 33304 propose a deep reinforcement learning-based 33305 framework to help the individual driver maximize 33306 their daily income via order selections and 33307 self-repositioning. We first formulated the taxi 33308 operation as a Markov Decision Process. Then we 33309 created a multi-agent simulation consisting of 33310 the taxi drivers that use different strategies. 33311 A deep Q network-based (DQN) framework is 33312 proposed for drivers to learn which orders to 33313 select and where to reposition. Our result shows 33314 the driver who adopts the DQN framework 33315 outperforms all other drivers. Furthermore, we 33316 also found that the optimal policy does not 33317 suggest the driver operating in particular areas 33318 but recommends selecting orders with $5 to $7.5 33319 taxi fare. 33320 </blockquote> 33321 </div> 33322 </div> 33323 </div> 33324 </div> 33325 <div class="slot-urls"></div> 33326 <a href="/wsc23papers/313.pdf" target="_blank">pdf</a 33327 ><br /> 33328 </div> 33329 <div class="slot-entry"> 33330 <a name="inv212" tabindex="-1"></a> 33331 <div class="slot-title-line"> 33332 <span class="slot-title" 33333 >Virtual Wearable Sensor Data Generation with 33334 Generative Adversarial Networks</span 33335 > 33336 </div> 33337 <div class="slot-authors"> 33338 Yining Huang and Hong Wan (North Carolina State 33339 University) and Xi Chen (Virginia Tech) 33340 </div> 33341 <div class="slot-abstract"> 33342 <div> 33343 <a 33344 class="clickable no-decoration" 33345 id="vhsjs_view_754_1707793552_866861" 33346 onclick="$('#vhsjs_view_754_1707793552_866861').hide(); 33347 $('#vhsjs_hide_754_1707793552_866861').show();
33348 $('#753_1707793552_8668532').slideDown(function() { 33349 if (typeof Masonry === 'function') { 33350 $('.use_masonry').masonry(); 33351 }; 33352 33353 });" 33354 ><i class="fa fa-caret-right"></i> 33355 <span class="hover_link">Abstract</span></a 33356 ><a 33357 class="clickable no-decoration" 33358 id="vhsjs_hide_754_1707793552_866861" 33359 onclick="$('#753_1707793552_8668532').hide(function() { 33360 if (typeof Masonry === 'function') { 33361 $('.use_masonry').masonry(); 33362 }; 33363 }); 33364 $('#vhsjs_hide_754_1707793552_866861').hide(); 33365 $('#vhsjs_view_754_1707793552_866861').show();" 33366 style="display: none" 33367 ><i class="fa fa-caret-down"></i> 33368 <span class="hover_link">Abstract</span></a 33369 > 33370 <div 33371 data-display-control="754_1707793552_866861" 33372 id="753_1707793552_8668532" 33373 style="display: none" 33374 > 33375 <div class="arrow-slidedown"> 33376 <blockquote> 33377 This study delves into the utilization of 33378 Generative Adversarial Networks (GANs) for 33379 generating subject-specific time series sensor 33380 data, offering an innovative alternative to 33381 traditional metamodel-based simulations. We 33382 undertake an in-depth analysis of DoppelGANger, 33383 a prominent GAN variant for time series data and 33384 metadata generation, evaluating its efficiency 33385 and efficacy. The sensor data for this 33386 investigation was sourced from the National 33387 Health and Nutrition Examination Survey, which 33388 served as the foundational training set. We 33389 scrutinized the synthesized sensor data 33390 corresponding to various physical attributes, 33391 focusing on the temporal and multi-dimensional 33392 statistical properties. Our empirical findings 33393 underscore the potential of GANs to adeptly 33394 capture the time-dependent correlations and the 33395 intricate statistical characteristics inherent 33396 in multi-dimensional data. This insight into 33397 GANs' capabilities is a crucial step towards 33398 more sophisticated synthetic data generation, 33399 with significant implications for future 33400 applications in wearable technology and 33401 personalized health monitoring systems. 33402 </blockquote> 33403 </div> 33404 </div> 33405 </div> 33406 </div> 33407 <div class="slot-urls"></div> 33408 <a href="/wsc23papers/314.pdf" target="_blank">pdf</a 33409 ><br /> 33410 </div> 33411 </div> 33412 </div> 33413 <div class="centered"> 33414 <div class="top-link"><a href="#top">Return to Top</a></div> 33415 </div> 33416 <hr /> 33417 </div> 33418 33419 33420 <div class="area-section"> 33421 <div class="centered"> 33422 <a name="ptrack133" tabindex="-1"></a> 33423 <div class="section-title">Vendor</div> 33424 </div> 33425 <div class="centered track-chair">
33426 <span class="track-chair-role" 33427 >Track Coordinator - Vendor: </span 33428 ><span class="track-chair-names" 33429 >Aristotelis Thanos (University of Miami), Edward Williams 33430 (PMC)</span 33431 > 33432 </div> 33433 <div class="section-entry"> 33434 <div class="session-entry"> 33435 <span class="session-event-type">Vendor Session</span 33436 ><span class="type-track-spacer"> · </span 33437 ><span class="program-track">Vendor</span><br /> 33438 <div class="session-title"> 33439 Simulation Software for Manufacturing 33440 </div> 33441 <div class="session-chair"> 33442 Chair: Nurcin Celik (University of Miami)<br /> 33443 </div> 33444 <div class="slot-entry"> 33445 <a name="vdra107" tabindex="-1"></a> 33446 <div class="slot-title-line"> 33447 <span class="slot-title" 33448 >Introducing Mozart Fab Wise: a Cloud-based Simulation 33449 Solution for Semiconductor Fabs</span 33450 > 33451 </div> 33452 <div class="slot-authors"> 33453 Keyhoon Ko (VMS Global, Inc.) 33454 </div> 33455 <div class="slot-abstract"> 33456 <div> 33457 <a 33458 class="clickable no-decoration" 33459 id="vhsjs_view_764_1707793552_8943706" 33460 onclick="$('#vhsjs_view_764_1707793552_8943706').hide(); 33461 $('#vhsjs_hide_764_1707793552_8943706').show(); 33462 $('#763_1707793552_8943622').slideDown(function() { 33463 if (typeof Masonry === 'function') { 33464 $('.use_masonry').masonry(); 33465 }; 33466 33467 });" 33468 ><i class="fa fa-caret-right"></i> 33469 <span class="hover_link">Abstract</span></a 33470 ><a 33471 class="clickable no-decoration" 33472 id="vhsjs_hide_764_1707793552_8943706" 33473 onclick="$('#763_1707793552_8943622').hide(function() { 33474 if (typeof Masonry === 'function') { 33475 $('.use_masonry').masonry(); 33476 }; 33477 }); 33478 $('#vhsjs_hide_764_1707793552_8943706').hide(); 33479 $('#vhsjs_view_764_1707793552_8943706').show();" 33480 style="display: none" 33481 ><i class="fa fa-caret-down"></i> 33482 <span class="hover_link">Abstract</span></a 33483 > 33484 <div 33485 data-display-control="764_1707793552_8943706" 33486 id="763_1707793552_8943622" 33487 style="display: none" 33488 > 33489 <div class="arrow-slidedown"> 33490 <blockquote> 33491 In response to the intricate planning and 33492 scheduling challenges encountered in the 33493 semiconductor industry, VMS leverages its 33494 extensive 20-year experience to introduce MOZART 33495 Fab WISE, a dedicated cloud-based simulation 33496 solution. Fab WISE offers an array of data 33497 interfaces, enabling the generation of 33498 comprehensive data and rich analytical reports. 33499 Customers have the flexibility to customize the 33500 level of modeling detail based on their specific 33501 objectives, with the capacity to conduct both 33502 short-term and long-term simulations. Remarkably 33503 adaptable, Fab WISE can function as a blueprint 33504 for capacity planning (CP), factory planning 33505 (FP), and real-time scheduling (RTS), making it 33506 a versatile solution tailored to 33507 customer-specific requirements. 33508 </blockquote> 33509 </div> 33510 </div> 33511 </div> 33512 </div> 33513 <div class="slot-urls"></div> 33514 </div> 33515 <div class="slot-entry"> 33516 <a name="vdra104" tabindex="-1"></a> 33517 <div class="slot-title-line">
33518 <span class="slot-title" 33519 >Chiaha Discrete Rate Simulation</span 33520 > 33521 </div> 33522 <div class="slot-authors"> 33523 Andrew Siprelle (Chiaha.ai) 33524 </div> 33525 <div class="slot-abstract"> 33526 <div> 33527 <a 33528 class="clickable no-decoration" 33529 id="vhsjs_view_766_1707793552_9021192" 33530 onclick="$('#vhsjs_view_766_1707793552_9021192').hide(); 33531 $('#vhsjs_hide_766_1707793552_9021192').show(); 33532 $('#765_1707793552_9021113').slideDown(function() { 33533 if (typeof Masonry === 'function') { 33534 $('.use_masonry').masonry(); 33535 }; 33536 33537 });" 33538 ><i class="fa fa-caret-right"></i> 33539 <span class="hover_link">Abstract</span></a 33540 ><a 33541 class="clickable no-decoration" 33542 id="vhsjs_hide_766_1707793552_9021192" 33543 onclick="$('#765_1707793552_9021113').hide(function() { 33544 if (typeof Masonry === 'function') { 33545 $('.use_masonry').masonry(); 33546 }; 33547 }); 33548 $('#vhsjs_hide_766_1707793552_9021192').hide(); 33549 $('#vhsjs_view_766_1707793552_9021192').show();" 33550 style="display: none" 33551 ><i class="fa fa-caret-down"></i> 33552 <span class="hover_link">Abstract</span></a 33553 > 33554 <div 33555 data-display-control="766_1707793552_9021192" 33556 id="765_1707793552_9021113" 33557 style="display: none" 33558 > 33559 <div class="arrow-slidedown"> 33560 <blockquote> 33561 Discrete Rate Simulation (DRS) has been a key 33562 enabling technology used to address canonical 33563 problems in high-speed manufacturing. In this 33564 talk, we review the history of DRS from its 33565 creation 25 years ago, to our revolutionary new 33566 DRS engine and associated tools. Let Chiaha help 33567 you accelerate your "raw data to prediction" 33568 journey! 33569 </blockquote> 33570 </div> 33571 </div> 33572 </div> 33573 </div> 33574 <div class="slot-urls"></div> 33575 </div> 33576 </div> 33577 <div class="session-entry"> 33578 <span class="session-event-type">Vendor Session</span 33579 ><span class="type-track-spacer"> · </span 33580 ><span class="program-track">Vendor</span><br /> 33581 <div class="session-title">Innovative Simulation Tools</div> 33582 <div class="session-chair"> 33583 Chair: Bahar Biller (SAS Institute, Inc)<br /> 33584 </div> 33585 <div class="slot-entry"> 33586 <a name="vdra105" tabindex="-1"></a> 33587 <div class="slot-title-line"> 33588 <span class="slot-title" 33589 >Three Recent Advances in Simio: Auto-create, Advanced 33590 Traffic Control, and DDMRP</span 33591 > 33592 </div> 33593 <div class="slot-authors"> 33594 Jeffrey Smith and David Sturrock (Simio LLC) 33595 </div> 33596 <div class="slot-abstract"> 33597 <div> 33598 <a 33599 class="clickable no-decoration" 33600 id="vhsjs_view_768_1707793552_908236" 33601 onclick="$('#vhsjs_view_768_1707793552_908236').hide(); 33602 $('#vhsjs_hide_768_1707793552_908236').show(); 33603 $('#767_1707793552_908228').slideDown(function() { 33604 if (typeof Masonry === 'function') { 33605 $('.use_masonry').masonry(); 33606 }; 33607 33608 });" 33609 ><i class="fa fa-caret-right"></i>
33610 <span class="hover_link">Abstract</span></a 33611 ><a 33612 class="clickable no-decoration" 33613 id="vhsjs_hide_768_1707793552_908236" 33614 onclick="$('#767_1707793552_908228').hide(function() { 33615 if (typeof Masonry === 'function') { 33616 $('.use_masonry').masonry(); 33617 }; 33618 }); 33619 $('#vhsjs_hide_768_1707793552_908236').hide(); 33620 $('#vhsjs_view_768_1707793552_908236').show();" 33621 style="display: none" 33622 ><i class="fa fa-caret-down"></i> 33623 <span class="hover_link">Abstract</span></a 33624 > 33625 <div 33626 data-display-control="768_1707793552_908236" 33627 id="767_1707793552_908228" 33628 style="display: none" 33629 > 33630 <div class="arrow-slidedown"> 33631 <blockquote> 33632 This talk discusses and demonstrates three 33633 recent advances in Simio. The first is 33634 Simio’s Data Driven/Data Generated 33635 modeling approach using Simio custom objects, 33636 data tables, and the AutoCreateInstance and 33637 Create Element methods. While the objects in the 33638 Simio Standard Library are very flexible, custom 33639 objects can take your models to the next level. 33640 Furthermore, the “Create Object From 33641 This” and “Update Property Defaults 33642 From This” functions make the creation and 33643 maintenance of custom objects extremely easy. 33644 The second topic is Simio’s advanced 33645 traffic control features which significantly 33646 simplify deadlock prevention and path planning 33647 for systems with bi-directional links. Finally, 33648 the third topic is Simio’s new DDMRP 33649 (Demand-driven Materials Requirement Planning) 33650 tools. These features include the DDMRP 33651 replenishment method as part of the existing 33652 Inventory Element, DDMRP specific calculators 33653 with associated data table schema/templates for 33654 inputs and outputs, and DDMRP Specific 33655 Dashboards. 33656 </blockquote> 33657 </div> 33658 </div> 33659 </div> 33660 </div> 33661 <div class="slot-urls"></div> 33662 </div> 33663 <div class="slot-entry"> 33664 <a name="vdra103" tabindex="-1"></a> 33665 <div class="slot-title-line"> 33666 <span class="slot-title" 33667 >Enterprise Resource Simulator: Simulating Without 33668 Limits</span 33669 > 33670 </div> 33671 <div class="slot-authors"> 33672 Michel Hoffmeijer (InControl Enterprise Dynamics) and 33673 Fred Jansma (Incontrol Enterprise Dynamics) 33674 </div> 33675 <div class="slot-abstract"> 33676 <div> 33677 <a 33678 class="clickable no-decoration" 33679 id="vhsjs_view_770_1707793552_9095228" 33680 onclick="$('#vhsjs_view_770_1707793552_9095228').hide(); 33681 $('#vhsjs_hide_770_1707793552_9095228').show(); 33682 $('#769_1707793552_9095144').slideDown(function() { 33683 if (typeof Masonry === 'function') { 33684 $('.use_masonry').masonry(); 33685 }; 33686 33687 });" 33688 ><i class="fa fa-caret-right"></i>
33689 <span class="hover_link">Abstract</span></a 33690 ><a 33691 class="clickable no-decoration" 33692 id="vhsjs_hide_770_1707793552_9095228" 33693 onclick="$('#769_1707793552_9095144').hide(function() { 33694 if (typeof Masonry === 'function') { 33695 $('.use_masonry').masonry(); 33696 }; 33697 }); 33698 $('#vhsjs_hide_770_1707793552_9095228').hide(); 33699 $('#vhsjs_view_770_1707793552_9095228').show();" 33700 style="display: none" 33701 ><i class="fa fa-caret-down"></i> 33702 <span class="hover_link">Abstract</span></a 33703 > 33704 <div 33705 data-display-control="770_1707793552_9095228" 33706 id="769_1707793552_9095144" 33707 style="display: none" 33708 > 33709 <div class="arrow-slidedown"> 33710 <blockquote> 33711 Enterprise Resource Simulator (ERS) is a 33712 simulation platform that focuses on speed and 33713 versatility. It allows for models that are very 33714 large while still offering good performance. ERS 33715 does this by utilizing the full capabilities of 33716 modern computers in terms of efficient and 33717 scalable multi-threading. In addition to pure 33718 scale, ERS allows the models to have more depth 33719 and complexity by allowing multiple different 33720 formalisms in the same model. In addition to the 33721 features of the models, ERS is built to support 33722 multiple programming languages and to allow a 33723 user or a developer to build a full application 33724 upon it. This means that ERS can fulfill all 33725 simulation needs. 33726 </blockquote> 33727 </div> 33728 </div> 33729 </div> 33730 </div> 33731 <div class="slot-urls"></div> 33732 </div> 33733 </div> 33734 <div class="session-entry"> 33735 <span class="session-event-type">Vendor Session</span 33736 ><span class="type-track-spacer"> · </span 33737 ><span class="program-track">Vendor</span><br /> 33738 <div class="session-title"> 33739 Integrating AI and Simulation 33740 </div> 33741 <div class="session-chair"> 33742 Chair: John Shortle (George Mason University)<br /> 33743 </div> 33744 <div class="slot-entry"> 33745 <a name="vdra101" tabindex="-1"></a> 33746 <div class="slot-title-line"> 33747 <span class="slot-title" 33748 >SmartFactory AI Productivity Utilizing 33749 Simulation</span 33750 > 33751 </div> 33752 <div class="slot-authors"> 33753 Samantha Duchscherer (Applied Materials) 33754 </div> 33755 <div class="slot-abstract"> 33756 <div> 33757 <a 33758 class="clickable no-decoration" 33759 id="vhsjs_view_772_1707793552_9128046" 33760 onclick="$('#vhsjs_view_772_1707793552_9128046').hide(); 33761 $('#vhsjs_hide_772_1707793552_9128046').show(); 33762 $('#771_1707793552_912797').slideDown(function() { 33763 if (typeof Masonry === 'function') { 33764 $('.use_masonry').masonry(); 33765 }; 33766 33767 });" 33768 ><i class="fa fa-caret-right"></i>
33769 <span class="hover_link">Abstract</span></a 33770 ><a 33771 class="clickable no-decoration" 33772 id="vhsjs_hide_772_1707793552_9128046" 33773 onclick="$('#771_1707793552_912797').hide(function() { 33774 if (typeof Masonry === 'function') { 33775 $('.use_masonry').masonry(); 33776 }; 33777 }); 33778 $('#vhsjs_hide_772_1707793552_9128046').hide(); 33779 $('#vhsjs_view_772_1707793552_9128046').show();" 33780 style="display: none" 33781 ><i class="fa fa-caret-down"></i> 33782 <span class="hover_link">Abstract</span></a 33783 > 33784 <div 33785 data-display-control="772_1707793552_9128046" 33786 id="771_1707793552_912797" 33787 style="display: none" 33788 > 33789 <div class="arrow-slidedown"> 33790 <blockquote> 33791 Accurately simulating a semiconductor 33792 environment is challenging. Tools and processing 33793 steps are constantly evolving due to 33794 advancements in technology nodes and other 33795 unforeseen manufacturing modifications. However, 33796 AutoSched has out of the box capabilities to 33797 accurately simulate a particular tooling area as 33798 well as an entire facility. Models are also 33799 customizable to handle robust scenarios ranging 33800 from modifying how routes are built to varying 33801 the number of bottleneck stations. This 33802 flexibility makes AutoSched a key component in 33803 the data preparation phase for deploying various 33804 AI use cases. Here we will demonstration the 33805 capabilities of AutoSched modeling key factors 33806 inherent to semiconductor manufacturing and 33807 showcase how this enables AI innovations and 33808 real operational efficiency gains. From 33809 predicting lot cycle time with a gradient 33810 boosting model to utilizing reinforcement 33811 learning for optimizing dispatching parameter 33812 values and scheduling constraints, simulation is 33813 empowering SmartFactory AI Productivity. 33814 </blockquote> 33815 </div> 33816 </div> 33817 </div> 33818 </div> 33819 <div class="slot-urls"></div> 33820 </div> 33821 <div class="slot-entry"> 33822 <a name="vdra106" tabindex="-1"></a> 33823 <div class="slot-title-line"> 33824 <span class="slot-title" 33825 >Data Driven Digital Twin – Benefits and 33826 Advantages in Real-time Systems</span 33827 > 33828 </div> 33829 <div class="slot-authors"> 33830 Hosni Adra (CreateASoft, Inc) 33831 </div> 33832 <div class="slot-abstract"> 33833 <div> 33834 <a 33835 class="clickable no-decoration" 33836 id="vhsjs_view_774_1707793552_9140036" 33837 onclick="$('#vhsjs_view_774_1707793552_9140036').hide(); 33838 $('#vhsjs_hide_774_1707793552_9140036').show(); 33839 $('#773_1707793552_9139955').slideDown(function() { 33840 if (typeof Masonry === 'function') { 33841 $('.use_masonry').masonry(); 33842 }; 33843 33844 });" 33845 ><i class="fa fa-caret-right"></i>
33846 <span class="hover_link">Abstract</span></a 33847 ><a 33848 class="clickable no-decoration" 33849 id="vhsjs_hide_774_1707793552_9140036" 33850 onclick="$('#773_1707793552_9139955').hide(function() { 33851 if (typeof Masonry === 'function') { 33852 $('.use_masonry').masonry(); 33853 }; 33854 }); 33855 $('#vhsjs_hide_774_1707793552_9140036').hide(); 33856 $('#vhsjs_view_774_1707793552_9140036').show();" 33857 style="display: none" 33858 ><i class="fa fa-caret-down"></i> 33859 <span class="hover_link">Abstract</span></a 33860 > 33861 <div 33862 data-display-control="774_1707793552_9140036" 33863 id="773_1707793552_9139955" 33864 style="display: none" 33865 > 33866 <div class="arrow-slidedown"> 33867 <blockquote> 33868 The term "digital twin" is akin to a chameleon 33869 in the industry, adopting various meanings and 33870 causing widespread confusion. In this 33871 presentation, we embark on a mission to 33872 demystify digital twins, categorize their 33873 diverse implementations, explore the realm of 33874 simulations, and unveil the myriad uses of this 33875 transformative technology. We explore the 33876 differences and benefits of each type with 33877 special emphasis on data-driven digital twins 33878 and their integration with AI (Artificial 33879 Intelligence), ML (Machine Learning) and DL 33880 (Deep Learning) technologies. 33881 www.createasoft.com 33882 </blockquote> 33883 </div> 33884 </div> 33885 </div> 33886 </div> 33887 <div class="slot-urls"></div> 33888 </div> 33889 </div> 33890 <div class="session-entry"> 33891 <span class="session-event-type">Vendor Session</span 33892 ><span class="type-track-spacer"> · </span 33893 ><span class="program-track">Vendor</span><br /> 33894 <div class="session-title"> 33895 Implementing Simulation Projects 33896 </div> 33897 <div class="session-chair"> 33898 Chair: John Shortle (George Mason University)<br /> 33899 </div> 33900 <div class="slot-entry"> 33901 <a name="vdra102" tabindex="-1"></a> 33902 <div class="slot-title-line"> 33903 <span class="slot-title" 33904 >Overcoming Real-world Challenges on Simulation 33905 Projects</span 33906 > 33907 </div> 33908 <div class="slot-authors"> 33909 Saurabh Parakh, Amy Greer, and Yusuke Legard (MOSIMTEC, 33910 LLC) 33911 </div> 33912 <div class="slot-abstract"> 33913 <div> 33914 <a 33915 class="clickable no-decoration" 33916 id="vhsjs_view_776_1707793552_9172266" 33917 onclick="$('#vhsjs_view_776_1707793552_9172266').hide(); 33918 $('#vhsjs_hide_776_1707793552_9172266').show(); 33919 $('#775_1707793552_917218').slideDown(function() { 33920 if (typeof Masonry === 'function') { 33921 $('.use_masonry').masonry(); 33922 }; 33923 33924 });" 33925 ><i class="fa fa-caret-right"></i>
33926 <span class="hover_link">Abstract</span></a 33927 ><a 33928 class="clickable no-decoration" 33929 id="vhsjs_hide_776_1707793552_9172266" 33930 onclick="$('#775_1707793552_917218').hide(function() { 33931 if (typeof Masonry === 'function') { 33932 $('.use_masonry').masonry(); 33933 }; 33934 }); 33935 $('#vhsjs_hide_776_1707793552_9172266').hide(); 33936 $('#vhsjs_view_776_1707793552_9172266').show();" 33937 style="display: none" 33938 ><i class="fa fa-caret-down"></i> 33939 <span class="hover_link">Abstract</span></a 33940 > 33941 <div 33942 data-display-control="776_1707793552_9172266" 33943 id="775_1707793552_917218" 33944 style="display: none" 33945 > 33946 <div class="arrow-slidedown"> 33947 <blockquote> 33948 MOSIMTEC expertly guides clients – from 33949 pharma to farming, from climate change to change 33950 management – through simulation modeling 33951 so they get the MOST knowledge, the MOST 33952 insight, and the MOST intelligent answers to 33953 Future Proof their Business. At this vendor 33954 track presentation, MOSIMTEC consultants will be 33955 sharing stories from implementing commercial 33956 simulation projects, along with tips for 33957 addressing real world challenges related to 33958 project management, stakeholder buy-in, project 33959 deadlines, and data scarcity. 33960 </blockquote> 33961 </div> 33962 </div> 33963 </div> 33964 </div> 33965 <div class="slot-urls"></div> 33966 </div> 33967 </div> 33968 </div> 33969 <div class="centered"> 33970 <div class="top-link"><a href="#top">Return to Top</a></div> 33971 </div> 33972 <hr /> 33973 </div> 33974 <div class="area-section"> 33975 <div class="centered"> 33976 <a name="ptrack138" tabindex="-1"></a> 33977 <div class="section-title">Poster</div> 33978 </div> 33979 <div class="section-entry"> 33980 <div class="session-entry"> 33981 <span class="session-event-type">Poster</span 33982 ><span class="type-track-spacer"> · </span 33983 ><span class="program-track">Poster</span><br /> 33984 <div class="session-title"> 33985 Poster Track Lightning Presentations 33986 </div> 33987 <div class="session-chair"> 33988 Chair: Zeyu Zheng (University of California, Berkeley); 33989 María Julia Blas (INGAR CONICET UTN)<br /> 33990 </div> 33991 <div class="slot-entry"> 33992 <a name="pos104" tabindex="-1"></a> 33993 <div class="slot-title-line"> 33994 <span class="slot-title" 33995 >Using Narratives to Facilitate Public Acceptance of 33996 Policies through Agent-Based Simulations</span 33997 > 33998 </div> 33999 <div class="slot-authors"> 34000 Yusuke Goto (Shibaura Institute of Technology) 34001 </div> 34002 <div class="slot-abstract"> 34003 <div> 34004 <a 34005 class="clickable no-decoration" 34006 id="vhsjs_view_778_1707793552_9306083" 34007 onclick="$('#vhsjs_view_778_1707793552_9306083').hide(); 34008 $('#vhsjs_hide_778_1707793552_9306083').show(); 34009 $('#777_1707793552_9305995').slideDown(function() { 34010 if (typeof Masonry === 'function') { 34011 $('.use_masonry').masonry(); 34012 }; 34013 34014 });" 34015 ><i class="fa fa-caret-right"></i>
34016 <span class="hover_link">Abstract</span></a 34017 ><a 34018 class="clickable no-decoration" 34019 id="vhsjs_hide_778_1707793552_9306083" 34020 onclick="$('#777_1707793552_9305995').hide(function() { 34021 if (typeof Masonry === 'function') { 34022 $('.use_masonry').masonry(); 34023 }; 34024 }); 34025 $('#vhsjs_hide_778_1707793552_9306083').hide(); 34026 $('#vhsjs_view_778_1707793552_9306083').show();" 34027 style="display: none" 34028 ><i class="fa fa-caret-down"></i> 34029 <span class="hover_link">Abstract</span></a 34030 > 34031 <div 34032 data-display-control="778_1707793552_9306083" 34033 id="777_1707793552_9305995" 34034 style="display: none" 34035 > 34036 <div class="arrow-slidedown"> 34037 <blockquote> 34038 In this paper, we introduce a conceptual 34039 framework of policy communication that is 34040 propelled by narratives generated via 34041 agent-based simulations. The framework 34042 demonstrates that public acceptance of polices 34043 is contingent upon the interplay between 34044 generated narratives and the stakeholders who 34045 receive them. Moreover, it illustrates a model 34046 that employs narratives to facilitate public 34047 acceptance of policies through agent-based 34048 simulations. Drawing on the proposed framework, 34049 we identify the following three challenges 34050 encountered in policy communication that is 34051 driven by narratives generated through 34052 agent-based simulations: developing a 34053 methodology of narrative design and 34054 visualization, identifying factors that 34055 influence public acceptance of policies, and 34056 providing the assurance of accountability as 34057 justified narratives. 34058 </blockquote> 34059 </div> 34060 </div> 34061 </div> 34062 </div> 34063 <div class="slot-urls"></div> 34064 <a href="/wsc23papers/pos104.pdf" target="_blank">pdf</a 34065 ><br /> 34066 </div> 34067 <div class="slot-entry"> 34068 <a name="pos105" tabindex="-1"></a> 34069 <div class="slot-title-line"> 34070 <span class="slot-title" 34071 >Digital Twin Readiness Assessment: Case Study at a 34072 Printing Company</span 34073 > 34074 </div> 34075 <div class="slot-authors"> 34076 Jānis Grabis (Riga Technical University) 34077 </div> 34078 <div class="slot-abstract"> 34079 <div> 34080 <a 34081 class="clickable no-decoration" 34082 id="vhsjs_view_780_1707793552_952876" 34083 onclick="$('#vhsjs_view_780_1707793552_952876').hide(); 34084 $('#vhsjs_hide_780_1707793552_952876').show(); 34085 $('#779_1707793552_952868').slideDown(function() { 34086 if (typeof Masonry === 'function') { 34087 $('.use_masonry').masonry(); 34088 }; 34089 34090 });" 34091 ><i class="fa fa-caret-right"></i>
34092 <span class="hover_link">Abstract</span></a 34093 ><a 34094 class="clickable no-decoration" 34095 id="vhsjs_hide_780_1707793552_952876" 34096 onclick="$('#779_1707793552_952868').hide(function() { 34097 if (typeof Masonry === 'function') { 34098 $('.use_masonry').masonry(); 34099 }; 34100 }); 34101 $('#vhsjs_hide_780_1707793552_952876').hide(); 34102 $('#vhsjs_view_780_1707793552_952876').show();" 34103 style="display: none" 34104 ><i class="fa fa-caret-down"></i> 34105 <span class="hover_link">Abstract</span></a 34106 > 34107 <div 34108 data-display-control="780_1707793552_952876" 34109 id="779_1707793552_952868" 34110 style="display: none" 34111 > 34112 <div class="arrow-slidedown"> 34113 <blockquote> 34114 Digital twins provide a way to control various 34115 manufacturing processes. To justify their 34116 implementation investment, a systematic 34117 readiness assessment is conducted at a printing 34118 company. The assessment highlights readiness 34119 gaps and provides basis for further 34120 implementation of digital twin technology. Three 34121 implementation scenarios are elaborated and 34122 evaluated jointly with the company’s 34123 representatives. A digital twin solution for 34124 optimization of the folding process to improve 34125 delivery time estimation is selected for further 34126 implementation. 34127 </blockquote> 34128 </div> 34129 </div> 34130 </div> 34131 </div> 34132 <div class="slot-urls"></div> 34133 <a href="/wsc23papers/pos105.pdf" target="_blank">pdf</a 34134 ><br /> 34135 </div> 34136 <div class="slot-entry"> 34137 <a name="pos106" tabindex="-1"></a> 34138 <div class="slot-title-line"> 34139 <span class="slot-title" 34140 >Constructing an ABM to Enhance Residents' Conviction 34141 Regarding the Effectiveness of Town Development 34142 Measures</span 34143 > 34144 </div> 34145 <div class="slot-authors"> 34146 Ibu Ueno and Shingo Takahashi (Waseda University) 34147 </div> 34148 <div class="slot-abstract"> 34149 <div> 34150 <a 34151 class="clickable no-decoration" 34152 id="vhsjs_view_782_1707793552_9553628" 34153 onclick="$('#vhsjs_view_782_1707793552_9553628').hide(); 34154 $('#vhsjs_hide_782_1707793552_9553628').show(); 34155 $('#781_1707793552_9553547').slideDown(function() { 34156 if (typeof Masonry === 'function') { 34157 $('.use_masonry').masonry(); 34158 }; 34159 34160 });" 34161 ><i class="fa fa-caret-right"></i> 34162 <span class="hover_link">Abstract</span></a 34163 ><a 34164 class="clickable no-decoration" 34165 id="vhsjs_hide_782_1707793552_9553628" 34166 onclick="$('#781_1707793552_9553547').hide(function() { 34167 if (typeof Masonry === 'function') { 34168 $('.use_masonry').masonry(); 34169 }; 34170 }); 34171 $('#vhsjs_hide_782_1707793552_9553628').hide(); 34172 $('#vhsjs_view_782_1707793552_9553628').show();" 34173 style="display: none" 34174 ><i class="fa fa-caret-down"></i>
34175 <span class="hover_link">Abstract</span></a 34176 > 34177 <div 34178 data-display-control="782_1707793552_9553628" 34179 id="781_1707793552_9553547" 34180 style="display: none" 34181 > 34182 <div class="arrow-slidedown"> 34183 <blockquote> 34184 When evaluating town development measures, 34185 social simulations have been attempted to be 34186 employed. In recent years, it is essential to 34187 involve diverse stakeholders in the modeling 34188 process and feedback of simulation results. This 34189 paper aims to construct a method using Gaming 34190 Simulation (GS) to allow participants to 34191 experience an Agent-Based Model (ABM), 34192 comprehend the model, and gain a sense of 34193 convincing from the simulation results. 34194 </blockquote> 34195 </div> 34196 </div> 34197 </div> 34198 </div> 34199 <div class="slot-urls"></div> 34200 <a href="/wsc23papers/pos106.pdf" target="_blank">pdf</a 34201 ><br /> 34202 </div> 34203 <div class="slot-entry"> 34204 <a name="pos108" tabindex="-1"></a> 34205 <div class="slot-title-line"> 34206 <span class="slot-title" 34207 >Integrated Modeling and Optimization of Spare Part 34208 Logistic Operations and Condition-based Maintenance 34209 Policies in a System of Geographically Distributed 34210 Assets</span 34211 > 34212 </div> 34213 <div class="slot-authors"> 34214 Po-Han Wang and Dragan Djurdjanovic (The University of 34215 Texas at Austin) 34216 </div> 34217 <div class="slot-abstract"> 34218 <div> 34219 <a 34220 class="clickable no-decoration" 34221 id="vhsjs_view_784_1707793552_9580333" 34222 onclick="$('#vhsjs_view_784_1707793552_9580333').hide(); 34223 $('#vhsjs_hide_784_1707793552_9580333').show(); 34224 $('#783_1707793552_958026').slideDown(function() { 34225 if (typeof Masonry === 'function') { 34226 $('.use_masonry').masonry(); 34227 }; 34228 34229 });" 34230 ><i class="fa fa-caret-right"></i> 34231 <span class="hover_link">Abstract</span></a 34232 ><a 34233 class="clickable no-decoration" 34234 id="vhsjs_hide_784_1707793552_9580333" 34235 onclick="$('#783_1707793552_958026').hide(function() { 34236 if (typeof Masonry === 'function') { 34237 $('.use_masonry').masonry(); 34238 }; 34239 }); 34240 $('#vhsjs_hide_784_1707793552_9580333').hide(); 34241 $('#vhsjs_view_784_1707793552_9580333').show();" 34242 style="display: none" 34243 ><i class="fa fa-caret-down"></i> 34244 <span class="hover_link">Abstract</span></a 34245 > 34246 <div 34247 data-display-control="784_1707793552_9580333" 34248 id="783_1707793552_958026" 34249 style="display: none" 34250 > 34251 <div class="arrow-slidedown"> 34252 <blockquote> 34253 This study presents joint optimization of Spare 34254 Parts Logistics (SPL) operations with 34255 condition-based maintenance (CBM) policies in a 34256 system of geographically distributed assets, 34257 each consisting of multiple degrading parts. The 34258 model considers facility location selection, 34259 network connectivity design, inventory levels 34260 for replenishment triggering, and CBM policies 34261 that minimize overall system operating costs. 34262 The solution is implemented as a sequential 34263 model consisting of two stages: the initial 34264 stage utilizes mathematical programming for 34265 facility location selection and network design. 34266 It is followed by a simulation-based method 34267 using Continuous Time Markov Chain to model 34268 degradation of spare parts and link it with 34269 inventory managements. Additionally, the 34270 maintenance operations model includes 34271 opportunistic maintenance, which enables further 34272 reduction of operating costs. Overall, the newly 34273 proposed approach addresses scale limitations 34274 and overly restrictive simplifications of 34275 previously published models, which enables a 34276 more comprehensive operational decision-making. 34277 </blockquote> 34278 </div> 34279 </div> 34280 </div> 34281 </div> 34282 <div class="slot-urls"></div> 34283 <a href="/wsc23papers/pos108.pdf" target="_blank">pdf</a 34284 ><br /> 34285 </div> 34286 <div class="slot-entry"> 34287 <a name="pos110" tabindex="-1"></a> 34288 <div class="slot-title-line">
34289 <span class="slot-title" 34290 >Potential Impact of a Diagnostic Test for Detecting 34291 Prepatent Guinea Worm Infections in Dogs</span 34292 > 34293 </div> 34294 <div class="slot-authors"> 34295 Hannah Smalley and Pinar Keskinocak (Georgia Institute 34296 of Technology); Julie Swann (North Carolina State 34297 University); Christopher Hanna (Global Project Partners, 34298 LLC); and Adam Weiss (The Carter Center) 34299 </div> 34300 <div class="slot-abstract"> 34301 <div> 34302 <a 34303 class="clickable no-decoration" 34304 id="vhsjs_view_786_1707793552_9606686" 34305 onclick="$('#vhsjs_view_786_1707793552_9606686').hide(); 34306 $('#vhsjs_hide_786_1707793552_9606686').show(); 34307 $('#785_1707793552_9606607').slideDown(function() { 34308 if (typeof Masonry === 'function') { 34309 $('.use_masonry').masonry(); 34310 }; 34311 34312 });" 34313 ><i class="fa fa-caret-right"></i> 34314 <span class="hover_link">Abstract</span></a 34315 ><a 34316 class="clickable no-decoration" 34317 id="vhsjs_hide_786_1707793552_9606686" 34318 onclick="$('#785_1707793552_9606607').hide(function() { 34319 if (typeof Masonry === 'function') { 34320 $('.use_masonry').masonry(); 34321 }; 34322 }); 34323 $('#vhsjs_hide_786_1707793552_9606686').hide(); 34324 $('#vhsjs_view_786_1707793552_9606686').show();" 34325 style="display: none" 34326 ><i class="fa fa-caret-down"></i> 34327 <span class="hover_link">Abstract</span></a 34328 > 34329 <div 34330 data-display-control="786_1707793552_9606686" 34331 id="785_1707793552_9606607" 34332 style="display: none" 34333 > 34334 <div class="arrow-slidedown"> 34335 <blockquote> 34336 Chad has seen a considerable reduction in cases 34337 of Guinea worm disease (or dracunculiasis) in 34338 domestic dogs in recent years but accelerating 34339 elimination of the disease may require 34340 additional tools. We investigate the potential 34341 benefits of a hypothetical diagnostic test 34342 capable of detecting pre-patent infections in 34343 dogs. We adapted an agent-based simulation model 34344 for analyzing disease transmission to examine 34345 the interaction of multiple test factors 34346 including sensitivity and specificity, infection 34347 detection timing, dog selection, and tethering 34348 compliance behaviors. We find that a diagnostic 34349 test could be successful in combination with 34350 existing interventions, and elimination can be 34351 achieved within two years with 80% or higher 34352 test sensitivity, 90% or higher specificity, 34353 systematic testing of each dog biannually, and 34354 long-term tethering of test-positive dogs. Due 34355 to the long incubation period (10-14 months) and 34356 lack of treatment, the testing rollout and 34357 response of dog owners are critical to the 34358 benefits of the test. 34359 </blockquote> 34360 </div> 34361 </div> 34362 </div> 34363 </div> 34364 <div class="slot-urls"></div> 34365 <a href="/wsc23papers/pos110.pdf" target="_blank">pdf</a 34366 ><br /> 34367 </div> 34368 <div class="slot-entry"> 34369 <a name="pos111" tabindex="-1"></a> 34370 <div class="slot-title-line">
34371 <span class="slot-title" 34372 >A Framework for Dynamic Control of Combat Support 34373 Exercises</span 34374 > 34375 </div> 34376 <div class="slot-authors"> 34377 Sean McCarty (Air Force Institute of Technology) 34378 </div> 34379 <div class="slot-abstract"> 34380 <div> 34381 <a 34382 class="clickable no-decoration" 34383 id="vhsjs_view_788_1707793552_9631584" 34384 onclick="$('#vhsjs_view_788_1707793552_9631584').hide(); 34385 $('#vhsjs_hide_788_1707793552_9631584').show(); 34386 $('#787_1707793552_963142').slideDown(function() { 34387 if (typeof Masonry === 'function') { 34388 $('.use_masonry').masonry(); 34389 }; 34390 34391 });" 34392 ><i class="fa fa-caret-right"></i> 34393 <span class="hover_link">Abstract</span></a 34394 ><a 34395 class="clickable no-decoration" 34396 id="vhsjs_hide_788_1707793552_9631584" 34397 onclick="$('#787_1707793552_963142').hide(function() { 34398 if (typeof Masonry === 'function') { 34399 $('.use_masonry').masonry(); 34400 }; 34401 }); 34402 $('#vhsjs_hide_788_1707793552_9631584').hide(); 34403 $('#vhsjs_view_788_1707793552_9631584').show();" 34404 style="display: none" 34405 ><i class="fa fa-caret-down"></i> 34406 <span class="hover_link">Abstract</span></a 34407 > 34408 <div 34409 data-display-control="788_1707793552_9631584" 34410 id="787_1707793552_963142" 34411 style="display: none" 34412 > 34413 <div class="arrow-slidedown"> 34414 <blockquote> 34415 Future armed conflict will be characterized by 34416 surprise as adversaries innovate and evolve. 34417 Current exercises provide inadequate 34418 opportunities for combat support forces to 34419 improvise. This research proposes a framework 34420 for human-in-the-loop control of exercises using 34421 a graph network for modeling combined with 34422 topological analysis and modifications to the 34423 zero one scheduling formulation. This framework 34424 is assessed using the United States Air Force 34425 Silver Flag exercise as a case study with 34426 promising results. 34427 </blockquote> 34428 </div> 34429 </div> 34430 </div> 34431 </div> 34432 <div class="slot-urls"></div> 34433 <a href="/wsc23papers/pos111.pdf" target="_blank">pdf</a 34434 ><br /> 34435 </div> 34436 <div class="slot-entry"> 34437 <a name="pos112" tabindex="-1"></a> 34438 <div class="slot-title-line"> 34439 <span class="slot-title" 34440 >Information Diffusion Model of SNS and Visualization 34441 Method</span 34442 > 34443 </div> 34444 <div class="slot-authors"> 34445 Kazumi Sekiguchi and Masakazu Furuichi (Nihon 34446 University) 34447 </div> 34448 <div class="slot-abstract"> 34449 <div> 34450 <a 34451 class="clickable no-decoration" 34452 id="vhsjs_view_790_1707793552_965745" 34453 onclick="$('#vhsjs_view_790_1707793552_965745').hide(); 34454 $('#vhsjs_hide_790_1707793552_965745').show(); 34455 $('#789_1707793552_965737').slideDown(function() { 34456 if (typeof Masonry === 'function') { 34457 $('.use_masonry').masonry(); 34458 }; 34459 34460 });" 34461 ><i class="fa fa-caret-right"></i>
34462 <span class="hover_link">Abstract</span></a 34463 ><a 34464 class="clickable no-decoration" 34465 id="vhsjs_hide_790_1707793552_965745" 34466 onclick="$('#789_1707793552_965737').hide(function() { 34467 if (typeof Masonry === 'function') { 34468 $('.use_masonry').masonry(); 34469 }; 34470 }); 34471 $('#vhsjs_hide_790_1707793552_965745').hide(); 34472 $('#vhsjs_view_790_1707793552_965745').show();" 34473 style="display: none" 34474 ><i class="fa fa-caret-down"></i> 34475 <span class="hover_link">Abstract</span></a 34476 > 34477 <div 34478 data-display-control="790_1707793552_965745" 34479 id="789_1707793552_965737" 34480 style="display: none" 34481 > 34482 <div class="arrow-slidedown"> 34483 <blockquote> 34484 The dissemination of social media has led to the 34485 explosion of fake news, other misinformation and 34486 disinformation, which significantly impacts 34487 society. They are sometimes based on information 34488 transmission by individuals, groups, and 34489 organizations. In order to analyze the influence 34490 of information diffusion, it is necessary not 34491 only to visualize the spread from a bird's eye 34492 view but also to examine the characteristics of 34493 local information propagation and the impact of 34494 the behavior. In this study, we developed a 34495 multi-agent information diffusion model of 34496 social networking service (SNS). We investigated 34497 a visualization method that simultaneously 34498 grasps the local information diffusion by 34499 individuals and the overarching information 34500 spread by multiple user clusters. This method 34501 facilitates the recognition of the information 34502 diffusion within a group and the final dispersal 34503 status in addition to the condition of 34504 information dissemination by each individual. 34505 </blockquote> 34506 </div> 34507 </div> 34508 </div> 34509 </div> 34510 <div class="slot-urls"></div> 34511 <a href="/wsc23papers/pos112.pdf" target="_blank">pdf</a 34512 ><br /> 34513 </div> 34514 <div class="slot-entry"> 34515 <a name="pos114" tabindex="-1"></a> 34516 <div class="slot-title-line"> 34517 <span class="slot-title" 34518 >Using a Discrete Event Simulation to Improve Check-in 34519 Operations at the Port of Dover</span 34520 > 34521 </div> 34522 <div class="slot-authors"> 34523 Siti Fariya (University of Kent, The Port of Dover); 34524 Kathy Kotiadis (University of Kent); Timothy van Vugt 34525 (The Port of Dover); and Jesse O'Hanley (University of 34526 Kent) 34527 </div> 34528 <div class="slot-abstract"> 34529 <div> 34530 <a 34531 class="clickable no-decoration" 34532 id="vhsjs_view_792_1707793552_9680367" 34533 onclick="$('#vhsjs_view_792_1707793552_9680367').hide(); 34534 $('#vhsjs_hide_792_1707793552_9680367').show(); 34535 $('#791_1707793552_968029').slideDown(function() { 34536 if (typeof Masonry === 'function') { 34537 $('.use_masonry').masonry(); 34538 }; 34539 34540 });" 34541 ><i class="fa fa-caret-right"></i>
34542 <span class="hover_link">Abstract</span></a 34543 ><a 34544 class="clickable no-decoration" 34545 id="vhsjs_hide_792_1707793552_9680367" 34546 onclick="$('#791_1707793552_968029').hide(function() { 34547 if (typeof Masonry === 'function') { 34548 $('.use_masonry').masonry(); 34549 }; 34550 }); 34551 $('#vhsjs_hide_792_1707793552_9680367').hide(); 34552 $('#vhsjs_view_792_1707793552_9680367').show();" 34553 style="display: none" 34554 ><i class="fa fa-caret-down"></i> 34555 <span class="hover_link">Abstract</span></a 34556 > 34557 <div 34558 data-display-control="792_1707793552_9680367" 34559 id="791_1707793552_968029" 34560 style="display: none" 34561 > 34562 <div class="arrow-slidedown"> 34563 <blockquote> 34564 This paper showcases our use of discrete event 34565 simulation (DES) to enhance check-in operations 34566 at the Port of Dover (PoD). PoD is the busiest 34567 international ferry port in the UK and since the 34568 UK left the European Union, the port has 34569 experienced increased processing times and 34570 considerable delays in passenger check-in. Three 34571 independent ferry operators run individual 34572 check-in systems for freight and tourist 34573 vehicles, leading to efficiency challenges, 34574 notably prolonged queuing times and limited 34575 throughput. Our study investigates two 34576 alternatives: a common check-in booth for all 34577 operators and vehicle types, and a system that 34578 retains operator-specific booths but merges the 34579 process for all traffic types. We aim to 34580 identify an improved operational model that 34581 reduces queue times and to explore a range of 34582 solutions that could improve check-in operations 34583 at the Port of Dover, which not only make the 34584 check-in process more efficient but also 34585 significantly reduces queuing times. 34586 </blockquote> 34587 </div> 34588 </div> 34589 </div> 34590 </div> 34591 <div class="slot-urls"></div> 34592 <a href="/wsc23papers/pos114.pdf" target="_blank">pdf</a 34593 ><br /> 34594 </div> 34595 <div class="slot-entry"> 34596 <a name="pos115" tabindex="-1"></a> 34597 <div class="slot-title-line"> 34598 <span class="slot-title" 34599 >Development and Application of the One-Stop Flow 34600 Analysis Framework Enabling Rapid Digital 34601 Engineering</span 34602 > 34603 </div> 34604 <div class="slot-authors"> 34605 Kengo Asada, Yuichi Matsuo, and Kozo Fujii (Tokyo 34606 University of Science) 34607 </div> 34608 <div class="slot-abstract"> 34609 <div> 34610 <a 34611 class="clickable no-decoration" 34612 id="vhsjs_view_794_1707793552_9707246" 34613 onclick="$('#vhsjs_view_794_1707793552_9707246').hide(); 34614 $('#vhsjs_hide_794_1707793552_9707246').show(); 34615 $('#793_1707793552_9707167').slideDown(function() { 34616 if (typeof Masonry === 'function') { 34617 $('.use_masonry').masonry(); 34618 }; 34619 34620 });" 34621 ><i class="fa fa-caret-right"></i>
34622 <span class="hover_link">Abstract</span></a 34623 ><a 34624 class="clickable no-decoration" 34625 id="vhsjs_hide_794_1707793552_9707246" 34626 onclick="$('#793_1707793552_9707167').hide(function() { 34627 if (typeof Masonry === 'function') { 34628 $('.use_masonry').masonry(); 34629 }; 34630 }); 34631 $('#vhsjs_hide_794_1707793552_9707246').hide(); 34632 $('#vhsjs_view_794_1707793552_9707246').show();" 34633 style="display: none" 34634 ><i class="fa fa-caret-down"></i> 34635 <span class="hover_link">Abstract</span></a 34636 > 34637 <div 34638 data-display-control="794_1707793552_9707246" 34639 id="793_1707793552_9707167" 34640 style="display: none" 34641 > 34642 <div class="arrow-slidedown"> 34643 <blockquote> 34644 This paper proposes a one-stop simulation 34645 framework from point cloud acquisition through 34646 flow analysis. Conventional flow analysis starts 34647 with computer-aided design (CAD) software to 34648 define the object shape and any mesh generator 34649 to build computational grids. However, CAD data 34650 of old buildings and rooms, including furniture, 34651 is hardly available. Thus, CAD data creation, 34652 which takes a lot of time, is required when 34653 conducting flow simulations of existing 34654 buildings first. The present study illustrates a 34655 simplified flow analysis procedure, which 34656 reduces this lead time by defining the object 34657 shape with point clouds and using a 34658 Cartesian-based flow solver. The proposed 34659 framework simplifies the design of heating, 34660 ventilation, and air conditioning (HVAC) and 34661 could improve its existing process and quality. 34662 </blockquote> 34663 </div> 34664 </div> 34665 </div> 34666 </div> 34667 <div class="slot-urls"></div> 34668 <a href="/wsc23papers/pos115.pdf" target="_blank">pdf</a 34669 ><br /> 34670 </div> 34671 <div class="slot-entry"> 34672 <a name="pos116" tabindex="-1"></a> 34673 <div class="slot-title-line"> 34674 <span class="slot-title" 34675 >Stochastically Constrained Level Set Approximation 34676 Via Probabilistic Branch and Bound</span 34677 > 34678 </div> 34679 <div class="slot-authors"> 34680 Hao Huang (Yuan Ze University), Shing Chih Tsai 34681 (National Cheng Kung University), and Chuljin Park 34682 (Hanyang University) 34683 </div> 34684 <div class="slot-abstract"> 34685 <div> 34686 <a 34687 class="clickable no-decoration" 34688 id="vhsjs_view_796_1707793552_9732878" 34689 onclick="$('#vhsjs_view_796_1707793552_9732878').hide(); 34690 $('#vhsjs_hide_796_1707793552_9732878').show(); 34691 $('#795_1707793552_9732797').slideDown(function() { 34692 if (typeof Masonry === 'function') { 34693 $('.use_masonry').masonry(); 34694 }; 34695 34696 });" 34697 ><i class="fa fa-caret-right"></i>
34698 <span class="hover_link">Abstract</span></a 34699 ><a 34700 class="clickable no-decoration" 34701 id="vhsjs_hide_796_1707793552_9732878" 34702 onclick="$('#795_1707793552_9732797').hide(function() { 34703 if (typeof Masonry === 'function') { 34704 $('.use_masonry').masonry(); 34705 }; 34706 }); 34707 $('#vhsjs_hide_796_1707793552_9732878').hide(); 34708 $('#vhsjs_view_796_1707793552_9732878').show();" 34709 style="display: none" 34710 ><i class="fa fa-caret-down"></i> 34711 <span class="hover_link">Abstract</span></a 34712 > 34713 <div 34714 data-display-control="796_1707793552_9732878" 34715 id="795_1707793552_9732797" 34716 style="display: none" 34717 > 34718 <div class="arrow-slidedown"> 34719 <blockquote> 34720 This paper investigates a simulation 34721 optimization problem with both stoch
34721astic 34722 objective and constraint functions with a 34723 discrete solution space. Our objective is to 34724 identify a set of near-optimal solutions within 34725 a specific quantile, such as the top 10%. To 34726 achieve this goal, we first employs a 34727 probabilistic branch-and-bound algorithm to find 34728 a level set of solutions. Then, we combine a 34729 penalty function approach with the probabilistic 34730 branch-and-bound algorithm to handle 34731 stochastically constrained problems. Both 34732 convergence analysis and experimental results 34733 are provided that demonstrate the superior 34734 efficiency of our proposed approaches over 34735 existing methods. 34736 </blockquote> 34737 </div> 34738 </div> 34739 </div> 34740 </div> 34741 <div class="slot-urls"></div> 34742 <a href="/wsc23papers/pos116.pdf" target="_blank">pdf</a 34743 ><br /> 34744 </div> 34745 <div class="slot-entry"> 34746 <a name="pos120" tabindex="-1"></a> 34747 <div class="slot-title-line"> 34748 <span class="slot-title" 34749 >A Standardized Method for Building Simulation-based 34750 Decision Support Systems Using High Level 34751 Architecture</span 34752 > 34753 </div> 34754 <div class="slot-authors"> 34755 Rana Ead, Yasser Mohamed, and Simaan AbouRizk 34756 (University of Alberta) 34757 </div> 34758 <div class="slot-abstract"> 34759 <div> 34760 <a 34761 class="clickable no-decoration" 34762 id="vhsjs_view_798_1707793552_9767175" 34763 onclick="$('#vhsjs_view_798_1707793552_9767175').hide(); 34764 $('#vhsjs_hide_798_1707793552_9767175').show(); 34765 $('#797_1707793552_9767098').slideDown(function() { 34766 if (typeof Masonry === 'function') { 34767 $('.use_masonry').masonry(); 34768 }; 34769 34770 });" 34771 ><i class="fa fa-caret-right"></i> 34772 <span class="hover_link">Abstract</span></a 34773 ><a 34774 class="clickable no-decoration" 34775 id="vhsjs_hide_798_1707793552_9767175" 34776 onclick="$('#797_1707793552_9767098').hide(function() { 34777 if (typeof Masonry === 'function') { 34778 $('.use_masonry').masonry(); 34779 }; 34780 }); 34781 $('#vhsjs_hide_798_1707793552_9767175').hide(); 34782 $('#vhsjs_view_798_1707793552_9767175').show();" 34783 style="display: none" 34784 ><i class="fa fa-caret-down"></i> 34785 <span class="hover_link">Abstract</span></a 34786 > 34787 <div 34788 data-display-control="798_1707793552_9767175" 34789 id="797_1707793552_9767098" 34790 style="display: none" 34791 > 34792 <div class="arrow-slidedown"> 34793 <blockquote> 34794 This research develops a standardized Federation 34795 Object Model (FOM) for Simulation-Based 34796 Decision-Support Systems (SB-DSS) in 34797 construction. SB-DSS are vital for tackling 34798 project complexities, but their development 34799 requires considerable time and expertise, 34800 leading to underdeveloped systems and limited 34801 adoption. To address this, the study adopts
34802 High-Level Architecture (HLA) standards, 34803 integrating autonomous simulations into a single 34804 distributed simulation. The FOM includes object 34805 classes, interactions, and datatype definitions, 34806 enabling efficient communication among 34807 federates. The initial FOM version was 34808 successfully tested with five federates, 34809 demonstrating its effectiveness. This 34810 standardized FOM promotes simulation 34811 reusability, interoperability, and data-driven 34812 decision-making, ultimately enhancing 34813 construction project execution and 34814 competitiveness. 34815 </blockquote> 34816 </div> 34817 </div> 34818 </div> 34819 </div> 34820 <div class="slot-urls"></div> 34821 <a href="/wsc23papers/pos120.pdf" target="_blank">pdf</a 34822 ><br /> 34823 </div> 34824 <div class="slot-entry"> 34825 <a name="pos121" tabindex="-1"></a> 34826 <div class="slot-title-line"> 34827 <span class="slot-title" 34828 >The Growth of Generative AI: Hype, Harm, and 34829 Control</span 34830 > 34831 </div> 34832 <div class="slot-authors"> 34833 Timothy Clancy (Dialectic Simulations); Asmeret Naugle 34834 (Sandia National Laboratories); and Ignacio J. 34835 Martinez-Moyano (Argonne National Laboratory, University 34836 of Chicago) 34837 </div> 34838 <div class="slot-abstract"> 34839 <div> 34840 <a 34841 class="clickable no-decoration" 34842 id="vhsjs_view_800_1707793552_9805233" 34843 onclick="$('#vhsjs_view_800_1707793552_9805233').hide(); 34844 $('#vhsjs_hide_800_1707793552_9805233').show(); 34845 $('#799_1707793552_9805148').slideDown(function() { 34846 if (typeof Masonry === 'function') { 34847 $('.use_masonry').masonry(); 34848 }; 34849 34850 });" 34851 ><i class="fa fa-caret-right"></i> 34852 <span class="hover_link">Abstract</span></a 34853 ><a 34854 class="clickable no-decoration" 34855 id="vhsjs_hide_800_1707793552_9805233" 34856 onclick="$('#799_1707793552_9805148').hide(function() { 34857 if (typeof Masonry === 'function') { 34858 $('.use_masonry').masonry(); 34859 }; 34860 }); 34861 $('#vhsjs_hide_800_1707793552_9805233').hide(); 34862 $('#vhsjs_view_800_1707793552_9805233').show();" 34863 style="display: none" 34864 ><i class="fa fa-caret-down"></i> 34865 <span class="hover_link">Abstract</span></a 34866 > 34867 <div 34868 data-display-control="800_1707793552_9805233" 34869 id="799_1707793552_9805148" 34870 style="display: none" 34871 > 34872 <div class="arrow-slidedown"> 34873 <blockquote> 34874 The hype-harm-control model investigates the 34875 societal impact of generative artificial 34876 intelligence (AI), given its growth, alignment 34877 with societal values, and controls. This system 34878 dynamics model was used to simulate the dynamics 34879 and impacts of generative AI over a 10-year time 34880 horizon. As the generative AI grows, hype and 34881 use increase, leading to both societal benefit 34882 and societal harm. This analysis found that 34883 while the balance of hype and societal harm 34884 determines the controls put on AI development 34885 and use, early societal harm creates a strong 34886 incentive to implement societal controls that 34887 limit the growth of generative AI overall. 34888 </blockquote> 34889 </div> 34890 </div> 34891 </div> 34892 </div> 34893 <div class="slot-urls"></div> 34894 <a href="/wsc23papers/pos121.pdf" target="_blank">pdf</a 34895 ><br /> 34896 </div> 34897 <div class="slot-entry"> 34898 <a name="pos122" tabindex="-1"></a> 34899 <div class="slot-title-line">
34900 <span class="slot-title" 34901 >A Virtual Training System Using Digital Twins Based 34902 on Discrete Event System Formalism</span 34903 > 34904 </div> 34905 <div class="slot-authors"> 34906 JinWoo Kim, GyuSik Ham, Sooyoung Jang, and Changbeom 34907 Choi (Hanbat National University) 34908 </div> 34909 <div class="slot-abstract"> 34910 <div> 34911 <a 34912 class="clickable no-decoration" 34913 id="vhsjs_view_802_1707793552_9832652" 34914 onclick="$('#vhsjs_view_802_1707793552_9832652').hide(); 34915 $('#vhsjs_hide_802_1707793552_9832652').show(); 34916 $('#801_1707793552_983257').slideDown(function() { 34917 if (typeof Masonry === 'function') { 34918 $('.use_masonry').masonry(); 34919 }; 34920 34921 });" 34922 ><i class="fa fa-caret-right"></i> 34923 <span class="hover_link">Abstract</span></a 34924 ><a 34925 class="clickable no-decoration" 34926 id="vhsjs_hide_802_1707793552_9832652" 34927 onclick="$('#801_1707793552_983257').hide(function() { 34928 if (typeof Masonry === 'function') { 34929 $('.use_masonry').masonry(); 34930 }; 34931 }); 34932 $('#vhsjs_hide_802_1707793552_9832652').hide(); 34933 $('#vhsjs_view_802_1707793552_9832652').show();" 34934 style="display: none" 34935 ><i class="fa fa-caret-down"></i> 34936 <span class="hover_link">Abstract</span></a 34937 > 34938 <div 34939 data-display-control="802_1707793552_9832652" 34940 id="801_1707793552_983257" 34941 style="display: none" 34942 > 34943 <div class="arrow-slidedown"> 34944 <blockquote> 34945 With the advancement of technology in education 34946 and training, it has become commonplace to 34947 conduct virtual rather than physical training to 34948 save time and money. In addition, various 34949 training hardware and software have been 34950 proposed to give immersive experiences to 34951 trainees to enhance the training effects in 34952 various domains. The training system can be 34953 regarded as a digital twin system, which 34954 collects data from the trainee, analyzes the 34955 data in the cyber world, and gives proper 34956 feedback to the trainee. This research proposes 34957 a virtual training system using digital twins 34958 based on discrete event system formalism. 34959 Especially, we focus on developing a 34960 cost-effective digital twin and helping the 34961 trainer to develop an evaluation system by 34962 composing models. The training system utilizes 34963 the webcam to collect skeleton data from the 34964 trainee and evaluate the data by composing 34965 discrete event system models. 34966 </blockquote> 34967 </div> 34968 </div> 34969 </div> 34970 </div> 34971 <div class="slot-urls"></div> 34972 <a href="/wsc23papers/pos122.pdf" target="_blank">pdf</a 34973 ><br /> 34974 </div> 34975 <div class="slot-entry"> 34976 <a name="pos125" tabindex="-1"></a> 34977 <div class="slot-title-line">
34978 <span class="slot-title" 34979 >Development of Production Digital Twin in 34980 Manufacturing Using Fischertechnik Factory Model</span 34981 > 34982 </div> 34983 <div class="slot-authors"> 34984 Yuichi Matsuo, Kengo Asada, and Kozo Fujii (Tokyo 34985 University of Science) 34986 </div> 34987 <div class="slot-abstract"> 34988 <div> 34989 <a 34990 class="clickable no-decoration" 34991 id="vhsjs_view_804_1707793552_9858158" 34992 onclick="$('#vhsjs_view_804_1707793552_9858158').hide(); 34993 $('#vhsjs_hide_804_1707793552_9858158').show(); 34994 $('#803_1707793552_985808').slideDown(function() { 34995 if (typeof Masonry === 'function') { 34996 $('.use_masonry').masonry(); 34997 }; 34998 34999 });" 35000 ><i class="fa fa-caret-right"></i> 35001 <span class="hover_link">Abstract</span></a 35002 ><a 35003 class="clickable no-decoration" 35004 id="vhsjs_hide_804_1707793552_9858158" 35005 onclick="$('#803_1707793552_985808').hide(function() { 35006 if (typeof Masonry === 'function') { 35007 $('.use_masonry').masonry(); 35008 }; 35009 }); 35010 $('#vhsjs_hide_804_1707793552_9858158').hide(); 35011 $('#vhsjs_view_804_1707793552_9858158').show();" 35012 style="display: none" 35013 ><i class="fa fa-caret-down"></i> 35014 <span class="hover_link">Abstract</span></a 35015 > 35016 <div 35017 data-display-control="804_1707793552_9858158" 35018 id="803_1707793552_985808" 35019 style="display: none" 35020 > 35021 <div class="arrow-slidedown"> 35022 <blockquote> 35023 Recently, there have been more opportunities to 35024 see and hear the term Digital Twin (DT) in 35025 various situations. However, the reality is that 35026 only the concept of DT precedes and that there 35027 is a lack of places and materials to absorb the 35028 DT content and its implementation. This paper 35029 presents a case study at Tokyo University of 35030 Science to develop the Production Digital Twin 35031 in manufacturing by using Fischertechnik factory 35032 model and Matlab/Simulink software tool. DT can 35033 support not only the education in universities 35034 but also human resource development in 35035 manufacturing industries through the study and 35036 practice concerning production line 35037 optimization, virtual commissioning, 35038 cyber-physical system implementation, real-time 35039 monitoring of production data, and furthermore 35040 lead the innovation in manufacturing in Japan. 35041 </blockquote> 35042 </div> 35043 </div> 35044 </div> 35045 </div> 35046 <div class="slot-urls"></div> 35047 <a href="/wsc23papers/pos125.pdf" target="_blank">pdf</a 35048 ><br /> 35049 </div> 35050 <div class="slot-entry"> 35051 <a name="pos128" tabindex="-1"></a> 35052 <div class="slot-title-line">
35053 <span class="slot-title" 35054 >Optimal Computing Budget Allocation for Monte Carlo 35055 Tree Search in Othello</span 35056 > 35057 </div> 35058 <div class="slot-authors"> 35059 Daniel Qiu (Thomas Jefferson High School) and Jie Xu 35060 (George Mason University) 35061 </div> 35062 <div class="slot-abstract"> 35063 <div> 35064 <a 35065 class="clickable no-decoration" 35066 id="vhsjs_view_806_1707793552_9883788" 35067 onclick="$('#vhsjs_view_806_1707793552_9883788').hide(); 35068 $('#vhsjs_hide_806_1707793552_9883788').show(); 35069 $('#805_1707793552_988371').slideDown(function() { 35070 if (typeof Masonry === 'function') { 35071 $('.use_masonry').masonry(); 35072 }; 35073 35074 });" 35075 ><i class="fa fa-caret-right"></i> 35076 <span class="hover_link">Abstract</span></a 35077 ><a 35078 class="clickable no-decoration" 35079 id="vhsjs_hide_806_1707793552_9883788" 35080 onclick="$('#805_1707793552_988371').hide(function() { 35081 if (typeof Masonry === 'function') { 35082 $('.use_masonry').masonry(); 35083 }; 35084 }); 35085 $('#vhsjs_hide_806_1707793552_9883788').hide(); 35086 $('#vhsjs_view_806_1707793552_9883788').show();" 35087 style="display: none" 35088 ><i class="fa fa-caret-down"></i> 35089 <span class="hover_link">Abstract</span></a 35090 > 35091 <div 35092 data-display-control="806_1707793552_9883788" 35093 id="805_1707793552_988371" 35094 style="display: none" 35095 > 35096 <div class="arrow-slidedown"> 35097 <blockquote> 35098 Upper Confidence bounds applied to Trees (UCT) 35099 is the most popular tree policy for Monte Carlo 35100 Tree Search (MCTS). However, UCT focuses on 35101 minimizing cumulative regret rather than 35102 maximizing the Probability of Correct Selection 35103 (PCS) of the best action, which is often 35104 preferred in game engines. To address this, we 35105 examine an Optimal Computing Budget Allocation 35106 (OCBA) tree policy that provides a rigorous way 35107 for maximizing the PCS rather than minimizing 35108 regret. MCTS-OCBA has been shown to work well 35109 with simple games such as Tic-Tac-Toe, where the 35110 search space is small enough to simulate 35111 through, but not unsolved games such as Othello 35112 or Go. We report numerical results showing that 35113 MCTS-OCBA performs better in Othello than 35114 MCTS-UCT and thus demonstrate OCBA is a more 35115 efficient tree policy for MCTS for game engines. 35116 </blockquote> 35117 </div> 35118 </div> 35119 </div> 35120 </div> 35121 <div class="slot-urls"></div> 35122 <a href="/wsc23papers/pos128.pdf" target="_blank">pdf</a 35123 ><br /> 35124 </div> 35125 <div class="slot-entry"> 35126 <a name="pos129" tabindex="-1"></a> 35127 <div class="slot-title-line">
35128 <span class="slot-title" 35129 >An Efficient Simulation-Based Optimization Algorithm 35130 for a Crane Scheduling Problem in a Steelmaking 35131 Shop</span 35132 > 35133 </div> 35134 <div class="slot-authors"> 35135 Woo-Jin Shin and Hyun-Jung Kim (Korea Advanced Institute 35136 of Science and Technology) 35137 </div> 35138 <div class="slot-abstract"> 35139 <div> 35140 <a 35141 class="clickable no-decoration" 35142 id="vhsjs_view_808_1707793552_990891" 35143 onclick="$('#vhsjs_view_808_1707793552_990891').hide(); 35144 $('#vhsjs_hide_808_1707793552_990891').show(); 35145 $('#807_1707793552_9908834').slideDown(function() { 35146 if (typeof Masonry === 'function') { 35147 $('.use_masonry').masonry(); 35148 }; 35149 35150 });" 35151 ><i class="fa fa-caret-right"></i> 35152 <span class="hover_link">Abstract</span></a 35153 ><a 35154 class="clickable no-decoration" 35155 id="vhsjs_hide_808_1707793552_990891" 35156 onclick="$('#807_1707793552_9908834').hide(function() { 35157 if (typeof Masonry === 'function') { 35158 $('.use_masonry').masonry(); 35159 }; 35160 }); 35161 $('#vhsjs_hide_808_1707793552_990891').hide(); 35162 $('#vhsjs_view_808_1707793552_990891').show();" 35163 style="display: none" 35164 ><i class="fa fa-caret-down"></i> 35165 <span class="hover_link">Abstract</span></a 35166 > 35167 <div 35168 data-display-control="808_1707793552_990891" 35169 id="807_1707793552_9908834" 35170 style="display: none" 35171 > 35172 <div class="arrow-slidedown"> 35173 <blockquote> 35174 This study addresses a crane scheduling problem 35175 in a steelmaking shop, where cranes are 35176 responsible for transporting ladles with molten 35177 steel between machines. To meet production 35178 schedules, the coordination between cranes and 35179 machines is crucial, performing the 35180 transportation of ladles at appropriate times. 35181 Also, multiple cranes share a common track, 35182 interference between them must be avoided. To 35183 address this problem, we propose an efficient 35184 algorithm based on iterative simulations. 35185 Several dominance rules are developed to reduce 35186 the solution space and accelerate the 35187 convergence of the algorithm. Experimental 35188 results show that our approach can derive 35189 high-quality solutions within a short time. 35190 </blockquote> 35191 </div> 35192 </div> 35193 </div> 35194 </div> 35195 <div class="slot-urls"></div> 35196 <a href="/wsc23papers/pos129.pdf" target="_blank">pdf</a 35197 ><br /> 35198 </div> 35199 <div class="slot-entry"> 35200 <a name="pos134" tabindex="-1"></a> 35201 <div class="slot-title-line"> 35202 <span class="slot-title" 35203 >Simulating Job Replication Versus Its Energy 35204 Usage</span 35205 > 35206 </div> 35207 <div class="slot-authors"> 35208 Vladimir Marbukh and Brian Cloteaux (NIST) 35209 </div> 35210 <div class="slot-abstract"> 35211 <div> 35212 <a 35213 class="clickable no-decoration" 35214 id="vhsjs_view_810_1707793552_9934044" 35215 onclick="$('#vhsjs_view_810_1707793552_9934044').hide(); 35216 $('#vhsjs_hide_810_1707793552_9934044').show(); 35217 $('#809_1707793552_9933965').slideDown(function() { 35218 if (typeof Masonry === 'function') { 35219 $('.use_masonry').masonry(); 35220 }; 35221 35222 });" 35223 ><i class="fa fa-caret-right"></i>
35224 <span class="hover_link">Abstract</span></a 35225 ><a 35226 class="clickable no-decoration" 35227 id="vhsjs_hide_810_1707793552_9934044" 35228 onclick="$('#809_1707793552_9933965').hide(function() { 35229 if (typeof Masonry === 'function') { 35230 $('.use_masonry').masonry(); 35231 }; 35232 }); 35233 $('#vhsjs_hide_810_1707793552_9934044').hide(); 35234 $('#vhsjs_view_810_1707793552_9934044').show();" 35235 style="display: none" 35236 ><i class="fa fa-caret-down"></i> 35237 <span class="hover_link">Abstract</span></a 35238 > 35239 <div 35240 data-display-control="810_1707793552_9934044" 35241 id="809_1707793552_9933965" 35242 style="display: none" 35243 > 35244 <div class="arrow-slidedown"> 35245 <blockquote> 35246 Due to the proliferation of computers in all 35247 aspects of our lives, the energy and ecological 35248 impacts of computing are becoming increasing 35249 important. Some of the transformative algorithms 35250 of recent years generate huge amounts of carbon 35251 dioxide, potentially damaging the environment. 35252 We have developed a set of simulations for 35253 understanding the trade-offs between distributed 35254 computing and its carbon impact. We briefly 35255 describe our current work and our future 35256 research aiming at finding practical algorithmic 35257 solutions. 35258 </blockquote> 35259 </div> 35260 </div> 35261 </div> 35262 </div> 35263 <div class="slot-urls"></div> 35264 <a href="/wsc23papers/pos134.pdf" target="_blank">pdf</a 35265 ><br /> 35266 </div> 35267 <div class="slot-entry"> 35268 <a name="pos139" tabindex="-1"></a> 35269 <div class="slot-title-line"> 35270 <span class="slot-title" 35271 >Bayesian Subset Selection for Near-Optimal 35272 Systems</span 35273 > 35274 </div> 35275 <div class="slot-authors"> 35276 Javier Gatica (Pontificia Universidad Católica de 35277 Chile) and Jinbo Zhao and David J. Eckman (Texas A&M 35278 University) 35279 </div> 35280 <div class="slot-abstract"> 35281 <div> 35282 <a 35283 class="clickable no-decoration" 35284 id="vhsjs_view_812_1707793552_9960136" 35285 onclick="$('#vhsjs_view_812_1707793552_9960136').hide(); 35286 $('#vhsjs_hide_812_1707793552_9960136').show(); 35287 $('#811_1707793552_9960055').slideDown(function() { 35288 if (typeof Masonry === 'function') { 35289 $('.use_masonry').masonry(); 35290 }; 35291 35292 });" 35293 ><i class="fa fa-caret-right"></i> 35294 <span class="hover_link">Abstract</span></a 35295 ><a 35296 class="clickable no-decoration" 35297 id="vhsjs_hide_812_1707793552_9960136" 35298 onclick="$('#811_1707793552_9960055').hide(function() { 35299 if (typeof Masonry === 'function') { 35300 $('.use_masonry').masonry(); 35301 }; 35302 }); 35303 $('#vhsjs_hide_812_1707793552_9960136').hide(); 35304 $('#vhsjs_view_812_1707793552_9960136').show();" 35305 style="display: none" 35306 ><i class="fa fa-caret-down"></i>
35307 <span class="hover_link">Abstract</span></a 35308 > 35309 <div 35310 data-display-control="812_1707793552_9960136" 35311 id="811_1707793552_9960055" 35312 style="display: none" 35313 > 35314 <div class="arrow-slidedown"> 35315 <blockquote> 35316 We study the ranking-and-selection problem of 35317 selecting a subset of simulated systems that 35318 with high probability contains a system with 35319 near-optimal performance. The posterior 35320 probability that at least one system in a given 35321 subset is near optimal - referred to as the 35322 posterior probability of good inclusion (pPGI) - 35323 can be expressed in terms of a sum of 35324 one-dimensional integrals and computed via 35325 numerical integration. Still, enumerating all 35326 possible subsets and computing their associated 35327 pPGI is impractical for large problem instances, 35328 thus we explore approximate solution methods. In 35329 particular, we investigate a greedy algorithm 35330 that builds a subset by iteratively adding the 35331 system that increases the pPGI the most. 35332 </blockquote> 35333 </div> 35334 </div> 35335 </div> 35336 </div> 35337 <div class="slot-urls"></div> 35338 <a href="/wsc23papers/pos139.pdf" target="_blank">pdf</a 35339 ><br /> 35340 </div> 35341 <div class="slot-entry"> 35342 <a name="pos140" tabindex="-1"></a> 35343 <div class="slot-title-line"> 35344 <span class="slot-title" 35345 >An Integrated Framework for Efficient Wireless 35346 Coverage Mapping Using Ray Tracing Acceleration</span 35347 > 35348 </div> 35349 <div class="slot-authors"> 35350 Hieu Le, Jian Tao, and Hernan Santos (Texas A&M) 35351 </div> 35352 <div class="slot-abstract"> 35353 <div> 35354 <a 35355 class="clickable no-decoration" 35356 id="vhsjs_view_814_1707793552_9985938" 35357 onclick="$('#vhsjs_view_814_1707793552_9985938').hide(); 35358 $('#vhsjs_hide_814_1707793552_9985938').show(); 35359 $('#813_1707793552_998586').slideDown(function() { 35360 if (typeof Masonry === 'function') { 35361 $('.use_masonry').masonry(); 35362 }; 35363 35364 });" 35365 ><i class="fa fa-caret-right"></i> 35366 <span class="hover_link">Abstract</span></a 35367 ><a 35368 class="clickable no-decoration" 35369 id="vhsjs_hide_814_1707793552_9985938" 35370 onclick="$('#813_1707793552_998586').hide(function() { 35371 if (typeof Masonry === 'function') { 35372 $('.use_masonry').masonry(); 35373 }; 35374 }); 35375 $('#vhsjs_hide_814_1707793552_9985938').hide(); 35376 $('#vhsjs_view_814_1707793552_9985938').show();" 35377 style="display: none" 35378 ><i class="fa fa-caret-down"></i> 35379 <span class="hover_link">Abstract</span></a 35380 > 35381 <div
35382 data-display-control="814_1707793552_9985938" 35383 id="813_1707793552_998586" 35384 style="display: none" 35385 > 35386 <div class="arrow-slidedown"> 35387 <blockquote> 35388 Evaluation of channel properties is one of the 35389 most important aspects in wireless 35390 communications. Ray tracing simulations have 35391 been widely used to estimate channel 35392 characteristics. In this poster, we put together 35393 many aspects of ray tracing techniques and 35394 signal estimation methods to build a coverage 35395 map. Acceleration structures for ray tracing are 35396 created to drastically reduce the computational 35397 time of the traversal of the ray-primitive 35398 intersections. Moreover, electromagnetics and 35399 wireless communications theories are studied to 35400 accurately estimate signal strength at an 35401 arbitrary point in the predefined area of the 35402 coverage map. 35403 </blockquote> 35404 </div> 35405 </div> 35406 </div> 35407 </div> 35408 <div class="slot-urls"></div> 35409 <a href="/wsc23papers/pos140.pdf" target="_blank">pdf</a 35410 ><br /> 35411 </div> 35412 </div> 35413 </div> 35414 <div class="centered"> 35415 <div class="top-link"><a href="#top">Return to Top</a></div> 35416 </div> 35417 <hr /> 35418 </div> 35419 <div class="area-section"> 35420 <div class="centered"> 35421 <a name="ptrack139" tabindex="-1"></a> 35422 <div class="section-title">PhD Colloquium</div> 35423 </div> 35424 <div class="centered track-chair"> 35425 <span class="track-chair-role" 35426 >Track Coordinator - Ph.D. Colloquium: </span 35427 ><span class="track-chair-names" 35428 >Anatoli Djanatliev (University of Erlangen-Nuremberg), 35429 Siyang Gao (City University of Hong Kong), Cristina 35430 Ruiz-Martín (Carleton University), Eunhye Song (Georgia 35431 Institute of Technology)</span 35432 > 35433 </div> 35434 <div class="section-entry"> 35435 <div class="session-entry"> 35436 <span class="session-event-type">PhD Colloquium</span 35437 ><span class="type-track-spacer"> · </span 35438 ><span class="program-track">PhD Colloquium</span><br /> 35439 <div class="session-title"> 35440 PhD Colloquium Keynote: Methods and Applications or 35441 Applications and Methods? 35442 </div> 35443 <div class="session-chair"> 35444 Chair: Siyang Gao (City University of Hong Kong)<br /> 35445 </div> 35446 <div class="slot-entry"> 35447 <a name="prog107" tabindex="-1"></a> 35448 <div class="slot-title-line"> 35449 <span class="slot-title" 35450 >Methods and Applications or Applications and 35451 Methods?</span 35452 > 35453 </div> 35454 <div class="slot-authors">Stephen Chick (INSEAD)</div> 35455 <div class="slot-abstract"> 35456 <div> 35457 <a 35458 class="clickable no-decoration" 35459 id="vhsjs_view_816_1707793553_0047011" 35460 onclick="$('#vhsjs_view_816_1707793553_0047011').hide(); 35461 $('#vhsjs_hide_816_1707793553_0047011').show(); 35462 $('#815_1707793553_004693').slideDown(function() { 35463 if (typeof Masonry === 'function') { 35464 $('.use_masonry').masonry(); 35465 }; 35466 35467 });" 35468 ><i class="fa fa-caret-right"></i>
35469 <span class="hover_link">Abstract</span></a 35470 ><a 35471 class="clickable no-decoration" 35472 id="vhsjs_hide_816_1707793553_0047011" 35473 onclick="$('#815_1707793553_004693').hide(function() { 35474 if (typeof Masonry === 'function') { 35475 $('.use_masonry').masonry(); 35476 }; 35477 }); 35478 $('#vhsjs_hide_816_1707793553_0047011').hide(); 35479 $('#vhsjs_view_816_1707793553_0047011').show();" 35480 style="display: none" 35481 ><i class="fa fa-caret-down"></i> 35482 <span class="hover_link">Abstract</span></a 35483 > 35484 <div 35485 data-display-control="816_1707793553_0047011" 35486 id="815_1707793553_004693" 35487 style="display: none" 35488 > 35489 <div class="arrow-slidedown"> 35490 <blockquote> 35491 Stochastic simulation is a powerful framework 35492 for supporting decision makers in a broad range 35493 of applications. Its methods draw upon applied 35494 probability, system dynamics, statistics, 35495 computing, and other fields. Simulation methods 35496 are interesting in and of themselves, including 35497 uncertainty modelling, stochastic optimization, 35498 the valuation of uncertainty, efficiency 35499 improvement, and the modelling of complex system 35500 behavior that might be hard to analyze through 35501 closed-form analysis. Applications may sometimes 35502 have standard approaches to support the analysis 35503 to inform a decision maker, but decision makers
35504 may also have criteria that are not reflected 35505 fully in a simulation model. And sometimes new 35506 applications give rise to very interesting 35507 structures that call for further analysis. In 35508 this talk, we discuss the feedback loop between 35509 methods development that allow new applications 35510 to be addressed, and new applications that give 35511 rise to new methods. 35512 </blockquote> 35513 </div> 35514 </div> 35515 </div> 35516 </div> 35517 <div class="slot-urls"></div> 35518 <a href="/wsc23papers/prog107.pdf" target="_blank">pdf</a 35519 ><br /> 35520 </div> 35521 </div> 35522 <div class="session-entry"> 35523 <span class="session-event-type">PhD Colloquium</span 35524 ><span class="type-track-spacer"> · </span 35525 ><span class="program-track">PhD Colloquium</span><br /> 35526 <div class="session-title">PhD Colloquium Session A1</div> 35527 <div class="session-chair"> 35528 Chair: Siyang Gao (City University of Hong Kong)<br /> 35529 </div> 35530 <div class="slot-entry"> 35531 <a name="doc101" tabindex="-1"></a> 35532 <div class="slot-title-line"> 35533 <span class="slot-title" 35534 >Reusing Historical Observations in Natural Policy 35535 Gradient</span 35536 > 35537 </div> 35538 <div class="slot-authors"> 35539 Yifan Lin (Georgia Institute of Technology) 35540 </div> 35541 <div class="slot-abstract"> 35542 <div> 35543 <a 35544 class="clickable no-decoration" 35545 id="vhsjs_view_818_1707793553_017338" 35546 onclick="$('#vhsjs_view_818_1707793553_017338').hide(); 35547 $('#vhsjs_hide_818_1707793553_017338').show(); 35548 $('#817_1707793553_0173295').slideDown(function() { 35549 if (typeof Masonry === 'function') { 35550 $('.use_masonry').masonry(); 35551 }; 35552 35553 });" 35554 ><i class="fa fa-caret-right"></i> 35555 <span class="hover_link">Abstract</span></a 35556 ><a 35557 class="clickable no-decoration" 35558 id="vhsjs_hide_818_1707793553_017338" 35559 onclick="$('#817_1707793553_0173295').hide(function() { 35560 if (typeof Masonry === 'function') { 35561 $('.use_masonry').masonry(); 35562 }; 35563 }); 35564 $('#vhsjs_hide_818_1707793553_017338').hide(); 35565 $('#vhsjs_view_818_1707793553_017338').show();" 35566 style="display: none" 35567 ><i class="fa fa-caret-down"></i> 35568 <span class="hover_link">Abstract</span></a 35569 > 35570 <div 35571 data-display-control="818_1707793553_017338" 35572 id="817_1707793553_0173295" 35573 style="display: none" 35574 > 35575 <div class="arrow-slidedown"> 35576 <blockquote> 35577 Reinforcement learning provides a framework for 35578 learning-based control, whose success largely 35579 depends on the amount of data it can utilize. 35580 The efficient utilization of historical samples 35581 obtained from previous iterations is essential 35582 for expediting policy optimization. Empirical 35583 evidence has shown that offline variants of 35584 policy gradient methods based on importance 35585 sampling work well. However, existing literature 35586 often neglect the interdependence between 35587 observations from different iterations, and the 35588 good empirical performance lacks a rigorous 35589 theoretical justification. In this paper, we 35590 study an offline variant of the natural policy 35591 gradient method with reusing historical 35592 observations. We show that the biases of the 35593 proposed estimators of Fisher information matrix 35594 and gradient are asymptotically negligible, and 35595 reusing historical observations reduces the 35596 conditional variance of the gradient estimator. 35597 The proposed algorithm and convergence analysis 35598 could be further applied to popular policy 35599 optimization algorithms such as trust region 35600 policy optimization. 35601 </blockquote> 35602 </div> 35603 </div> 35604 </div> 35605 </div> 35606 <div class="slot-urls"></div> 35607 <a href="/wsc23papers/doc101.pdf" target="_blank">pdf</a 35608 ><br /> 35609 </div> 35610 <div class="slot-entry"> 35611 <a name="doc102" tabindex="-1"></a> 35612 <div class="slot-title-line">
35613 <span class="slot-title" 35614 >Dispatching in Real Frontend Fabs With Industrial 35615 Grade Discrete-Event Simulations by Deep Reinforcement 35616 Learning With Evolution Strategies</span 35617 > 35618 </div> 35619 <div class="slot-authors"> 35620 Patrick Stöckermann (Infineon Technologies AG) 35621 </div> 35622 <div class="slot-abstract"> 35623 <div> 35624 <a 35625 class="clickable no-decoration" 35626 id="vhsjs_view_820_1707793553_0412428" 35627 onclick="$('#vhsjs_view_820_1707793553_0412428').hide(); 35628 $('#vhsjs_hide_820_1707793553_0412428').show(); 35629 $('#819_1707793553_0412345').slideDown(function() { 35630 if (typeof Masonry === 'function') { 35631 $('.use_masonry').masonry(); 35632 }; 35633 35634 });" 35635 ><i class="fa fa-caret-right"></i> 35636 <span class="hover_link">Abstract</span></a 35637 ><a 35638 class="clickable no-decoration" 35639 id="vhsjs_hide_820_1707793553_0412428" 35640 onclick="$('#819_1707793553_0412345').hide(function() { 35641 if (typeof Masonry === 'function') { 35642 $('.use_masonry').masonry(); 35643 }; 35644 }); 35645 $('#vhsjs_hide_820_1707793553_0412428').hide(); 35646 $('#vhsjs_view_820_1707793553_0412428').show();" 35647 style="display: none" 35648 ><i class="fa fa-caret-down"></i> 35649 <span class="hover_link">Abstract</span></a 35650 > 35651 <div 35652 data-display-control="820_1707793553_0412428" 35653 id="819_1707793553_0412345" 35654 style="display: none" 35655 > 35656 <div class="arrow-slidedown"> 35657 <blockquote> 35658 Scheduling is a fundamental task in each 35659 production facility with implications on the 35660 overall efficiency of the facility. While 35661 classic job-shop scheduling problems become 35662 intractable when the number of machines and jobs 35663 increases, the problem gets even more complex in 35664 the context of semiconductor manufacturing, 35665 where flexible production control and stochastic 35666 event handling are required. In this paper, we 35667 propose a Deep Reinforcement Learning approach 35668 for lot dispatching to minimize the Flow Factor 35669 (FF) of a digital twin of a real-world, 35670 stochastic, large-scale semiconductor 35671 manufacturing facility. We present the first 35672 application of Reinforcement Learning (RL) to an 35673 industrial grade semiconductor manufacturing 35674 scenario of that size. Our approach leverages a 35675 self-attention mechanism to learn an effective 35676 dispatching policy for the manufacturing 35677 facility and is able to reduce the global FF of 35678 the fab. 35679 </blockquote> 35680 </div> 35681 </div> 35682 </div> 35683 </div> 35684 <div class="slot-urls"></div> 35685 <a href="/wsc23papers/doc102.pdf" target="_blank">pdf</a 35686 ><br /> 35687 </div> 35688 <div class="slot-entry"> 35689 <a name="doc103" tabindex="-1"></a> 35690 <div class="slot-title-line">
35691 <span class="slot-title" 35692 >Cutting through the Noise: Machine Learning Proxies 35693 for High Dimensional Nested Simulation</span 35694 > 35695 </div> 35696 <div class="slot-authors"> 35697 Xintong Li (University of Waterloo) 35698 </div> 35699 <div class="slot-abstract"> 35700 <div> 35701 <a 35702 class="clickable no-decoration" 35703 id="vhsjs_view_822_1707793553_043828" 35704 onclick="$('#vhsjs_view_822_1707793553_043828').hide(); 35705 $('#vhsjs_hide_822_1707793553_043828').show(); 35706 $('#821_1707793553_0438201').slideDown(function() { 35707 if (typeof Masonry === 'function') { 35708 $('.use_masonry').masonry(); 35709 }; 35710 35711 });" 35712 ><i class="fa fa-caret-right"></i> 35713 <span class="hover_link">Abstract</span></a 35714 ><a 35715 class="clickable no-decoration" 35716 id="vhsjs_hide_822_1707793553_043828" 35717 onclick="$('#821_1707793553_0438201').hide(function() { 35718 if (typeof Masonry === 'function') { 35719 $('.use_masonry').masonry(); 35720 }; 35721 }); 35722 $('#vhsjs_hide_822_1707793553_043828').hide(); 35723 $('#vhsjs_view_822_1707793553_043828').show();" 35724 style="display: none" 35725 ><i class="fa fa-caret-down"></i> 35726 <span class="hover_link">Abstract</span></a 35727 > 35728 <div 35729 data-display-control="822_1707793553_043828" 35730 id="821_1707793553_0438201" 35731 style="display: none" 35732 > 35733 <div class="arrow-slidedown"> 35734 <blockquote> 35735 Deep learning models have gained great success 35736 in many applications, but their adoption in 35737 financial and actuarial applications have been 35738 received by regulators with trepidation. The 35739 lack of transparency and interpretability of 35740 these models raises skepticism about their 35741 resilience and reliability, which are important 35742 factors for financial stability and insurance 35743 benefit fulfillment. In this study, we use 35744 stochastic simulation as a data generator to 35745 examine deep learning models under controlled 35746 settings. Our study shows interesting findings 35747 in fundamental questions like “What do 35748 deep learning models learn from noisy 35749 data?” and “How well do they learn 35750 from noisy data?”. Based on our findings, 35751 we propose an efficient nested simulation 35752 procedure that uses deep learning models as 35753 proxies to estimate tail risk measures of 35754 hedging errors for variable annuities. The 35755 proposed procedure uses deep learning to 35756 concentrate simulation budget on tail scenarios 35757 while maintaining transparency in estimation. 35758 </blockquote> 35759 </div> 35760 </div> 35761 </div> 35762 </div> 35763 <div class="slot-urls"></div> 35764 <a href="/wsc23papers/doc103.pdf" target="_blank">pdf</a 35765 ><br /> 35766 </div> 35767 <div class="slot-entry"> 35768 <a name="doc105" tabindex="-1"></a> 35769 <div class="slot-title-line">
35770 <span class="slot-title" 35771 >Solving Deadlock Situations in Intralogistics with 35772 Reinforcement Learning</span 35773 > 35774 </div> 35775 <div class="slot-authors"> 35776 Marcel Müller (Otto von Guericke University 35777 Magdeburg) 35778 </div> 35779 <div class="slot-abstract"> 35780 <div> 35781 <a 35782 class="clickable no-decoration" 35783 id="vhsjs_view_824_1707793553_0462027" 35784 onclick="$('#vhsjs_view_824_1707793553_0462027').hide(); 35785 $('#vhsjs_hide_824_1707793553_0462027').show(); 35786 $('#823_1707793553_0461938').slideDown(function() { 35787 if (typeof Masonry === 'function') { 35788 $('.use_masonry').masonry(); 35789 }; 35790 35791 });" 35792 ><i class="fa fa-caret-right"></i> 35793 <span class="hover_link">Abstract</span></a 35794 ><a 35795 class="clickable no-decoration" 35796 id="vhsjs_hide_824_1707793553_0462027" 35797 onclick="$('#823_1707793553_0461938').hide(function() { 35798 if (typeof Masonry === 'function') { 35799 $('.use_masonry').masonry(); 35800 }; 35801 }); 35802 $('#vhsjs_hide_824_1707793553_0462027').hide(); 35803 $('#vhsjs_view_824_1707793553_0462027').show();" 35804 style="display: none" 35805 ><i class="fa fa-caret-down"></i> 35806 <span class="hover_link">Abstract</span></a 35807 > 35808 <div 35809 data-display-control="824_1707793553_0462027" 35810 id="823_1707793553_0461938" 35811 style="display: none" 35812 > 35813 <div class="arrow-slidedown"> 35814 <blockquote> 35815 Intralogistics faces challenges from global 35816 disruptions such as the COVID-19 pandemic, 35817 geopolitical tensions, and wars, emphasizing the 35818 need for increased flexibility of logistic 35819 systems. Compounded by staff shortages in 35820 industrial countries, automation continues to 35821 rise, evidenced by the growing number of 35822 industrial robots. This rise in automation 35823 demands enhanced capabilities for intralogistic 35824 systems, including handling deadlocks. This 35825 research delves into the potential of 35826 reinforcement learning (RL) in addressing 35827 deadlocks, aiming to increase the efficiency, 35828 flexibility, and resilience of intralogistic 35829 systems. 35830 </blockquote> 35831 </div> 35832 </div> 35833 </div> 35834 </div> 35835 <div class="slot-urls"></div> 35836 <a href="/wsc23papers/doc105.pdf" target="_blank">pdf</a 35837 ><br /> 35838 </div> 35839 <div class="slot-entry"> 35840 <a name="doc108" tabindex="-1"></a> 35841 <div class="slot-title-line"> 35842 <span class="slot-title" 35843 >Feature Selection in Generalized Linear models via 35844 the Lasso: To Scale or Not to Scale?</span 35845 > 35846 </div> 35847 <div class="slot-authors"> 35848 Anant Mathur (University of New South Wales) 35849 </div> 35850 <div class="slot-abstract"> 35851 <div> 35852 <a 35853 class="clickable no-decoration" 35854 id="vhsjs_view_826_1707793553_0487568" 35855 onclick="$('#vhsjs_view_826_1707793553_0487568').hide(); 35856 $('#vhsjs_hide_826_1707793553_0487568').show(); 35857 $('#825_1707793553_0487487').slideDown(function() { 35858 if (typeof Masonry === 'function') { 35859 $('.use_masonry').masonry(); 35860 }; 35861 35862 });" 35863 ><i class="fa fa-caret-right"></i>
35864 <span class="hover_link">Abstract</span></a 35865 ><a 35866 class="clickable no-decoration" 35867 id="vhsjs_hide_826_1707793553_0487568" 35868 onclick="$('#825_1707793553_0487487').hide(function() { 35869 if (typeof Masonry === 'function') { 35870 $('.use_masonry').masonry(); 35871 }; 35872 }); 35873 $('#vhsjs_hide_826_1707793553_0487568').hide(); 35874 $('#vhsjs_view_826_1707793553_0487568').show();" 35875 style="display: none" 35876 ><i class="fa fa-caret-down"></i> 35877 <span class="hover_link">Abstract</span></a 35878 > 35879 <div 35880 data-display-control="826_1707793553_0487568" 35881 id="825_1707793553_0487487" 35882 style="display: none" 35883 > 35884 <div class="arrow-slidedown"> 35885 <blockquote> 35886 The Lasso regression is a popular regularization 35887 method for feature selection in statistics. 35888 Prior to computing the Lasso estimator in both 35889 linear and generalized linear models, it is 35890 common to conduct a preliminary rescaling of the 35891 feature matrix to ensure that all the features 35892 are standardized. Without this standardization, 35893 it is argued, the Lasso estimate will, 35894 unfortunately, depend on the units used to 35895 measure the features. We propose a new type of 35896 iterative rescaling of the features in the 35897 context of generalized linear models. Whilst 35898 existing Lasso algorithms perform a single 35899 scaling as a preprocessing step, the proposed 35900 rescaling is applied iteratively throughout the 35901 Lasso computation until convergence. We provide 35902 numerical examples, with both real and simulated 35903 data, illustrating that the proposed iterative 35904 rescaling can significantly improve the 35905 statistical performance of the Lasso estimator 35906 without incurring any significant additional 35907 computational cost. 35908 </blockquote> 35909 </div> 35910 </div> 35911 </div> 35912 </div> 35913 <div class="slot-urls"></div> 35914 <a href="/wsc23papers/doc108.pdf" target="_blank">pdf</a 35915 ><br /> 35916 </div> 35917 <div class="slot-entry"> 35918 <a name="doc112" tabindex="-1"></a> 35919 <div class="slot-title-line"> 35920 <span class="slot-title" 35921 >Hyperheuristic Optimization as Decision Suport for 35922 the Operative Service Delivery Planning in the Context 35923 of Product-Service Systems</span 35924 > 35925 </div> 35926 <div class="slot-authors"> 35927 Enes Alp (Ruhr-Universität Bochum) 35928 </div> 35929 <div class="slot-abstract"> 35930 <div> 35931 <a 35932 class="clickable no-decoration" 35933 id="vhsjs_view_828_1707793553_0510802" 35934 onclick="$('#vhsjs_view_828_1707793553_0510802').hide(); 35935 $('#vhsjs_hide_828_1707793553_0510802').show(); 35936 $('#827_1707793553_0510726').slideDown(function() { 35937 if (typeof Masonry === 'function') { 35938 $('.use_masonry').masonry(); 35939 }; 35940 35941 });" 35942 ><i class="fa fa-caret-right"></i>
35943 <span class="hover_link">Abstract</span></a 35944 ><a 35945 class="clickable no-decoration" 35946 id="vhsjs_hide_828_1707793553_0510802" 35947 onclick="$('#827_1707793553_0510726').hide(function() { 35948 if (typeof Masonry === 'function') { 35949 $('.use_masonry').masonry(); 35950 }; 35951 }); 35952 $('#vhsjs_hide_828_1707793553_0510802').hide(); 35953 $('#vhsjs_view_828_1707793553_0510802').show();" 35954 style="display: none" 35955 ><i class="fa fa-caret-down"></i> 35956 <span class="hover_link">Abstract</span></a 35957 > 35958 <div 35959 data-display-control="828_1707793553_0510802" 35960 id="827_1707793553_0510726" 35961 style="display: none" 35962 > 35963 <div class="arrow-slidedown"> 35964 <blockquote> 35965 In the pursuit of differentiation and revenue 35966 increment, numerous manufacturing enterprises 35967 are innovating their business models through the 35968 introduction of Product-Service Systems (PSS). 35969 In these business models, the efficacy of 35970 service delivery assumes paramount significance, 35971 leading to challenges in the planning. The 35972 objective of this PhD project is the 35973 conceptualization and development of a decision 35974 support system for operative service delivery 35975 planning within the context of PSS. 35976 </blockquote> 35977 </div> 35978 </div> 35979 </div> 35980 </div> 35981 <div class="slot-urls"></div> 35982 <a href="/wsc23papers/doc112.pdf" target="_blank">pdf</a 35983 ><br /> 35984 </div> 35985 <div class="slot-entry"> 35986 <a name="doc114" tabindex="-1"></a> 35987 <div class="slot-title-line"> 35988 <span class="slot-title" 35989 >System Simulation and Machine Learning-Based 35990 Maintenance Optimization for an Inland Waterway 35991 Transportation System</span 35992 > 35993 </div> 35994 <div class="slot-authors"> 35995 Maryam Aghamohammadghasem (University of Arkansas) 35996 </div> 35997 <div class="slot-abstract"> 35998 <div> 35999 <a 36000 class="clickable no-decoration" 36001 id="vhsjs_view_830_1707793553_0536418" 36002 onclick="$('#vhsjs_view_830_1707793553_0536418').hide(); 36003 $('#vhsjs_hide_830_1707793553_0536418').show(); 36004 $('#829_1707793553_0536335').slideDown(function() { 36005 if (typeof Masonry === 'function') { 36006 $('.use_masonry').masonry(); 36007 }; 36008 36009 });" 36010 ><i class="fa fa-caret-right"></i> 36011 <span class="hover_link">Abstract</span></a 36012 ><a 36013 class="clickable no-decoration" 36014 id="vhsjs_hide_830_1707793553_0536418" 36015 onclick="$('#829_1707793553_0536335').hide(function() { 36016 if (typeof Masonry === 'function') { 36017 $('.use_masonry').masonry(); 36018 }; 36019 }); 36020 $('#vhsjs_hide_830_1707793553_0536418').hide(); 36021 $('#vhsjs_view_830_1707793553_0536418').show();" 36022 style="display: none" 36023 ><i class="fa fa-caret-down"></i>
36024 <span class="hover_link">Abstract</span></a 36025 > 36026 <div 36027 data-display-control="830_1707793553_0536418" 36028 id="829_1707793553_0536335" 36029 style="display: none" 36030 > 36031 <div class="arrow-slidedown"> 36032 <blockquote> 36033 To keep an inland waterway transportation system 36034 (IWTS) up and running, the interconnected 36035 infrastructure, including lock and dam systems, 36036 must remain in good operating condition. 36037 However, unexpected disruptions often occur, 36038 causing significant transportation delays and 36039 economic losses. To evaluate the impacts of such 36040 disruptions, a Python-enhanced NetLogo 36041 simulation tool is developed, in which extreme 36042 natural events are considered and characterized 36043 by a spatiotemporal model. With this tool, 36044 optimal maintenance strategies that maximize the 36045 total cargo throughput of the IWTS are 36046 determined via deep reinforcement learning. A 36047 case study of the lower Mississippi River system 36048 and the McClellan-Kerr Arkansas River Navigation 36049 System is conducted to illustrate the capability 36050 of the developed simulation and machine 36051 learning-based method for IWTS maintenance 36052 optimization. 36053 </blockquote> 36054 </div> 36055 </div> 36056 </div> 36057 </div> 36058 <div class="slot-urls"></div> 36059 <a href="/wsc23papers/doc114.pdf" target="_blank">pdf</a 36060 ><br /> 36061 </div> 36062 <div class="slot-entry"> 36063 <a name="doc117" tabindex="-1"></a> 36064 <div class="slot-title-line"> 36065 <span class="slot-title" 36066 >Strengthening Emergency Department Resilience: 36067 Simulation-Based Surge Management</span 36068 > 36069 </div> 36070 <div class="slot-authors"> 36071 Eman Ouda (Khalifa University) 36072 </div> 36073 <div class="slot-abstract"> 36074 <div> 36075 <a 36076 class="clickable no-decoration" 36077 id="vhsjs_view_832_1707793553_0560594" 36078 onclick="$('#vhsjs_view_832_1707793553_0560594').hide(); 36079 $('#vhsjs_hide_832_1707793553_0560594').show(); 36080 $('#831_1707793553_0560515').slideDown(function() { 36081 if (typeof Masonry === 'function') { 36082 $('.use_masonry').masonry(); 36083 }; 36084 36085 });" 36086 ><i class="fa fa-caret-right"></i> 36087 <span class="hover_link">Abstract</span></a 36088 ><a 36089 class="clickable no-decoration" 36090 id="vhsjs_hide_832_1707793553_0560594" 36091 onclick="$('#831_1707793553_0560515').hide(function() { 36092 if (typeof Masonry === 'function') { 36093 $('.use_masonry').masonry(); 36094 }; 36095 }); 36096 $('#vhsjs_hide_832_1707793553_0560594').hide();
36097 $('#vhsjs_view_832_1707793553_0560594').show();" 36098 style="display: none" 36099 ><i class="fa fa-caret-down"></i> 36100 <span class="hover_link">Abstract</span></a 36101 > 36102 <div 36103 data-display-control="832_1707793553_0560594" 36104 id="831_1707793553_0560515" 36105 style="display: none" 36106 > 36107 <div class="arrow-slidedown"> 36108 <blockquote> 36109 This study aims to improve the resilience of the 36110 Emergency Department (ED) to handle demand 36111 surges through a combination of Discrete Event 36112 Simulation (DES) and resilience assessment 36113 techniques. By evaluating resistance and 36114 recoverability components, the analysis examines 36115 the resilience of the ED, patient flow dynamics, 36116 and resource requirements. A dedicated 36117 simulation model is developed to uncover how the 36118 ED performs during normal operations and demand 36119 surges, exploring the effects of alterations and 36120 additional resources on resilience using the 36121 resilience triangle framework for optimized 36122 resource allocation. This research improves our 36123 understanding of ED resilience, paving the way 36124 for further investigations into performance 36125 improvement during demand spikes, and the 36126 results suggest new patient flow strategies to 36127 enhance resilience. 36128 </blockquote> 36129 </div> 36130 </div> 36131 </div> 36132 </div> 36133 <div class="slot-urls"></div> 36134 <a href="/wsc23papers/doc117.pdf" target="_blank">pdf</a 36135 ><br /> 36136 </div> 36137 <div class="slot-entry"> 36138 <a name="doc120" tabindex="-1"></a> 36139 <div class="slot-title-line"> 36140 <span class="slot-title" 36141 >Expediting Stochastic Derivative-free 36142 Optimization</span 36143 > 36144 </div> 36145 <div class="slot-authors"> 36146 Yunsoo Ha (North Carolina State University) 36147 </div> 36148 <div class="slot-abstract"> 36149 <div> 36150 <a 36151 class="clickable no-decoration" 36152 id="vhsjs_view_834_1707793553_0585487" 36153 onclick="$('#vhsjs_view_834_1707793553_0585487').hide(); 36154 $('#vhsjs_hide_834_1707793553_0585487').show(); 36155 $('#833_1707793553_058541').slideDown(function() { 36156 if (typeof Masonry === 'function') { 36157 $('.use_masonry').masonry(); 36158 }; 36159 36160 });" 36161 ><i class="fa fa-caret-right"></i> 36162 <span class="hover_link">Abstract</span></a 36163 ><a 36164 class="clickable no-decoration" 36165 id="vhsjs_hide_834_1707793553_0585487" 36166 onclick="$('#833_1707793553_058541').hide(function() { 36167 if (typeof Masonry === 'function') { 36168 $('.use_masonry').masonry(); 36169 }; 36170 }); 36171 $('#vhsjs_hide_834_1707793553_0585487').hide(); 36172 $('#vhsjs_view_834_1707793553_0585487').show();" 36173 style="display: none" 36174 ><i class="fa fa-caret-down"></i>
36175 <span class="hover_link">Abstract</span></a 36176 > 36177 <div 36178 data-display-control="834_1707793553_0585487" 36179 id="833_1707793553_058541" 36180 style="display: none" 36181 > 36182 <div class="arrow-slidedown"> 36183 <blockquote> 36184 Adaptive sampling-based trust-region 36185 optimization has emerged as an efficient solver 36186 for nonlinear and nonconvex problems in noisy 36187 derivative-free environments. This class of 36188 algorithms proceeds by iteratively constructing 36189 local models on objective function estimates 36190 that use a carefully chosen number of calls to 36191 the stochastic oracle. To expedite this class of 36192 algorithms, we introduce four refinements: (a) 36193 quadratic local models with diagonal Hessian, 36194 (b) a direct search, (c) a reusing strategy, and 36195 (d) common random numbers. We have substantiated 36196 that the introduced refinements enable the 36197 algorithm to achieve accelerated convergence, 36198 both in numerical simulations and in theoretical 36199 analyses. 36200 </blockquote> 36201 </div> 36202 </div> 36203 </div> 36204 </div> 36205 <div class="slot-urls"></div> 36206 <a href="/wsc23papers/doc120.pdf" target="_blank">pdf</a 36207 ><br /> 36208 </div> 36209 <div class="slot-entry"> 36210 <a name="doc121" tabindex="-1"></a> 36211 <div class="slot-title-line"> 36212 <span class="slot-title" 36213 >Conditional Importance Sampling for Convex Rare-Event 36214 Sets</span 36215 > 36216 </div> 36217 <div class="slot-authors"> 36218 Lewen Zheng (The Chinese University of Hong Kong) 36219 </div> 36220 <div class="slot-abstract"> 36221 <div> 36222 <a 36223 class="clickable no-decoration" 36224 id="vhsjs_view_836_1707793553_0610836" 36225 onclick="$('#vhsjs_view_836_1707793553_0610836').hide(); 36226 $('#vhsjs_hide_836_1707793553_0610836').show(); 36227 $('#835_1707793553_0610754').slideDown(function() { 36228 if (typeof Masonry === 'function') { 36229 $('.use_masonry').masonry(); 36230 }; 36231 36232 });" 36233 ><i class="fa fa-caret-right"></i> 36234 <span class="hover_link">Abstract</span></a 36235 ><a 36236 class="clickable no-decoration" 36237 id="vhsjs_hide_836_1707793553_0610836" 36238 onclick="$('#835_1707793553_0610754').hide(function() { 36239 if (typeof Masonry === 'function') { 36240 $('.use_masonry').masonry(); 36241 }; 36242 }); 36243 $('#vhsjs_hide_836_1707793553_0610836').hide(); 36244 $('#vhsjs_view_836_1707793553_0610836').show();" 36245 style="display: none" 36246 ><i class="fa fa-caret-down"></i> 36247 <span class="hover_link">Abstract</span></a 36248 > 36249 <div
36250 data-display-control="836_1707793553_0610836" 36251 id="835_1707793553_0610754" 36252 style="display: none" 36253 > 36254 <div class="arrow-slidedown"> 36255 <blockquote> 36256 This paper studies the efficient estimation of 36257 expectations defined on convex rare-event sets 36258 using importance sampling. Classical importance 36259 sampling methods often neglect the geometry of 36260 the target set, resulting in a significant 36261 number of samples falling outside the target 36262 set. This can lead to an increase in the 36263 relative error of the estimator as the target 36264 event becomes rarer. To address this issue, we 36265 develop a conditional importance sampling scheme 36266 that achieves bounded relative error by changing 36267 the sampling distribution to ensure that a 36268 majority of samples lie inside the target set. 36269 The proposed method is easy to implement and 36270 significantly outperforms the existing 36271 approaches in various numerical experiments. 36272 </blockquote> 36273 </div> 36274 </div> 36275 </div> 36276 </div> 36277 <div class="slot-urls"></div> 36278 <a href="/wsc23papers/doc121.pdf" target="_blank">pdf</a 36279 ><br /> 36280 </div> 36281 <div class="slot-entry"> 36282 <a name="doc123" tabindex="-1"></a> 36283 <div class="slot-title-line"> 36284 <span class="slot-title" 36285 >Efficient Input Uncertainty Quantification for 36286 Regenerative Simulation</span 36287 > 36288 </div> 36289 <div class="slot-authors"> 36290 Linyun He (Georgia Institute of Technology) 36291 </div> 36292 <div class="slot-abstract"> 36293 <div> 36294 <a 36295 class="clickable no-decoration" 36296 id="vhsjs_view_838_1707793553_0636256" 36297 onclick="$('#vhsjs_view_838_1707793553_0636256').hide(); 36298 $('#vhsjs_hide_838_1707793553_0636256').show(); 36299 $('#837_1707793553_0636175').slideDown(function() { 36300 if (typeof Masonry === 'function') { 36301 $('.use_masonry').masonry(); 36302 }; 36303 36304 });" 36305 ><i class="fa fa-caret-right"></i> 36306 <span class="hover_link">Abstract</span></a 36307 ><a 36308 class="clickable no-decoration" 36309 id="vhsjs_hide_838_1707793553_0636256" 36310 onclick="$('#837_1707793553_0636175').hide(function() { 36311 if (typeof Masonry === 'function') { 36312 $('.use_masonry').masonry(); 36313 }; 36314 }); 36315 $('#vhsjs_hide_838_1707793553_0636256').hide(); 36316 $('#vhsjs_view_838_1707793553_0636256').show();" 36317 style="display: none" 36318 ><i class="fa fa-caret-down"></i> 36319 <span class="hover_link">Abstract</span></a 36320 > 36321 <div 36322 data-display-control="838_1707793553_0636256" 36323 id="837_1707793553_0636175" 36324 style="display: none" 36325 > 36326 <div class="arrow-slidedown"> 36327 <blockquote> 36328 The initial bias in steady-state simulation can 36329 be characterized as the bias of a ratio 36330 estimator if the simulation model has a 36331 regenerative structure. This work tackles input 36332 uncertainty quantification for a regenerative 36333 simulation model when its input distributions 36334 are estimated from finite data. Our aim is to 36335 construct a bootstrap-based confidence interval 36336 (CI) for the true simulation output mean 36337 performance that provides a correct coverage 36338 with significantly less computational cost than 36339 the traditional methods. Exploiting the 36340 regenerative structure, we propose a k-nearest 36341 neighbor (kNN) ratio estimator for the 36342 steady-state performance measure at each set of 36343 bootstrapped input models and construct a 36344 bootstrap CI from the computed estimators. 36345 Asymptotically optimal choices for k and 36346 bootstrap sample size are discussed. We further 36347 improve the CI by combining the kNN and 36348 likelihood ratio methods. We empirically compare 36349 the efficiency of the proposed estimators with 36350 the standard estimator using queueing examples. 36351 </blockquote> 36352 </div> 36353 </div> 36354 </div> 36355 </div> 36356 <div class="slot-urls"></div> 36357 <a href="/wsc23papers/doc123.pdf" target="_blank">pdf</a 36358 ><br /> 36359 </div> 36360 </div> 36361 <div class="session-entry">
36362 <span class="session-event-type">PhD Colloquium</span 36363 ><span class="type-track-spacer"> · </span 36364 ><span class="program-track">PhD Colloquium</span><br /> 36365 <div class="session-title">PhD Colloquium Session B1</div> 36366 <div class="session-chair"> 36367 Chair: Enlu Zhou (Georgia Institute of Technology)<br /> 36368 </div> 36369 <div class="slot-entry"> 36370 <a name="doc130" tabindex="-1"></a> 36371 <div class="slot-title-line"> 36372 <span class="slot-title" 36373 >Shapley-Shubik Explanations of Feature 36374 Importance</span 36375 > 36376 </div> 36377 <div class="slot-authors"> 36378 Gayane Grigoryan (Old Dominion University) 36379 </div> 36380 <div class="slot-abstract"> 36381 <div> 36382 <a 36383 class="clickable no-decoration" 36384 id="vhsjs_view_840_1707793553_0688407" 36385 onclick="$('#vhsjs_view_840_1707793553_0688407').hide(); 36386 $('#vhsjs_hide_840_1707793553_0688407').show(); 36387 $('#839_1707793553_0688329').slideDown(function() { 36388 if (typeof Masonry === 'function') { 36389 $('.use_masonry').masonry(); 36390 }; 36391 36392 });" 36393 ><i class="fa fa-caret-right"></i> 36394 <span class="hover_link">Abstract</span></a 36395 ><a 36396 class="clickable no-decoration" 36397 id="vhsjs_hide_840_1707793553_0688407" 36398 onclick="$('#839_1707793553_0688329').hide(function() { 36399 if (typeof Masonry === 'function') { 36400 $('.use_masonry').masonry(); 36401 }; 36402 }); 36403 $('#vhsjs_hide_840_1707793553_0688407').hide(); 36404 $('#vhsjs_view_840_1707793553_0688407').show();" 36405 style="display: none" 36406 ><i class="fa fa-caret-down"></i> 36407 <span class="hover_link">Abstract</span></a 36408 > 36409 <div 36410 data-display-control="840_1707793553_0688407" 36411 id="839_1707793553_0688329" 36412 style="display: none" 36413 > 36414 <div class="arrow-slidedown"> 36415 <blockquote> 36416 Explaining feature importance values in models 36417 is a central concern in the realm of explainable 36418 artificial intelligence (XAI). While the Shapley 36419 value has garnered significant attention, there 36420 are other promising cooperative game theory 36421 (CGT) solutions, such as the Shapley-Shubik, 36422 that have not received the same amount of 36423 attention. In this paper, we explore the 36424 potential of the Shapley-Shubik method for 36425 elucidating feature importance values in 36426 simulations and machine learning models. 36427 </blockquote> 36428 </div> 36429 </div> 36430 </div> 36431 </div> 36432 <div class="slot-urls"></div> 36433 <a href="/wsc23papers/doc130.pdf" target="_blank">pdf</a 36434 ><br /> 36435 </div> 36436 <div class="slot-entry"> 36437 <a name="doc132" tabindex="-1"></a> 36438 <div class="slot-title-line">
36439 <span class="slot-title" 36440 >Breaking the Monotony: Promoting Diversity in 36441 High-dimensional Batch Surrogate Optimization</span 36442 > 36443 </div> 36444 <div class="slot-authors"> 36445 Nazanin Nezami (University of Illinois Chicago) 36446 </div> 36447 <div class="slot-abstract"> 36448 <div> 36449 <a 36450 class="clickable no-decoration" 36451 id="vhsjs_view_842_1707793553_0715108" 36452 onclick="$('#vhsjs_view_842_1707793553_0715108').hide(); 36453 $('#vhsjs_hide_842_1707793553_0715108').show(); 36454 $('#841_1707793553_071503').slideDown(function() { 36455 if (typeof Masonry === 'function') { 36456 $('.use_masonry').masonry(); 36457 }; 36458 36459 });" 36460 ><i class="fa fa-caret-right"></i> 36461 <span class="hover_link">Abstract</span></a 36462 ><a 36463 class="clickable no-decoration" 36464 id="vhsjs_hide_842_1707793553_0715108" 36465 onclick="$('#841_1707793553_071503').hide(function() { 36466 if (typeof Masonry === 'function') { 36467 $('.use_masonry').masonry(); 36468 }; 36469 }); 36470 $('#vhsjs_hide_842_1707793553_0715108').hide(); 36471 $('#vhsjs_view_842_1707793553_0715108').show();" 36472 style="display: none" 36473 ><i class="fa fa-caret-down"></i> 36474 <span class="hover_link">Abstract</span></a 36475 > 36476 <div 36477 data-display-control="842_1707793553_0715108" 36478 id="841_1707793553_071503" 36479 style="display: none" 36480 > 36481 <div class="arrow-slidedown"> 36482 <blockquote> 36483 In the realm of high-dimensional batch surrogate 36484 optimization, the challenge of fostering 36485 diversity while pursuing optimal solutions is 36486 paramount. Traditional approaches often result 36487 in monotonous exploration patterns, hindering 36488 the discovery of promising solutions and 36489 reducing efficiency. This thesis introduces 36490 innovative strategies, prioritizing diversity 36491 and exploration to break free from the monotony 36492 inherent in such tasks. Additionally, the thesis 36493 explores the impact of algorithmic 36494 hyperparameters on the exploration-exploitation 36495 trade-off to establish a robust framework. The 36496 "Elevating Exploration" strategies prioritize 36497 diverse candidate batch generation through 36498 adaptive sampling techniques, infusing vitality 36499 into the optimization process and effectively 36500 exploring uncharted regions of the search space. 36501 Empirical validation on optimization problems 36502 confirms their effectiveness in navigating 36503 complex landscapes. Beyond theoretical 36504 advancements and empirical validation, this 36505 thesis lays the groundwork for a paradigm shift, 36506 empowering practitioners to approach complex 36507 optimization challenges with renewed vigor and 36508 precision by promoting diversity and elevated 36509 exploration. 36510 </blockquote> 36511 </div> 36512 </div> 36513 </div> 36514 </div> 36515 <div class="slot-urls"></div> 36516 <a href="/wsc23papers/doc132.pdf" target="_blank">pdf</a 36517 ><br /> 36518 </div> 36519 <div class="slot-entry"> 36520 <a name="doc137" tabindex="-1"></a> 36521 <div class="slot-title-line">
36522 <span class="slot-title" 36523 >A Calibration Model for Bot-Like Behaviors in 36524 Agent-Based Anagram Game Simulation</span 36525 > 36526 </div> 36527 <div class="slot-authors"> 36528 Xueying Liu (Virginia Polytechnic Institute and State 36529 University) 36530 </div> 36531 <div class="slot-abstract"> 36532 <div> 36533 <a 36534 class="clickable no-decoration" 36535 id="vhsjs_view_844_1707793553_0739608" 36536 onclick="$('#vhsjs_view_844_1707793553_0739608').hide(); 36537 $('#vhsjs_hide_844_1707793553_0739608').show(); 36538 $('#843_1707793553_0739524').slideDown(function() { 36539 if (typeof Masonry === 'function') { 36540 $('.use_masonry').masonry(); 36541 }; 36542 36543 });" 36544 ><i class="fa fa-caret-right"></i> 36545 <span class="hover_link">Abstract</span></a 36546 ><a 36547 class="clickable no-decoration" 36548 id="vhsjs_hide_844_1707793553_0739608" 36549 onclick="$('#843_1707793553_0739524').hide(function() { 36550 if (typeof Masonry === 'function') { 36551 $('.use_masonry').masonry(); 36552 }; 36553 }); 36554 $('#vhsjs_hide_844_1707793553_0739608').hide(); 36555 $('#vhsjs_view_844_1707793553_0739608').show();" 36556 style="display: none" 36557 ><i class="fa fa-caret-down"></i> 36558 <span class="hover_link">Abstract</span></a 36559 > 36560 <div 36561 data-display-control="844_1707793553_0739608" 36562 id="843_1707793553_0739524" 36563 style="display: none" 36564 > 36565 <div class="arrow-slidedown"> 36566 <blockquote> 36567 Experiments that are games played among a 36568 network of players are widely used to study 36569 human behavior. Furthermore, bots or intelligent 36570 systems can be used in these games to produce 36571 contexts that elicit particular types of human 36572 responses. Bot behaviors could be specified 36573 solely based on experimental data. In this work, 36574 we take a different perspective, called the 36575 Probability Calibration (PC) approach, to 36576 simulate networked group anagram games with 36577 certain players having bot-like behaviors. The 36578 proposed method starts with data-driven models 36579 and calibrates in principled ways the parameters 36580 that alter player behaviors. It can alter the 36581 performance of each type of agent (e.g., bot) 36582 action, per player, in group anagram games. 36583 Further, statistical methods are used to test 36584 whether the PC models produce results that are 36585 statistically different from those of the 36586 original models. Case studies demonstrate the 36587 merits of the proposed method. 36588 </blockquote> 36589 </div> 36590 </div> 36591 </div> 36592 </div> 36593 <div class="slot-urls"></div> 36594 <a href="/wsc23papers/doc137.pdf" target="_blank">pdf</a 36595 ><br /> 36596 </div> 36597 <div class="slot-entry"> 36598 <a name="doc106" tabindex="-1"></a> 36599 <div class="slot-title-line">
36600 <span class="slot-title" 36601 >An Additive Decomposition for Discrete Simulation 36602 Optimization Using Gaussian Markov Random Fields</span 36603 > 36604 </div> 36605 <div class="slot-authors"> 36606 Harun Avci (Northwestern University) 36607 </div> 36608 <div class="slot-abstract"> 36609 <div> 36610 <a 36611 class="clickable no-decoration" 36612 id="vhsjs_view_846_1707793553_0765352" 36613 onclick="$('#vhsjs_view_846_1707793553_0765352').hide(); 36614 $('#vhsjs_hide_846_1707793553_0765352').show(); 36615 $('#845_1707793553_0765269').slideDown(function() { 36616 if (typeof Masonry === 'function') { 36617 $('.use_masonry').masonry(); 36618 }; 36619 36620 });" 36621 ><i class="fa fa-caret-right"></i> 36622 <span class="hover_link">Abstract</span></a 36623 ><a 36624 class="clickable no-decoration" 36625 id="vhsjs_hide_846_1707793553_0765352" 36626 onclick="$('#845_1707793553_0765269').hide(function() { 36627 if (typeof Masonry === 'function') { 36628 $('.use_masonry').masonry(); 36629 }; 36630 }); 36631 $('#vhsjs_hide_846_1707793553_0765352').hide(); 36632 $('#vhsjs_view_846_1707793553_0765352').show();" 36633 style="display: none" 36634 ><i class="fa fa-caret-down"></i> 36635 <span class="hover_link">Abstract</span></a 36636 > 36637 <div 36638 data-display-control="846_1707793553_0765352" 36639 id="845_1707793553_0765269" 36640 style="display: none" 36641 > 36642 <div class="arrow-slidedown"> 36643 <blockquote> 36644 We consider a discrete optimization via 36645 simulation problem with high-dimensional, 36646 integer-ordered decision variables. One of the 36647 methods to solve such a problem is Bayesian 36648 optimization (BO). Although BO can provide rapid 36649 solution improvement within a tight 36650 computational budget, the posterior update 36651 creates a significant computational overhead for 36652 large-scale problems. To overcome this 36653 challenge, we propose an algorithm that 36654 decomposes the prior distribution into an 36655 additive form as an approximation. Despite this 36656 approximation, our numerical analysis reveals 36657 that the algorithm can obtain rapid improvement. 36658 </blockquote> 36659 </div> 36660 </div> 36661 </div> 36662 </div> 36663 <div class="slot-urls"></div> 36664 <a href="/wsc23papers/doc106.pdf" target="_blank">pdf</a 36665 ><br /> 36666 </div> 36667 <div class="slot-entry"> 36668 <a name="doc110" tabindex="-1"></a> 36669 <div class="slot-title-line"> 36670 <span class="slot-title" 36671 >Simulation-Based Resolution of Deadlocks in Automated 36672 Guided Vehicles using Multi-Agent Reinforcement 36673 Learning in Intralogistic</span 36674 > 36675 </div> 36676 <div class="slot-authors"> 36677 Mustafa Jelibaghu (Technische Hochschule Aschaffenburg) 36678 </div> 36679 <div class="slot-abstract"> 36680 <div> 36681 <a 36682 class="clickable no-decoration" 36683 id="vhsjs_view_848_1707793553_079066" 36684 onclick="$('#vhsjs_view_848_1707793553_079066').hide(); 36685 $('#vhsjs_hide_848_1707793553_079066').show(); 36686 $('#847_1707793553_0790577').slideDown(function() { 36687 if (typeof Masonry === 'function') { 36688 $('.use_masonry').masonry(); 36689 }; 36690 36691 });" 36692 ><i class="fa fa-caret-right"></i>
36693 <span class="hover_link">Abstract</span></a 36694 ><a 36695 class="clickable no-decoration" 36696 id="vhsjs_hide_848_1707793553_079066" 36697 onclick="$('#847_1707793553_0790577').hide(function() { 36698 if (typeof Masonry === 'function') { 36699 $('.use_masonry').masonry(); 36700 }; 36701 }); 36702 $('#vhsjs_hide_848_1707793553_079066').hide(); 36703 $('#vhsjs_view_848_1707793553_079066').show();" 36704 style="display: none" 36705 ><i class="fa fa-caret-down"></i> 36706 <span class="hover_link">Abstract</span></a 36707 > 36708 <div 36709 data-display-control="848_1707793553_079066" 36710 id="847_1707793553_0790577" 36711 style="display: none" 36712 > 36713 <div class="arrow-slidedown"> 36714 <blockquote> 36715 This abstract presents a novel approach to 36716 address deadlock scenarios in Automated Guided 36717 Vehicle (AGV) systems utilizing Multi-Agent 36718 Reinforcement Learning (MARL) within a 36719 simulation framework. Deadlocks, frequently 36720 encountered in AGV operations, impede system 36721 efficiency. Traditional resolution methods can 36722 be complex and suboptimal. This study proposes a 36723 MARL-based solution, capitalizing on the 36724 decentralized decision-making prowess of agents 36725 to navigate AGVs out of deadlocks. A simulated 36726 environment accurately mimics real-world AGV 36727 dynamics, enabling agents to learn deadlock 36728 resolution strategies through trial and error. 36729 The results demonstrate that the MARL approach 36730 significantly mitigates deadlocks, enhancing 36731 overall system performance. This research 36732 contributes to the synergy between simulation, 36733 multi-agent systems, and reinforcement learning, 36734 offering an efficient deadlock resolution 36735 paradigm with potential real-world AGV 36736 application. 36737 </blockquote> 36738 </div> 36739 </div> 36740 </div> 36741 </div> 36742 <div class="slot-urls"></div> 36743 <a href="/wsc23papers/doc110.pdf" target="_blank">pdf</a 36744 ><br /> 36745 </div> 36746 <div class="slot-entry"> 36747 <a name="doc113" tabindex="-1"></a> 36748 <div class="slot-title-line"> 36749 <span class="slot-title" 36750 >How People's Beliefs Determine Society's Disease 36751 Resistence</span 36752 > 36753 </div> 36754 <div class="slot-authors"> 36755 Geonsik Yu (Purdue University) 36756 </div> 36757 <div class="slot-abstract"> 36758 <div> 36759 <a 36760 class="clickable no-decoration" 36761 id="vhsjs_view_850_1707793553_081651" 36762 onclick="$('#vhsjs_view_850_1707793553_081651').hide(); 36763 $('#vhsjs_hide_850_1707793553_081651').show(); 36764 $('#849_1707793553_0816429').slideDown(function() { 36765 if (typeof Masonry === 'function') { 36766 $('.use_masonry').masonry(); 36767 }; 36768 36769 });" 36770 ><i class="fa fa-caret-right"></i>
36771 <span class="hover_link">Abstract</span></a 36772 ><a 36773 class="clickable no-decoration" 36774 id="vhsjs_hide_850_1707793553_081651" 36775 onclick="$('#849_1707793553_0816429').hide(function() { 36776 if (typeof Masonry === 'function') { 36777 $('.use_masonry').masonry(); 36778 }; 36779 }); 36780 $('#vhsjs_hide_850_1707793553_081651').hide(); 36781 $('#vhsjs_view_850_1707793553_081651').show();" 36782 style="display: none" 36783 ><i class="fa fa-caret-down"></i> 36784 <span class="hover_link">Abstract</span></a 36785 > 36786 <div 36787 data-display-control="850_1707793553_081651" 36788 id="849_1707793553_0816429" 36789 style="display: none" 36790 > 36791 <div class="arrow-slidedown"> 36792 <blockquote> 36793 Protecting public health from infectious 36794 diseases often relies on people’s beliefs, 36795 especially when self-care interventions are the 36796 only viable tools for disease mitigation. In 36797 this study, we focus on how public opinion and 36798 its surrounding factors affect disease spread. 36799 We propose an agent-based simulation framework 36800 that incorporates opinion dynamics with an 36801 epidemic model. We demonstrate that the model 36802 can replicate the patterns of opinion and 36803 disease dynamics observed in 15 countries during 36804 the COVID-19 pandemic. Based on the fitted 36805 models, we examine how various opinion-related 36806 factors influence the consequences of the 36807 epidemic. For our explanatory model, we employ 36808 the random forest algorithm and assess the 36809 permutation importance of these factors. Partial 36810 dependence plots are also investigated to 36811 observe the direction of the factors’ 36812 impacts. Our results reveal that the initial 36813 level of public opinion on preventive 36814 interventions has a dominant impact on the total 36815 count of new infections. 36816 </blockquote> 36817 </div> 36818 </div> 36819 </div> 36820 </div> 36821 <div class="slot-urls"></div> 36822 <a href="/wsc23papers/doc113.pdf" target="_blank">pdf</a 36823 ><br /> 36824 </div> 36825 <div class="slot-entry"> 36826 <a name="doc115" tabindex="-1"></a> 36827 <div class="slot-title-line"> 36828 <span class="slot-title" 36829 >Marine Ecosystem Services Disruption and Social 36830 Violence</span 36831 > 36832 </div> 36833 <div class="slot-authors"> 36834 Rafael Hurtado (University of Central Florida) 36835 </div> 36836 <div class="slot-abstract"> 36837 <div> 36838 <a 36839 class="clickable no-decoration" 36840 id="vhsjs_view_852_1707793553_0842397" 36841 onclick="$('#vhsjs_view_852_1707793553_0842397').hide(); 36842 $('#vhsjs_hide_852_1707793553_0842397').show(); 36843 $('#851_1707793553_0842314').slideDown(function() { 36844 if (typeof Masonry === 'function') { 36845 $('.use_masonry').masonry(); 36846 }; 36847 36848 });" 36849 ><i class="fa fa-caret-right"></i>
36850 <span class="hover_link">Abstract</span></a 36851 ><a 36852 class="clickable no-decoration" 36853 id="vhsjs_hide_852_1707793553_0842397" 36854 onclick="$('#851_1707793553_0842314').hide(function() { 36855 if (typeof Masonry === 'function') { 36856 $('.use_masonry').masonry(); 36857 }; 36858 }); 36859 $('#vhsjs_hide_852_1707793553_0842397').hide(); 36860 $('#vhsjs_view_852_1707793553_0842397').show();" 36861 style="display: none" 36862 ><i class="fa fa-caret-down"></i> 36863 <span class="hover_link">Abstract</span></a 36864 > 36865 <div 36866 data-display-control="852_1707793553_0842397" 36867 id="851_1707793553_0842314" 36868 style="display: none" 36869 > 36870 <div class="arrow-slidedown"> 36871 <blockquote> 36872 Marine ecosystem services support coastal 36873 communities by offering essential sustenance, 36874 protection, and cultural benefits. However, the 36875 global decline in these ecosystems has disrupted 36876 these services, impacting the communities 36877 reliant on them. The Archipelago of San Andres 36878 Providencia and Santa Catalina (ASAPSC) in the 36879 Colombian Caribbean exemplifies this decline, 36880 coinciding with a rise in violent crimes and 36881 homicide rates. This study employs an 36882 agent-based model (ABM) to simulate the ASAPSC 36883 case and examine the potential links between 36884 marine ecosystem depletion and the escalation of 36885 social violence. The simulation results suggest 36886 a link between disruption of ecosystem services 36887 and social violence and set the stage for future 36888 empirical research in environmental security. 36889 </blockquote> 36890 </div> 36891 </div> 36892 </div> 36893 </div> 36894 <div class="slot-urls"></div> 36895 <a href="/wsc23papers/doc115.pdf" target="_blank">pdf</a 36896 ><br /> 36897 </div> 36898 <div class="slot-entry"> 36899 <a name="doc116" tabindex="-1"></a> 36900 <div class="slot-title-line"> 36901 <span class="slot-title" 36902 >Focused Flexibility in Workforce Scheduling</span 36903 > 36904 </div> 36905 <div class="slot-authors"> 36906 Johanna Wiesflecker (The University of Edinburgh) 36907 </div> 36908 <div class="slot-abstract"> 36909 <div> 36910 <a 36911 class="clickable no-decoration" 36912 id="vhsjs_view_854_1707793553_0867696" 36913 onclick="$('#vhsjs_view_854_1707793553_0867696').hide(); 36914 $('#vhsjs_hide_854_1707793553_0867696').show(); 36915 $('#853_1707793553_0867612').slideDown(function() { 36916 if (typeof Masonry === 'function') { 36917 $('.use_masonry').masonry(); 36918 }; 36919 36920 });" 36921 ><i class="fa fa-caret-right"></i> 36922 <span class="hover_link">Abstract</span></a 36923 ><a 36924 class="clickable no-decoration" 36925 id="vhsjs_hide_854_1707793553_0867696" 36926 onclick="$('#853_1707793553_0867612').hide(function() { 36927 if (typeof Masonry === 'function') { 36928 $('.use_masonry').masonry(); 36929 }; 36930 }); 36931 $('#vhsjs_hide_854_1707793553_0867696').hide(); 36932 $('#vhsjs_view_854_1707793553_0867696').show();" 36933 style="display: none" 36934 ><i class="fa fa-caret-down"></i>
36935 <span class="hover_link">Abstract</span></a 36936 > 36937 <div 36938 data-display-control="854_1707793553_0867696" 36939 id="853_1707793553_0867612" 36940 style="display: none" 36941 > 36942 <div class="arrow-slidedown"> 36943 <blockquote> 36944 In many industries, work schedules often go 36945 through lengthy approval processes. Once 36946 approved, schedules may be locked in for long 36947 time horizons (e.g., months). Working 36948 regulations allow for partial changes 36949 (re-rostering) in a small number of extreme 36950 cases. Most other disruptions (staff 36951 absenteeism, change in demand pattern, etc.) 36952 will be dealt with only at huge costs. Injecting 36953 flexibility (affordable, case-specific 36954 re-rostering options) from the very outset 36955 (schedule approval stage) can foster schedule 36956 robustness at lower costs. This work shows how 36957 to jointly adopt simulation and Adaptive Large 36958 Neighborhood Search to do just that. At each 36959 iteration of the proposed Sim-ALNS algorithm, 36960 ALNS selects a combination of levels of 36961 flexibility (within guidelines set by the 36962 organization), while a Monte-Carlo simulation 36963 scheme evaluates the performance of the 36964 solution. Experiments in an airport security 36965 setting show that the method leads to a 27% 36966 decrease in average weekly re-rostering cost. 36967 </blockquote> 36968 </div> 36969 </div> 36970 </div> 36971 </div> 36972 <div class="slot-urls"></div> 36973 <a href="/wsc23papers/doc116.pdf" target="_blank">pdf</a 36974 ><br /> 36975 </div> 36976 <div class="slot-entry"> 36977 <a name="doc118" tabindex="-1"></a> 36978 <div class="slot-title-line"> 36979 <span class="slot-title" 36980 >A Combined Simulation Optimization Framework to 36981 Improve Logistics Processes in the Production of 36982 Specialty Chemicals</span 36983 > 36984 </div> 36985 <div class="slot-authors"> 36986 Maximilian Kiefer (TU Dortmund University, Graduate 36987 School of Logistics / Institute of Transport Logistics) 36988 </div> 36989 <div class="slot-abstract"> 36990 <div> 36991 <a 36992 class="clickable no-decoration" 36993 id="vhsjs_view_856_1707793553_0893302" 36994 onclick="$('#vhsjs_view_856_1707793553_0893302').hide(); 36995 $('#vhsjs_hide_856_1707793553_0893302').show(); 36996 $('#855_1707793553_0893216').slideDown(function() { 36997 if (typeof Masonry === 'function') { 36998 $('.use_masonry').masonry(); 36999 }; 37000 37001 });" 37002 ><i class="fa fa-caret-right"></i> 37003 <span class="hover_link">Abstract</span></a 37004 ><a 37005 class="clickable no-decoration" 37006 id="vhsjs_hide_856_1707793553_0893302" 37007 onclick="$('#855_1707793553_0893216').hide(function() { 37008 if (typeof Masonry === 'function') { 37009 $('.use_masonry').masonry(); 37010 }; 37011 }); 37012 $('#vhsjs_hide_856_1707793553_0893302').hide(); 37013 $('#vhsjs_view_856_1707793553_0893302').show();" 37014 style="display: none" 37015 ><i class="fa fa-caret-down"></i>
37016 <span class="hover_link">Abstract</span></a 37017 > 37018 <div 37019 data-display-control="856_1707793553_0893302" 37020 id="855_1707793553_0893216" 37021 style="display: none" 37022 > 37023 <div class="arrow-slidedown"> 37024 <blockquote> 37025 The chemical industry is experiencing shifts in 37026 market conditions, leading to an increasing need 37027 for fast and individual-engineered chemicals. 37028 This trend causes a change from mass production 37029 to the production of small, demand-driven 37030 quantities. This results in various variants and 37031 container types, requiring efficient logistics 37032 management to handle the complexity. A 37033 methodical framework should enable the user to 37034 fulfill the specific requirements of the 37035 logistics processes and thus make the complex 37036 planning manageable. In particular, supply and 37037 disposal methods and container management are 37038 under special consideration. Therefore, a 37039 simulation and optimization framework is 37040 developed. First, the motivation of the research 37041 project is presented. Afterward, a framework for 37042 planning logistics processes is designed, 37043 consisting of data preparation, mathematical 37044 optimization, and simulation. 37045 </blockquote> 37046 </div> 37047 </div> 37048 </div> 37049 </div> 37050 <div class="slot-urls"></div> 37051 <a href="/wsc23papers/doc118.pdf" target="_blank">pdf</a 37052 ><br /> 37053 </div> 37054 </div> 37055 <div class="session-entry"> 37056 <span class="session-event-type">PhD Colloquium</span 37057 ><span class="type-track-spacer"> · </span 37058 ><span class="program-track">PhD Colloquium</span><br /> 37059 <div class="session-title">PhD Colloquium Session A2</div> 37060 <div class="session-chair"> 37061 Chair: Siyang Gao (City University of Hong Kong)<br /> 37062 </div> 37063 <div class="slot-entry"> 37064 <a name="doc124" tabindex="-1"></a> 37065 <div class="slot-title-line"> 37066 <span class="slot-title" 37067 >Computer Simulation-based Templates for Lean 37068 Implementation in Small and Medium Construction 37069 Enterprises</span 37070 > 37071 </div> 37072 <div class="slot-authors"> 37073 Prashanth Kumar Sreram (Indian Institute of Technology 37074 Bombay, National Institute of Construction Management 37075 and Research Hyderabad) 37076 </div> 37077 <div class="slot-abstract"> 37078 <div> 37079 <a 37080 class="clickable no-decoration" 37081 id="vhsjs_view_858_1707793553_094595" 37082 onclick="$('#vhsjs_view_858_1707793553_094595').hide(); 37083 $('#vhsjs_hide_858_1707793553_094595').show(); 37084 $('#857_1707793553_094587').slideDown(function() { 37085 if (typeof Masonry === 'function') { 37086 $('.use_masonry').masonry(); 37087 }; 37088 37089 });" 37090 ><i class="fa fa-caret-right"></i>
37091 <span class="hover_link">Abstract</span></a 37092 ><a 37093 class="clickable no-decoration" 37094 id="vhsjs_hide_858_1707793553_094595" 37095 onclick="$('#857_1707793553_094587').hide(function() { 37096 if (typeof Masonry === 'function') { 37097 $('.use_masonry').masonry(); 37098 }; 37099 }); 37100 $('#vhsjs_hide_858_1707793553_094595').hide(); 37101 $('#vhsjs_view_858_1707793553_094595').show();" 37102 style="display: none" 37103 ><i class="fa fa-caret-down"></i> 37104 <span class="hover_link">Abstract</span></a 37105 > 37106 <div 37107 data-display-control="858_1707793553_094595" 37108 id="857_1707793553_094587" 37109 style="display: none" 37110 > 37111 <div class="arrow-slidedown"> 37112 <blockquote> 37113 A country's economic advancement hinges on the 37114 construction sector, but its growth is marred by 37115 the global construction industry's chief 37116 predicament: tangible and intangible waste. Lean 37117 construction employs strategies such as Value 37118 Stream Mapping (VSM), yielding crucial time and 37119 cost savings. Presently, VSM's execution is 37120 limited to static process representation, 37121 segregating preparation, and assessment of 37122 enhancement alternatives. In the era of 37123 construction 4.0, embracing technological and 37124 digital shifts is imperative, enhancing 37125 performance via simulation. Hence, uniting Lean 37126 Construction with Simulation becomes essential, 37127 validating lean principles through simulation 37128 models and aiding improved project 37129 decision-making. Thus, research concentrates on 37130 crafting VSM-based discrete event simulation 37131 (DES) models tailored for small and medium 37132 enterprises in the offsite construction realm. 37133 The current focus is offsite construction, while 37134 forthcoming research addresses complex 37135 activities, refining simulation models as 37136 valuable tools for industry practitioners. 37137 </blockquote> 37138 </div> 37139 </div> 37140 </div> 37141 </div> 37142 <div class="slot-urls"></div> 37143 <a href="/wsc23papers/doc124.pdf" target="_blank">pdf</a 37144 ><br /> 37145 </div> 37146 <div class="slot-entry"> 37147 <a name="doc125" tabindex="-1"></a> 37148 <div class="slot-title-line"> 37149 <span class="slot-title" 37150 >Causal Dynamic Bayesian Networks for Simulation 37151 Metamodeling</span 37152 > 37153 </div> 37154 <div class="slot-authors"> 37155 Pracheta Amaranath (University of Massachusetts Amherst) 37156 </div> 37157 <div class="slot-abstract"> 37158 <div> 37159 <a 37160 class="clickable no-decoration" 37161 id="vhsjs_view_860_1707793553_0971394" 37162 onclick="$('#vhsjs_view_860_1707793553_0971394').hide(); 37163 $('#vhsjs_hide_860_1707793553_0971394').show(); 37164 $('#859_1707793553_0971315').slideDown(function() { 37165 if (typeof Masonry === 'function') { 37166 $('.use_masonry').masonry(); 37167 }; 37168 37169 });" 37170 ><i class="fa fa-caret-right"></i>
37171 <span class="hover_link">Abstract</span></a 37172 ><a 37173 class="clickable no-decoration" 37174 id="vhsjs_hide_860_1707793553_0971394" 37175 onclick="$('#859_1707793553_0971315').hide(function() { 37176 if (typeof Masonry === 'function') { 37177 $('.use_masonry').masonry(); 37178 }; 37179 }); 37180 $('#vhsjs_hide_860_1707793553_0971394').hide(); 37181 $('#vhsjs_view_860_1707793553_0971394').show();" 37182 style="display: none" 37183 ><i class="fa fa-caret-down"></i> 37184 <span class="hover_link">Abstract</span></a 37185 > 37186 <div 37187 data-display-control="860_1707793553_0971394" 37188 id="859_1707793553_0971315" 37189 style="display: none" 37190 > 37191 <div class="arrow-slidedown"> 37192 <blockquote> 37193 A traditional metamodel for a discrete-event 37194 simulation approximates a real-valued 37195 performance measure as a function of the 37196 input-parameter values. We introduce a novel 37197 class of metamodels based on modular dynamic 37198 Bayesian networks (MDBNs), a subclass of 37199 probabilistic graphical models which can be used 37200 to efficiently answer a rich class of 37201 probabilistic and causal queries (PCQs). Such 37202 queries represent the joint probability 37203 distribution of the system state at multiple 37204 time points, given observations of, and 37205 interventions on, other state variables and 37206 input parameters. This paper is a first 37207 demonstration of how the extensive theory and 37208 technology of causal graphical models can be 37209 used to enhance simulation metamodeling. We 37210 demonstrate this potential by showing how a 37211 single MDBN for an M/M/1 queue can be learned 37212 from simulation data and then be used to quickly 37213 and accurately answer a variety of PCQs, most of 37214 which are out-of-scope for existing metamodels. 37215 </blockquote> 37216 </div> 37217 </div> 37218 </div> 37219 </div> 37220 <div class="slot-urls"></div> 37221 <a href="/wsc23papers/doc125.pdf" target="_blank">pdf</a 37222 ><br /> 37223 </div> 37224 <div class="slot-entry"> 37225 <a name="doc129" tabindex="-1"></a> 37226 <div class="slot-title-line"> 37227 <span class="slot-title" 37228 >Improving Buffer Storage Performance in Ceramic Tile 37229 Industry Via Simulation</span 37230 > 37231 </div> 37232 <div class="slot-authors"> 37233 Marco Taccini (University of Modena and Reggio Emilia) 37234 </div> 37235 <div class="slot-abstract"> 37236 <div> 37237 <a 37238 class="clickable no-decoration" 37239 id="vhsjs_view_862_1707793553_0996168" 37240 onclick="$('#vhsjs_view_862_1707793553_0996168').hide(); 37241 $('#vhsjs_hide_862_1707793553_0996168').show(); 37242 $('#861_1707793553_0996087').slideDown(function() { 37243 if (typeof Masonry === 'function') { 37244 $('.use_masonry').masonry(); 37245 }; 37246 37247 });" 37248 ><i class="fa fa-caret-right"></i>
37249 <span class="hover_link">Abstract</span></a 37250 ><a 37251 class="clickable no-decoration" 37252 id="vhsjs_hide_862_1707793553_0996168" 37253 onclick="$('#861_1707793553_0996087').hide(function() { 37254 if (typeof Masonry === 'function') { 37255 $('.use_masonry').masonry(); 37256 }; 37257 }); 37258 $('#vhsjs_hide_862_1707793553_0996168').hide(); 37259 $('#vhsjs_view_862_1707793553_0996168').show();" 37260 style="display: none" 37261 ><i class="fa fa-caret-down"></i> 37262 <span class="hover_link">Abstract</span></a 37263 > 37264 <div 37265 data-display-control="862_1707793553_0996168" 37266 id="861_1707793553_0996087" 37267 style="display: none" 37268 > 37269 <div class="arrow-slidedown"> 37270 <blockquote> 37271 This study aims at identifying the best strategy 37272 to temporarily store products within a buffer 37273 area in an Italian ceramic tile company. The 37274 storage policy is analyzed to maximize the 37275 storage capacity, facilitate operators' 37276 activities, and, consequently, improve the 37277 warehouse logistics performance. A discrete 37278 event simulation was conducted using Salabim, a 37279 Python based open-source software, in order to 37280 determine the best policy. We compare the 37281 performance of the current storage policy, based 37282 on technical production properties of products, 37283 and a newly proposed one, based on products' 37284 downstream destination. The results suggested 37285 that the proposed strategy significantly 37286 improves the performance of the buffer area 37287 management. The approach can be applied to 37288 different applications, contributing to the 37289 literature on simulation-based decision-making 37290 in material management. Furthermore, the study 37291 provides a functional case study showing the 37292 potential and achievable results of Salabim for 37293 modeling complex systems. 37294 </blockquote> 37295 </div> 37296 </div> 37297 </div> 37298 </div> 37299 <div class="slot-urls"></div> 37300 <a href="/wsc23papers/doc129.pdf" target="_blank">pdf</a 37301 ><br /> 37302 </div> 37303 <div class="slot-entry"> 37304 <a name="doc133" tabindex="-1"></a> 37305 <div class="slot-title-line"> 37306 <span class="slot-title" 37307 >Integrating AI and Simulation for Intelligent 37308 Material Handling</span 37309 > 37310 </div> 37311 <div class="slot-authors"> 37312 Sriparvathi Shaji Bhattathiri (Rochester Institute of 37313 Technology) 37314 </div> 37315 <div class="slot-abstract"> 37316 <div> 37317 <a 37318 class="clickable no-decoration" 37319 id="vhsjs_view_864_1707793553_1021829" 37320 onclick="$('#vhsjs_view_864_1707793553_1021829').hide(); 37321 $('#vhsjs_hide_864_1707793553_1021829').show(); 37322 $('#863_1707793553_1021745').slideDown(function() { 37323 if (typeof Masonry === 'function') { 37324 $('.use_masonry').masonry(); 37325 }; 37326 37327 });" 37328 ><i class="fa fa-caret-right"></i>
37329 <span class="hover_link">Abstract</span></a 37330 ><a 37331 class="clickable no-decoration" 37332 id="vhsjs_hide_864_1707793553_1021829" 37333 onclick="$('#863_1707793553_1021745').hide(function() { 37334 if (typeof Masonry === 'function') { 37335 $('.use_masonry').masonry(); 37336 }; 37337 }); 37338 $('#vhsjs_hide_864_1707793553_1021829').hide(); 37339 $('#vhsjs_view_864_1707793553_1021829').show();" 37340 style="display: none" 37341 ><i class="fa fa-caret-down"></i> 37342 <span class="hover_link">Abstract</span></a 37343 > 37344 <div 37345 data-display-control="864_1707793553_1021829" 37346 id="863_1707793553_1021745" 37347 style="display: none" 37348 > 37349 <div class="arrow-slidedown"> 37350 <blockquote> 37351 With the increasing integration of autonomous 37352 mobile robots in warehouse facilities for 37353 storage and retrieval, the need arises to make 37354 intelligent dispatching decisions to maximize 37355 operational efficiency and meet shipping 37356 deadlines. The aim of this research is to enable 37357 effective real-time, dispatching decisions 37358 taking into consideration both travel distance 37359 and due date. In particular, we develop a 37360 reinforcement learning method for task selection 37361 in a multi-agent warehouse environment. A Monte 37362 Carlo simulation approach is used to train the 37363 Artificial Intelligence model and assess its 37364 capabilities and limitations. The performance of 37365 the proposed model is compared with that of 37366 rule-based task selection methods. The 37367 preliminary experimental results indicate strong 37368 potential in employing reinforcement learning 37369 for real-time dispatch in warehouse 37370 environments. 37371 </blockquote> 37372 </div> 37373 </div> 37374 </div> 37375 </div> 37376 <div class="slot-urls"></div> 37377 <a href="/wsc23papers/doc133.pdf" target="_blank">pdf</a 37378 ><br /> 37379 </div> 37380 <div class="slot-entry"> 37381 <a name="doc136" tabindex="-1"></a> 37382 <div class="slot-title-line"> 37383 <span class="slot-title" 37384 >Model Predictive Control in Optimal Intervention of 37385 Covid-19 with Mixed Epistemic-aleatoric 37386 Uncertainty</span 37387 > 37388 </div> 37389 <div class="slot-authors"> 37390 Jinming Wan (Binghamton University) 37391 </div> 37392 <div class="slot-abstract"> 37393 <div> 37394 <a 37395 class="clickable no-decoration" 37396 id="vhsjs_view_866_1707793553_1054616" 37397 onclick="$('#vhsjs_view_866_1707793553_1054616').hide(); 37398 $('#vhsjs_hide_866_1707793553_1054616').show(); 37399 $('#865_1707793553_105447').slideDown(function() { 37400 if (typeof Masonry === 'function') { 37401 $('.use_masonry').masonry(); 37402 }; 37403 37404 });" 37405 ><i class="fa fa-caret-right"></i>
37406 <span class="hover_link">Abstract</span></a 37407 ><a 37408 class="clickable no-decoration" 37409 id="vhsjs_hide_866_1707793553_1054616" 37410 onclick="$('#865_1707793553_105447').hide(function() { 37411 if (typeof Masonry === 'function') { 37412 $('.use_masonry').masonry(); 37413 }; 37414 }); 37415 $('#vhsjs_hide_866_1707793553_1054616').hide(); 37416 $('#vhsjs_view_866_1707793553_1054616').show();" 37417 style="display: none" 37418 ><i class="fa fa-caret-down"></i> 37419 <span class="hover_link">Abstract</span></a 37420 > 37421 <div 37422 data-display-control="866_1707793553_1054616" 37423 id="865_1707793553_105447" 37424 style="display: none" 37425 > 37426 <div class="arrow-slidedown"> 37427 <blockquote> 37428 Non-pharmaceutical interventions (NPI) have been 37429 proven vital in the fight against the COVID-19 37430 pandemic before the massive rollout of 37431 vaccinations. Considering the inherent 37432 epistemic-aleatoric uncertainty of parameters, 37433 accurate simulation and modeling of the 37434 interplay between the NPI and contagion dynamics 37435 are critical to the optimal design of 37436 intervention policies. We propose a modified 37437 SIRD-MPC model that combines a modified 37438 stochastic 37439 Susceptible-Infected-Recovered-Deceased (SIRD) 37440 compartment model with mixed epistemic-aleatoric 37441 parameters and Model Predictive Control (MPC), 37442 to develop robust NPI control policies to 37443 contain the infection of the COVID-19 pandemic 37444 with minimum economic impact. 37445 </blockquote> 37446 </div> 37447 </div> 37448 </div> 37449 </div> 37450 <div class="slot-urls"></div> 37451 <a href="/wsc23papers/doc136.pdf" target="_blank">pdf</a 37452 ><br /> 37453 </div> 37454 <div class="slot-entry"> 37455 <a name="doc142" tabindex="-1"></a> 37456 <div class="slot-title-line"> 37457 <span class="slot-title" 37458 >Perishable Inventory Management: Human Milk Banking 37459 Case Study</span 37460 > 37461 </div> 37462 <div class="slot-authors"> 37463 Marta Staff (University of Exeter) 37464 </div> 37465 <div class="slot-abstract"> 37466 <div> 37467 <a 37468 class="clickable no-decoration" 37469 id="vhsjs_view_868_1707793553_1090696" 37470 onclick="$('#vhsjs_view_868_1707793553_1090696').hide(); 37471 $('#vhsjs_hide_868_1707793553_1090696').show(); 37472 $('#867_1707793553_109056').slideDown(function() { 37473 if (typeof Masonry === 'function') { 37474 $('.use_masonry').masonry(); 37475 }; 37476 37477 });" 37478 ><i class="fa fa-caret-right"></i> 37479 <span class="hover_link">Abstract</span></a 37480 ><a 37481 class="clickable no-decoration" 37482 id="vhsjs_hide_868_1707793553_1090696" 37483 onclick="$('#867_1707793553_109056').hide(function() { 37484 if (typeof Masonry === 'function') { 37485 $('.use_masonry').masonry(); 37486 }; 37487 }); 37488 $('#vhsjs_hide_868_1707793553_1090696').hide(); 37489 $('#vhsjs_view_868_1707793553_1090696').show();" 37490 style="display: none" 37491 ><i class="fa fa-caret-down"></i>
37492 <span class="hover_link">Abstract</span></a 37493 > 37494 <div 37495 data-display-control="868_1707793553_1090696" 37496 id="867_1707793553_109056" 37497 style="display: none" 37498 > 37499 <div class="arrow-slidedown"> 37500 <blockquote> 37501 Despite providing lifesaving donor human milk to 37502 vulnerable premature infants, human milk banking 37503 is greatly overlooked from an Operations 37504 Research perspective, with yet to be explored 37505 distinctive characteristics, offering attractive 37506 prospects for Modelling and Simulation research. 37507 The effective management of inventory, where 37508 products have limited shelf life, adds to its 37509 complexity. The commonly utilized newsvendor 37510 model to study inventory decisions is unlikely 37511 to capture the intricacies of items with 37512 extended shelf lives. A milk donor typically 37513 accumulates milk over time, resulting in the 37514 donation of a “stash” consisting of 37515 milk units with different expiry dates. The 37516 decision of whether to treat it as a whole, or 37517 split it, when the “stash” is 37518 progressed out of the ingress inventory into 37519 production, will affect the remaining shelf life 37520 of the final product, but also the associated 37521 production costs. Hence DES is being utilized to 37522 investigate the cost-benefit analysis of batch 37523 splitting. 37524 </blockquote> 37525 </div> 37526 </div> 37527 </div> 37528 </div> 37529 <div class="slot-urls"></div> 37530 <a href="/wsc23papers/doc142.pdf" target="_blank">pdf</a 37531 ><br /> 37532 </div> 37533 <div class="slot-entry"> 37534 <a name="doc143" tabindex="-1"></a> 37535 <div class="slot-title-line"> 37536 <span class="slot-title" 37537 >Estimating Treatment Effects from Simulation Samples 37538 of Population-scale Models</span 37539 > 37540 </div> 37541 <div class="slot-authors"> 37542 Abdulrahman Ahmed (University of Pittsburgh) 37543 </div> 37544 <div class="slot-abstract"> 37545 <div> 37546 <a 37547 class="clickable no-decoration" 37548 id="vhsjs_view_870_1707793553_1130123" 37549 onclick="$('#vhsjs_view_870_1707793553_1130123').hide(); 37550 $('#vhsjs_hide_870_1707793553_1130123').show(); 37551 $('#869_1707793553_1129987').slideDown(function() { 37552 if (typeof Masonry === 'function') { 37553 $('.use_masonry').masonry(); 37554 }; 37555 37556 });" 37557 ><i class="fa fa-caret-right"></i> 37558 <span class="hover_link">Abstract</span></a 37559 ><a 37560 class="clickable no-decoration" 37561 id="vhsjs_hide_870_1707793553_1130123" 37562 onclick="$('#869_1707793553_1129987').hide(function() { 37563 if (typeof Masonry === 'function') { 37564 $('.use_masonry').masonry(); 37565 }; 37566 }); 37567 $('#vhsjs_hide_870_1707793553_1130123').hide(); 37568 $('#vhsjs_view_870_1707793553_1130123').show();" 37569 style="display: none" 37570 ><i class="fa fa-caret-down"></i>
37571 <span class="hover_link">Abstract</span></a 37572 > 37573 <div 37574 data-display-control="870_1707793553_1130123" 37575 id="869_1707793553_1129987" 37576 style="display: none" 37577 > 37578 <div class="arrow-slidedown"> 37579 <blockquote> 37580 Large-scale models require an exhaustive amount 37581 of computational power to simulate, especially 37582 when there are multiple treatment conditions to 37583 be evaluated across large geographical regions. 37584 Therefore, developing an efficient method to 37585 distribute computational resources efficiently 37586 is essential for conducting large-scale 37587 simulations. Agent-based modeling can generate 37588 accurate simulation samples, and our goal is to 37589 use them for estimating treatment effects to 37590 optimize potential interventions with as few 37591 simulation samples as possible. In this 37592 abstract, I will show methods that perform 37593 better than benchmarks by taking into account 37594 the uncertainty in the estimation of treatment 37595 effects dynamically and discuss our next steps 37596 for improving them. 37597 </blockquote> 37598 </div> 37599 </div> 37600 </div> 37601 </div> 37602 <div class="slot-urls"></div> 37603 <a href="/wsc23papers/doc143.pdf" target="_blank">pdf</a 37604 ><br /> 37605 </div> 37606 <div class="slot-entry"> 37607 <a name="doc144" tabindex="-1"></a> 37608 <div class="slot-title-line"> 37609 <span class="slot-title" 37610 >Adaptive Ranking and Selection Based Genetic 37611 Algorithms For Data-driven Problems</span 37612 > 37613 </div> 37614 <div class="slot-authors"> 37615 Kimia Vahdat (North Carolina State University) 37616 </div> 37617 <div class="slot-abstract"> 37618 <div> 37619 <a 37620 class="clickable no-decoration" 37621 id="vhsjs_view_872_1707793553_1170123" 37622 onclick="$('#vhsjs_view_872_1707793553_1170123').hide(); 37623 $('#vhsjs_hide_872_1707793553_1170123').show(); 37624 $('#871_1707793553_1169987').slideDown(function() { 37625 if (typeof Masonry === 'function') { 37626 $('.use_masonry').masonry(); 37627 }; 37628 37629 });" 37630 ><i class="fa fa-caret-right"></i> 37631 <span class="hover_link">Abstract</span></a 37632 ><a 37633 class="clickable no-decoration" 37634 id="vhsjs_hide_872_1707793553_1170123" 37635 onclick="$('#871_1707793553_1169987').hide(function() { 37636 if (typeof Masonry === 'function') { 37637 $('.use_masonry').masonry(); 37638 }; 37639 }); 37640 $('#vhsjs_hide_872_1707793553_1170123').hide(); 37641 $('#vhsjs_view_872_1707793553_1170123').show();" 37642 style="display: none" 37643 ><i class="fa fa-caret-down"></i> 37644 <span class="hover_link">Abstract</span></a 37645 > 37646 <div
37647 data-display-control="872_1707793553_1170123" 37648 id="871_1707793553_1169987" 37649 style="display: none" 37650 > 37651 <div class="arrow-slidedown"> 37652 <blockquote> 37653 We present ARGA, the Adaptive Robust Genetic 37654 Algorithm, for optimizing simulation problems 37655 with binary variables affected by input 37656 uncertainty and Monte Carlo noise. In this 37657 method, a population evolves as more information 37658 about the high-dimensional, stochastic problem 37659 becomes available. ARGA conducts ranking and 37660 selection with a debiasing mechanism of fitness 37661 values using fast iterated bootstraps economized 37662 with control variates. Debiasing reduces the 37663 model risk due to input uncertainty bias, 37664 leading to a more accurate ranking of designs. 37665 Given the double loop of function evaluations, 37666 we incorporate adaptive budget allocation 37667 throughout the search only if the current 37668 population's proximity to optimality signals the 37669 need for a smaller standard error. In that case, 37670 we allocate replications to the input model of 37671 the design most responsible for risk. Empirical 37672 results with a fixed optimization budget show 37673 that ARGA obtains significantly better solutions 37674 in feature selection problems across various 37675 datasets. 37676 </blockquote> 37677 </div> 37678 </div> 37679 </div> 37680 </div> 37681 <div class="slot-urls"></div> 37682 <a href="/wsc23papers/doc144.pdf" target="_blank">pdf</a 37683 ><br /> 37684 </div> 37685 <div class="slot-entry"> 37686 <a name="doc107" tabindex="-1"></a> 37687 <div class="slot-title-line"> 37688 <span class="slot-title" 37689 >Enhancing Parallel Large-Scale Ranking and Selection 37690 Using Clustering Techniques</span 37691 > 37692 </div> 37693 <div class="slot-authors"> 37694 Zishi Zhang (Guanghua School of Management,Peking 37695 University) 37696 </div> 37697 <div class="slot-abstract"> 37698 <div> 37699 <a 37700 class="clickable no-decoration" 37701 id="vhsjs_view_874_1707793553_1220815" 37702 onclick="$('#vhsjs_view_874_1707793553_1220815').hide(); 37703 $('#vhsjs_hide_874_1707793553_1220815').show(); 37704 $('#873_1707793553_1220677').slideDown(function() { 37705 if (typeof Masonry === 'function') { 37706 $('.use_masonry').masonry(); 37707 }; 37708 37709 });" 37710 ><i class="fa fa-caret-right"></i> 37711 <span class="hover_link">Abstract</span></a 37712 ><a 37713 class="clickable no-decoration" 37714 id="vhsjs_hide_874_1707793553_1220815" 37715 onclick="$('#873_1707793553_1220677').hide(function() { 37716 if (typeof Masonry === 'function') { 37717 $('.use_masonry').masonry(); 37718 }; 37719 }); 37720 $('#vhsjs_hide_874_1707793553_1220815').hide(); 37721 $('#vhsjs_view_874_1707793553_1220815').show();" 37722 style="display: none" 37723 ><i class="fa fa-caret-down"></i>
37724 <span class="hover_link">Abstract</span></a 37725 > 37726 <div 37727 data-display-control="874_1707793553_1220815" 37728 id="873_1707793553_1220677" 37729 style="display: none" 37730 > 37731 <div class="arrow-slidedown"> 37732 <blockquote> 37733 We explore the use of correlation-based 37734 clustering techniques to enhance large-scale R&S 37735 procedures under parallel computing environment. 37736 Both theoretical analysis and numerical 37737 experiments convincingly demonstrate that 37738 clustering techniques can significantly improve 37739 the sample efficiency of existing R&S methods. 37740 </blockquote> 37741 </div> 37742 </div> 37743 </div> 37744 </div> 37745 <div class="slot-urls"></div> 37746 <a href="/wsc23papers/doc107.pdf" target="_blank">pdf</a 37747 ><br /> 37748 </div> 37749 <div class="slot-entry"> 37750 <a name="doc109" tabindex="-1"></a> 37751 <div class="slot-title-line"> 37752 <span class="slot-title" 37753 >Reliable Adaptive Stochastic Optimization with High 37754 Probability Guarantees</span 37755 > 37756 </div> 37757 <div class="slot-authors"> 37758 Miaolan Xie (Cornell University) 37759 </div> 37760 <div class="slot-abstract"> 37761 <div> 37762 <a 37763 class="clickable no-decoration" 37764 id="vhsjs_view_876_1707793553_126048" 37765 onclick="$('#vhsjs_view_876_1707793553_126048').hide(); 37766 $('#vhsjs_hide_876_1707793553_126048').show(); 37767 $('#875_1707793553_126035').slideDown(function() { 37768 if (typeof Masonry === 'function') { 37769 $('.use_masonry').masonry(); 37770 }; 37771 37772 });" 37773 ><i class="fa fa-caret-right"></i> 37774 <span class="hover_link">Abstract</span></a 37775 ><a 37776 class="clickable no-decoration" 37777 id="vhsjs_hide_876_1707793553_126048" 37778 onclick="$('#875_1707793553_126035').hide(function() { 37779 if (typeof Masonry === 'function') { 37780 $('.use_masonry').masonry(); 37781 }; 37782 }); 37783 $('#vhsjs_hide_876_1707793553_126048').hide(); 37784 $('#vhsjs_view_876_1707793553_126048').show();" 37785 style="display: none" 37786 ><i class="fa fa-caret-down"></i> 37787 <span class="hover_link">Abstract</span></a 37788 > 37789 <div 37790 data-display-control="876_1707793553_126048" 37791 id="875_1707793553_126035" 37792 style="display: none" 37793 > 37794 <div class="arrow-slidedown"> 37795 <blockquote> 37796 To handle real-world data that is noisy, biased 37797 and even corrupted, we consider a simple 37798 adaptive framework for stochastic optimization 37799 where the step size is adaptively adjusted 37800 according to the algorithm's progress instead of 37801 manual tuning or using a pre-specified sequence. 37802 Function value, gradient and possibly Hessian 37803 estimates are provided by probabilistic oracles 37804 and can be biased and arbitrarily corrupted, 37805 capturing multiple settings including expected 37806 loss minimization in machine learning, 37807 zeroth-order and low-precision optimization. 37808 This framework is very general and encompasses 37809 stochastic variants of line search, 37810 quasi-Newton, cubic regularized Newton and
37810SQP 37811 methods for unconstrained and constrained 37812 problems. Under reasonable conditions on the 37813 oracles, we show high probability bounds on the 37814 sample and iteration complexity of the 37815 algorithms. 37816 </blockquote> 37817 </div> 37818 </div> 37819 </div> 37820 </div> 37821 <div class="slot-urls"></div> 37822 <a href="/wsc23papers/doc109.pdf" target="_blank">pdf</a 37823 ><br /> 37824 </div> 37825 </div> 37826 <div class="session-entry"> 37827 <span class="session-event-type">PhD Colloquium</span 37828 ><span class="type-track-spacer"> · </span 37829 ><span class="program-track">PhD Colloquium</span><br /> 37830 <div class="session-title">PhD Colloquium Session B2</div> 37831 <div class="session-chair"> 37832 Chair: Enlu Zhou (Georgia Institute of Technology)<br /> 37833 </div> 37834 <div class="slot-entry"> 37835 <a name="doc119" tabindex="-1"></a> 37836 <div class="slot-title-line"> 37837 <span class="slot-title" 37838 >Sustainability-Integrated Digital Framework for 37839 Decision Making in Interior Construction Design</span 37840 > 37841 </div> 37842 <div class="slot-authors"> 37843 Rongxu Liu (University of Exeter) 37844 </div> 37845 <div class="slot-abstract"> 37846 <div> 37847 <a 37848 class="clickable no-decoration" 37849 id="vhsjs_view_878_1707793553_134667" 37850 onclick="$('#vhsjs_view_878_1707793553_134667').hide(); 37851 $('#vhsjs_hide_878_1707793553_134667').show(); 37852 $('#877_1707793553_1346533').slideDown(function() { 37853 if (typeof Masonry === 'function') { 37854 $('.use_masonry').masonry(); 37855 }; 37856 37857 });" 37858 ><i class="fa fa-caret-right"></i> 37859 <span class="hover_link">Abstract</span></a 37860 ><a 37861 class="clickable no-decoration" 37862 id="vhsjs_hide_878_1707793553_134667" 37863 onclick="$('#877_1707793553_1346533').hide(function() { 37864 if (typeof Masonry === 'function') { 37865 $('.use_masonry').masonry(); 37866 }; 37867 }); 37868 $('#vhsjs_hide_878_1707793553_134667').hide(); 37869 $('#vhsjs_view_878_1707793553_134667').show();" 37870 style="display: none" 37871 ><i class="fa fa-caret-down"></i> 37872 <span class="hover_link">Abstract</span></a 37873 > 37874 <div 37875 data-display-control="878_1707793553_134667" 37876 id="877_1707793553_1346533" 37877 style="display: none" 37878 > 37879 <div class="arrow-slidedown"> 37880 <blockquote> 37881 The present study presents a novel digital tool 37882 that is seamlessly integrated with cutting-edge 37883 Industry 4.0 technologies. The primary objective 37884 of this tool is to effectively cater to the 37885 diverse requirements of stakeholders involved in 37886 interior construction projects. This research 37887 endeavor explores the various challenges faced 37888 by stakeholders, examines the significance of 37889 digital tools in facilitating the integration of 37890 cutting-edge technologies, and assesses the 37891 effectiveness of the proposed application in 37892 improving project results. The anticipated 37893 outcomes hold the potential to fundamentally 37894 transform the landscape of construction project 37895 management in the future. This transformation 37896 will be achieved through the integration of 37897 stakeholder requirements and the utilization of 37898 cutting-edge technological advancements. 37899 </blockquote> 37900 </div> 37901 </div> 37902 </div> 37903 </div> 37904 <div class="slot-urls"></div> 37905 <a href="/wsc23papers/doc119.pdf" target="_blank">pdf</a 37906 ><br /> 37907 </div> 37908 <div class="slot-entry"> 37909 <a name="doc122" tabindex="-1"></a> 37910 <div class="slot-title-line">
37911 <span class="slot-title" 37912 >Dynamic Weapon Target Assignment via Simulation, 37913 Reinforcement Learning and Graph Neural Network</span 37914 > 37915 </div> 37916 <div class="slot-authors"> 37917 Seung Heon Oh (Seoul National University) 37918 </div> 37919 <div class="slot-abstract"> 37920 <div> 37921 <a 37922 class="clickable no-decoration" 37923 id="vhsjs_view_880_1707793553_1379755" 37924 onclick="$('#vhsjs_view_880_1707793553_1379755').hide(); 37925 $('#vhsjs_hide_880_1707793553_1379755').show(); 37926 $('#879_1707793553_137962').slideDown(function() { 37927 if (typeof Masonry === 'function') { 37928 $('.use_masonry').masonry(); 37929 }; 37930 37931 });" 37932 ><i class="fa fa-caret-right"></i> 37933 <span class="hover_link">Abstract</span></a 37934 ><a 37935 class="clickable no-decoration" 37936 id="vhsjs_hide_880_1707793553_1379755" 37937 onclick="$('#879_1707793553_137962').hide(function() { 37938 if (typeof Masonry === 'function') { 37939 $('.use_masonry').masonry(); 37940 }; 37941 }); 37942 $('#vhsjs_hide_880_1707793553_1379755').hide(); 37943 $('#vhsjs_view_880_1707793553_1379755').show();" 37944 style="display: none" 37945 ><i class="fa fa-caret-down"></i> 37946 <span class="hover_link">Abstract</span></a 37947 > 37948 <div 37949 data-display-control="880_1707793553_1379755" 37950 id="879_1707793553_137962" 37951 style="display: none" 37952 > 37953 <div class="arrow-slidedown"> 37954 <blockquote> 37955 DWTA (dynamic weapon target assignment problem) 37956 is the important resource scheduling problem in 37957 battlefield. In this paper, deep reinforcement 37958 learning and graph neural network optimize the 37959 performance of the decision making of DWTA. The 37960 proposed method is evaluated experimentally for 37961 some cases and compared with other heuristic 37962 methods. 37963 </blockquote> 37964 </div> 37965 </div> 37966 </div> 37967 </div> 37968 <div class="slot-urls"></div> 37969 <a href="/wsc23papers/doc122.pdf" target="_blank">pdf</a 37970 ><br /> 37971 </div> 37972 <div class="slot-entry"> 37973 <a name="doc126" tabindex="-1"></a> 37974 <div class="slot-title-line"> 37975 <span class="slot-title" 37976 >A Simulation Framework for Clearing Function-based 37977 Release Date Optimization in a Material Requirements 37978 Planned Planned Production System</span 37979 > 37980 </div> 37981 <div class="slot-authors"> 37982 Wolfgang Seiringer (University of Applied Science Upper 37983 Austria) 37984 </div> 37985 <div class="slot-abstract"> 37986 <div> 37987 <a 37988 class="clickable no-decoration" 37989 id="vhsjs_view_882_1707793553_1422124" 37990 onclick="$('#vhsjs_view_882_1707793553_1422124').hide(); 37991 $('#vhsjs_hide_882_1707793553_1422124').show(); 37992 $('#881_1707793553_1421983').slideDown(function() { 37993 if (typeof Masonry === 'function') { 37994 $('.use_masonry').masonry(); 37995 }; 37996 37997 });" 37998 ><i class="fa fa-caret-right"></i>
37999 <span class="hover_link">Abstract</span></a 38000 ><a 38001 class="clickable no-decoration" 38002 id="vhsjs_hide_882_1707793553_1422124" 38003 onclick="$('#881_1707793553_1421983').hide(function() { 38004 if (typeof Masonry === 'function') { 38005 $('.use_masonry').masonry(); 38006 }; 38007 }); 38008 $('#vhsjs_hide_882_1707793553_1422124').hide(); 38009 $('#vhsjs_view_882_1707793553_1422124').show();" 38010 style="display: none" 38011 ><i class="fa fa-caret-down"></i> 38012 <span class="hover_link">Abstract</span></a 38013 > 38014 <div 38015 data-display-control="882_1707793553_1422124" 38016 id="881_1707793553_1421983" 38017 style="display: none" 38018 > 38019 <div class="arrow-slidedown"> 38020 <blockquote> 38021 In this research work a simulation framework is 38022 developed helping to overcome the missing 38023 capacity limitation of material requirements 38024 planning (MRP) to obtain more reliable planning 38025 results. Therefore, the concept of clearing 38026 functions (CF) are integrated as constraints 38027 into a mathematical optimization problem. When 38028 using CF as capacity constraints it is possible 38029 to identify how much of the current workload is 38030 realistic to be processed on the shop floor of a 38031 production. The CF based release dates will 38032 replace the fixed planned lead time of MRP, 38033 which is unable to handle capacity limitations. 38034 To evaluate the performance of CF based release 38035 date planning a comparison with standard MRP 38036 using a simulation experiment is done. First 38037 results show the potential of the CF approach, 38038 but due to the complexity of the release 38039 mechanism adjustments to the planning and 38040 optimization component in the simulation are 38041 necessary. 38042 </blockquote> 38043 </div> 38044 </div> 38045 </div> 38046 </div> 38047 <div class="slot-urls"></div> 38048 <a href="/wsc23papers/doc126.pdf" target="_blank">pdf</a 38049 ><br /> 38050 </div> 38051 <div class="slot-entry"> 38052 <a name="doc127" tabindex="-1"></a> 38053 <div class="slot-title-line"> 38054 <span class="slot-title" 38055 >To What Extent Can Simulation Optimization be Used in 38056 Wildlife Reserve Design?</span 38057 > 38058 </div> 38059 <div class="slot-authors"> 38060 Shengjie Zhou (Lancaster University) 38061 </div> 38062 <div class="slot-abstract"> 38063 <div> 38064 <a 38065 class="clickable no-decoration" 38066 id="vhsjs_view_884_1707793553_1461406" 38067 onclick="$('#vhsjs_view_884_1707793553_1461406').hide(); 38068 $('#vhsjs_hide_884_1707793553_1461406').show(); 38069 $('#883_1707793553_1461267').slideDown(function() { 38070 if (typeof Masonry === 'function') { 38071 $('.use_masonry').masonry(); 38072 }; 38073 38074 });" 38075 ><i class="fa fa-caret-right"></i>
38076 <span class="hover_link">Abstract</span></a 38077 ><a 38078 class="clickable no-decoration" 38079 id="vhsjs_hide_884_1707793553_1461406" 38080 onclick="$('#883_1707793553_1461267').hide(function() { 38081 if (typeof Masonry === 'function') { 38082 $('.use_masonry').masonry(); 38083 }; 38084 }); 38085 $('#vhsjs_hide_884_1707793553_1461406').hide(); 38086 $('#vhsjs_view_884_1707793553_1461406').show();" 38087 style="display: none" 38088 ><i class="fa fa-caret-down"></i> 38089 <span class="hover_link">Abstract</span></a 38090 > 38091 <div 38092 data-display-control="884_1707793553_1461406" 38093 id="883_1707793553_1461267" 38094 style="display: none" 38095 > 38096 <div class="arrow-slidedown"> 38097 <blockquote> 38098 Wildlife reserves serve as a critical tool for 38099 conserving wildlife species. The design of such 38100 reserves can be formulated as a simulation 38101 optimization problem, with the objective of 38102 minimizing conservation costs while satisfying 38103 species survival constraints. Our research 38104 explores this problem formulation and the 38105 relevant solution methods, with a particular 38106 focus on the Chance Constrained Selection of the 38107 Best algorithm. We formulate the problem using a 38108 deterministic objective function subject to a 38109 probabilistic constraint. To estimate the 38110 survival probability under various policies, we 38111 have developed a Gray Wolf (Canis lupus) model 38112 that simulates the wolves’ dispersal, 38113 breeding, and death processes in discrete time 38114 steps. Our poster presents three scenarios that 38115 demonstrate the potential use of Simulation 38116 Optimization techniques in wildlife 38117 conservation. 38118 </blockquote> 38119 </div> 38120 </div> 38121 </div> 38122 </div> 38123 <div class="slot-urls"></div> 38124 <a href="/wsc23papers/doc127.pdf" target="_blank">pdf</a 38125 ><br /> 38126 </div> 38127 <div class="slot-entry"> 38128 <a name="doc128" tabindex="-1"></a> 38129 <div class="slot-title-line"> 38130 <span class="slot-title" 38131 >Real-time Delay Prediction for Kidney Transplantation 38132 System</span 38133 > 38134 </div> 38135 <div class="slot-authors"> 38136 Najiya Fatma (Indian Institute of Technology Delhi) 38137 </div> 38138 <div class="slot-abstract"> 38139 <div> 38140 <a 38141 class="clickable no-decoration" 38142 id="vhsjs_view_886_1707793553_1503417" 38143 onclick="$('#vhsjs_view_886_1707793553_1503417').hide(); 38144 $('#vhsjs_hide_886_1707793553_1503417').show(); 38145 $('#885_1707793553_1503272').slideDown(function() { 38146 if (typeof Masonry === 'function') { 38147 $('.use_masonry').masonry(); 38148 }; 38149 38150 });" 38151 ><i class="fa fa-caret-right"></i>
38152 <span class="hover_link">Abstract</span></a 38153 ><a 38154 class="clickable no-decoration" 38155 id="vhsjs_hide_886_1707793553_1503417" 38156 onclick="$('#885_1707793553_1503272').hide(function() { 38157 if (typeof Masonry === 'function') { 38158 $('.use_masonry').masonry(); 38159 }; 38160 }); 38161 $('#vhsjs_hide_886_1707793553_1503417').hide(); 38162 $('#vhsjs_view_886_1707793553_1503417').show();" 38163 style="display: none" 38164 ><i class="fa fa-caret-down"></i> 38165 <span class="hover_link">Abstract</span></a 38166 > 38167 <div 38168 data-display-control="886_1707793553_1503417" 38169 id="885_1707793553_1503272" 38170 style="display: none" 38171 > 38172 <div class="arrow-slidedown"> 38173 <blockquote> 38174 We present a combined simulation and machine 38175 learning framework for predicting, at the time 38176 of end-stage renal disease patient’s 38177 registration on the kidney transplantation 38178 waitlist, whether the patient will receive a 38179 transplant before their health deteriorates. If 38180 the patient is predicted to receive a 38181 transplant, we predict their time on the 38182 waitlist before receiving the transplant. We 38183 accomplish this by developing a discrete-event 38184 simulation model of the kidney transplantation 38185 system using patient-related and organ 38186 donor-related information. We use the validated 38187 model to record clinical and operational 38188 features for each patient at the time of their 38189 registration, which is then used to train 38190 machine learning algorithms to predict the 38191 transplantation waitlist outcome, and, in turn, 38192 the organ allocation time. Our approach is 38193 suitable for generating real-time delay 38194 predictions for complex queuing systems where 38195 data regarding state of the queueing system that 38196 can be used to train ML methods is not 38197 maintained. 38198 </blockquote> 38199 </div> 38200 </div> 38201 </div> 38202 </div> 38203 <div class="slot-urls"></div> 38204 <a href="/wsc23papers/doc128.pdf" target="_blank">pdf</a 38205 ><br /> 38206 </div> 38207 <div class="slot-entry"> 38208 <a name="doc131" tabindex="-1"></a> 38209 <div class="slot-title-line"> 38210 <span class="slot-title" 38211 >Epydemia: an Open-source Agent-based Model for 38212 Infectious Disease Modeling</span 38213 > 38214 </div> 38215 <div class="slot-authors"> 38216 Sebastian Rodriguez Cartes (North Carolina State 38217 University) 38218 </div> 38219 <div class="slot-abstract"> 38220 <div> 38221 <a 38222 class="clickable no-decoration" 38223 id="vhsjs_view_888_1707793553_1543803" 38224 onclick="$('#vhsjs_view_888_1707793553_1543803').hide(); 38225 $('#vhsjs_hide_888_1707793553_1543803').show(); 38226 $('#887_1707793553_1543665').slideDown(function() { 38227 if (typeof Masonry === 'function') { 38228 $('.use_masonry').masonry(); 38229 }; 38230 38231 });" 38232 ><i class="fa fa-caret-right"></i>
38233 <span class="hover_link">Abstract</span></a 38234 ><a 38235 class="clickable no-decoration" 38236 id="vhsjs_hide_888_1707793553_1543803" 38237 onclick="$('#887_1707793553_1543665').hide(function() { 38238 if (typeof Masonry === 'function') { 38239 $('.use_masonry').masonry(); 38240 }; 38241 }); 38242 $('#vhsjs_hide_888_1707793553_1543803').hide(); 38243 $('#vhsjs_view_888_1707793553_1543803').show();" 38244 style="display: none" 38245 ><i class="fa fa-caret-down"></i> 38246 <span class="hover_link">Abstract</span></a 38247 > 38248 <div 38249 data-display-control="888_1707793553_1543803" 38250 id="887_1707793553_1543665" 38251 style="display: none" 38252 > 38253 <div class="arrow-slidedown"> 38254 <blockquote> 38255 Agent-based models provide a flexible framework 38256 for the modeling of infectious diseases. We 38257 propose an open-source simulation framework, 38258 EPyDEMIA, that allows modeling multiple diseases 38259 infecting a population, implementing complex 38260 agent behaviors, and different interventions. 38261 The framework was designed as a discrete-event 38262 simulator and was implemented using Python. 38263 Infections throughout a population are driven 38264 using a network of multiple independent layers. 38265 We highlight the utility of our framework by 38266 showcasing a two-disease outbreak example. The 38267 proposed tool's modularity facilitates the 38268 implementation of disease transmission models, 38269 streamlining the analysis of the health impacts 38270 of infections. 38271 </blockquote> 38272 </div> 38273 </div> 38274 </div> 38275 </div> 38276 <div class="slot-urls"></div> 38277 <a href="/wsc23papers/doc131.pdf" target="_blank">pdf</a 38278 ><br /> 38279 </div> 38280 <div class="slot-entry"> 38281 <a name="doc134" tabindex="-1"></a> 38282 <div class="slot-title-line"> 38283 <span class="slot-title" 38284 >Developing a Bi-Level and Interoperable Framework for 38285 Digital Twins: An Application For The Underground 38286 Mining Industry</span 38287 > 38288 </div> 38289 <div class="slot-authors"> 38290 Mostafa DadkhahKalateh (Polytechnique Montréal) 38291 </div> 38292 <div class="slot-abstract"> 38293 <div> 38294 <a 38295 class="clickable no-decoration" 38296 id="vhsjs_view_890_1707793553_1579275" 38297 onclick="$('#vhsjs_view_890_1707793553_1579275').hide(); 38298 $('#vhsjs_hide_890_1707793553_1579275').show(); 38299 $('#889_1707793553_1579137').slideDown(function() { 38300 if (typeof Masonry === 'function') { 38301 $('.use_masonry').masonry(); 38302 }; 38303 38304 });" 38305 ><i class="fa fa-caret-right"></i> 38306 <span class="hover_link">Abstract</span></a 38307 ><a 38308 class="clickable no-decoration" 38309 id="vhsjs_hide_890_1707793553_1579275" 38310 onclick="$('#889_1707793553_1579137').hide(function() { 38311 if (typeof Masonry === 'function') { 38312 $('.use_masonry').masonry(); 38313 }; 38314 }); 38315 $('#vhsjs_hide_890_1707793553_1579275').hide(); 38316 $('#vhsjs_view_890_1707793553_1579275').show();" 38317 style="display: none" 38318 ><i class="fa fa-caret-down"></i>
38319 <span class="hover_link">Abstract</span></a 38320 > 38321 <div 38322 data-display-control="890_1707793553_1579275" 38323 id="889_1707793553_1579137" 38324 style="display: none" 38325 > 38326 <div class="arrow-slidedown"> 38327 <blockquote> 38328 The study presents an innovative modular, 38329 technical, and bi-level Digital Twin 38330 architecture, specifically designed for 38331 underground mining systems. Aligned with 38332 Industry 4.0 principles, it aspires to integrate 38333 and enhance mining activities across the mining 38334 value chain. Spanning its entire value chain, 38335 the architecture considers lifecycle phases, 38336 physical assets and operations in six functional 38337 layers, addressing interoperability between the 38338 IoT, data, and various models. This holistic 38339 design facilitates remote control of underground 38340 operations and provides flexibility to craft 38341 decision tools tailored to individual 38342 configurations. The focus is on merging 38343 real-time data with decision tools to achieve a 38344 granular system portrayal and facilitate 38345 informed operational decisions. The architecture 38346 adopts a service-oriented approach, 38347 necessitating the partitioning of data and 38348 decision models, ensuring a flexible, extensible 38349 lower-level Fleet Management System using UML 38350 methodologies. Ultimately, this architecture is 38351 poised to revolutionize mining processes and 38352 resiliency, driving operational efficiency, 38353 safety and adaptability to new heights. 38354 </blockquote> 38355 </div> 38356 </div> 38357 </div> 38358 </div> 38359 <div class="slot-urls"></div> 38360 <a href="/wsc23papers/doc134.pdf" target="_blank">pdf</a 38361 ><br /> 38362 </div> 38363 <div class="slot-entry"> 38364 <a name="doc135" tabindex="-1"></a> 38365 <div class="slot-title-line"> 38366 <span class="slot-title" 38367 >Towards a Hybrid Discrete Event Simulation 38368 Agent-based Model for the Texas State Mental Hospital 38369 System</span 38370 > 38371 </div> 38372 <div class="slot-authors"> 38373 Maria Tomasso (Texas State University) 38374 </div> 38375 <div class="slot-abstract"> 38376 <div> 38377 <a 38378 class="clickable no-decoration" 38379 id="vhsjs_view_892_1707793553_1621745" 38380 onclick="$('#vhsjs_view_892_1707793553_1621745').hide(); 38381 $('#vhsjs_hide_892_1707793553_1621745').show(); 38382 $('#891_1707793553_162147').slideDown(function() { 38383 if (typeof Masonry === 'function') { 38384 $('.use_masonry').masonry(); 38385 }; 38386 38387 });" 38388 ><i class="fa fa-caret-right"></i> 38389 <span class="hover_link">Abstract</span></a 38390 ><a 38391 class="clickable no-decoration" 38392 id="vhsjs_hide_892_1707793553_1621745" 38393 onclick="$('#891_1707793553_162147').hide(function() { 38394 if (typeof Masonry === 'function') { 38395 $('.use_masonry').masonry(); 38396 }; 38397 }); 38398 $('#vhsjs_hide_892_1707793553_1621745').hide(); 38399 $('#vhsjs_view_892_1707793553_1621745').show();" 38400 style="display: none" 38401 ><i class="fa fa-caret-down"></i>
38402 <span class="hover_link">Abstract</span></a 38403 > 38404 <div 38405 data-display-control="892_1707793553_1621745" 38406 id="891_1707793553_162147" 38407 style="display: none" 38408 > 38409 <div class="arrow-slidedown"> 38410 <blockquote> 38411 State mental health hospitals provide a vital 38412 service to individuals who pose a threat to 38413 themselves or others. However, in recent years, 38414 these facilities have struggled to meet demand, 38415 resulting in a waitlist of over one thousand 38416 patients. Despite legislative efforts to address 38417 this issue, waitlist lengths persist and 38418 continue to grow. This study employs a hybrid 38419 discrete event simulation agent-based model 38420 (DES-ABM), trained on publicly available 38421 aggregate data, to model waitlists for state 38422 mental health hospitals in Texas. Once trained, 38423 the model enables projections of the impact of 38424 various policy interventions and resource 38425 allocation strategies on the waitlist. The model 38426 successfully approximated waitlist lengths from 38427 2020-2022, and we tested two interventions 38428 involving the expansion of available beds, 38429 recording their effects on the waitlists. 38430 </blockquote> 38431 </div> 38432 </div> 38433 </div> 38434 </div> 38435 <div class="slot-urls"></div> 38436 <a href="/wsc23papers/doc135.pdf" target="_blank">pdf</a 38437 ><br /> 38438 </div> 38439 <div class="slot-entry"> 38440 <a name="doc140" tabindex="-1"></a> 38441 <div class="slot-title-line"> 38442 <span class="slot-title" 38443 >Significance of Traffic Loading for Evacuation and 38444 Percolation-based Control Strategies</span 38445 > 38446 </div> 38447 <div class="slot-authors"> 38448 Ruqing Huang (The University of Tennessee, Knoxville) 38449 </div> 38450 <div class="slot-abstract"> 38451 <div> 38452 <a 38453 class="clickable no-decoration" 38454 id="vhsjs_view_894_1707793553_1662385" 38455 onclick="$('#vhsjs_view_894_1707793553_1662385').hide(); 38456 $('#vhsjs_hide_894_1707793553_1662385').show(); 38457 $('#893_1707793553_1662245').slideDown(function() { 38458 if (typeof Masonry === 'function') { 38459 $('.use_masonry').masonry(); 38460 }; 38461 38462 });" 38463 ><i class="fa fa-caret-right"></i> 38464 <span class="hover_link">Abstract</span></a 38465 ><a 38466 class="clickable no-decoration" 38467 id="vhsjs_hide_894_1707793553_1662385" 38468 onclick="$('#893_1707793553_1662245').hide(function() { 38469 if (typeof Masonry === 'function') { 38470 $('.use_masonry').masonry(); 38471 }; 38472 }); 38473 $('#vhsjs_hide_894_1707793553_1662385').hide(); 38474 $('#vhsjs_view_894_1707793553_1662385').show();" 38475 style="display: none" 38476 ><i class="fa fa-caret-down"></i>
38477 <span class="hover_link">Abstract</span></a 38478 > 38479 <div 38480 data-display-control="894_1707793553_1662385" 38481 id="893_1707793553_1662245" 38482 style="display: none" 38483 > 38484 <div class="arrow-slidedown"> 38485 <blockquote> 38486 This paper investigates the significance of 38487 traffic loading rate for evacuation efficiency 38488 through large-scale evacuation simulation on a 38489 20*20 grid network, emphasizing the emergency 38490 evacuation of the central 10*10 CBD area. There 38491 exists an equilibrium between the loading flow 38492 into the CBD and the exiting flow out of the 38493 CBD, which simultaneously optimizes evacuation 38494 efficiency. Loading can be excessive, over, 38495 equilibrium, or under-loaded, with overloading 38496 causing widespread jams and potential gridlocks. 38497 Using percolation theory, we also proposed 38498 several strategies that limit congestion spread 38499 to the CBD's edge, achieving equilibrium with 38500 optimal evacuee exit rates. 38501 </blockquote> 38502 </div> 38503 </div> 38504 </div> 38505 </div> 38506 <div class="slot-urls"></div> 38507 <a href="/wsc23papers/doc140.pdf" target="_blank">pdf</a 38508 ><br /> 38509 </div> 38510 <div class="slot-entry"> 38511 <a name="doc141" tabindex="-1"></a> 38512 <div class="slot-title-line"> 38513 <span class="slot-title" 38514 >Assessing the Impact of Social Network Settings on 38515 COVID-19 Transmission in Cruise Ships: An Agent-Based 38516 Modeling Approach</span 38517 > 38518 </div> 38519 <div class="slot-authors"> 38520 Akane Fujimoto Wakabayashi (Georgia Institute of 38521 Technology) 38522 </div> 38523 <div class="slot-abstract"> 38524 <div> 38525 <a 38526 class="clickable no-decoration" 38527 id="vhsjs_view_896_1707793553_170498" 38528 onclick="$('#vhsjs_view_896_1707793553_170498').hide(); 38529 $('#vhsjs_hide_896_1707793553_170498').show(); 38530 $('#895_1707793553_1704838').slideDown(function() { 38531 if (typeof Masonry === 'function') { 38532 $('.use_masonry').masonry(); 38533 }; 38534 38535 });" 38536 ><i class="fa fa-caret-right"></i> 38537 <span class="hover_link">Abstract</span></a 38538 ><a 38539 class="clickable no-decoration" 38540 id="vhsjs_hide_896_1707793553_170498" 38541 onclick="$('#895_1707793553_1704838').hide(function() { 38542 if (typeof Masonry === 'function') { 38543 $('.use_masonry').masonry(); 38544 }; 38545 }); 38546 $('#vhsjs_hide_896_1707793553_170498').hide(); 38547 $('#vhsjs_view_896_1707793553_170498').show();" 38548 style="display: none" 38549 ><i class="fa fa-caret-down"></i> 38550 <span class="hover_link">Abstract</span></a 38551 > 38552 <div
38553 data-display-control="896_1707793553_170498" 38554 id="895_1707793553_1704838" 38555 style="display: none" 38556 > 38557 <div class="arrow-slidedown"> 38558 <blockquote> 38559 Cruise ship operations faced significant 38560 disruptions during the COVID-19 pandemic. Close 38561 quarters and dense populations of domestic and 38562 international travelers are an environment where 38563 viruses can spread easily. The cruise industry 38564 and public health partners continue to develop 38565 guidelines to control the spread of disease 38566 within these settings. In this study, we 38567 developed an agent-based model to simulate the 38568 spread of COVID-19 in cruise ship environments. 38569 The model considers various types of 38570 interactions, including passenger-passenger, 38571 passenger-crew, and crew-crew interactions 38572 within networks and the cruise ship population. 38573 We evaluated the impact of different social 38574 network settings, such as group travel sizes, 38575 intensity of interactions, and initial number of 38576 infection seeds on the spread of disease. The 38577 findings provide insights for public health 38578 decision-makers and the modeling framework can 38579 inform other modeling activities that rely on 38580 similar data streams. 38581 </blockquote> 38582 </div> 38583 </div> 38584 </div> 38585 </div> 38586 <div class="slot-urls"></div> 38587 <a href="/wsc23papers/doc141.pdf" target="_blank">pdf</a 38588 ><br /> 38589 </div> 38590 </div> 38591 </div> 38592 <div class="centered"> 38593 <div class="top-link"><a href="#top">Return to Top</a></div> 38594 </div> 38595 <hr /> 38596 </div> 38597 <div class="area-section"> 38598 <div class="centered"> 38599 <a name="other" tabindex="-1"></a> 38600 <div class="section-title">Other</div> 38601 </div> 38602 <div class="section-entry"> 38603 <div class="session-entry"> 38604 <span class="session-event-type">Plenary</span><br /> 38605 <div class="session-title">In Memoriam</div> 38606 <div class="session-chair"> 38607 Chair: James Wilson (North Carolina State University)<br /> 38608 </div> 38609 <div class="slot-entry"> 38610 <a name="prog102" tabindex="-1"></a> 38611 <div class="slot-title-line"> 38612 <span class="slot-title" 38613 >In Memoriam: Peter D. Welch (1928‒2023)</span 38614 > 38615 </div> 38616 <div class="slot-authors"> 38617 James Wilson (North Carolina State University) 38618 </div> 38619 <div class="slot-abstract"></div> 38620 <div class="slot-urls"></div> 38621 <a href="/wsc23papers/prog102.pdf" target="_blank">pdf</a 38622 ><br /> 38623 </div> 38624 </div> 38625 </div> 38626 <div class="centered"> 38627 <div class="top-link"><a href="#top">Return to Top</a></div> 38628 </div> 38629 <hr /> 38630 </div> 38631 </td> 38632 </tr> 38633 </table> 38634 </div> 38635 </div> 38636 <div class="created-date righted">Created 2024-2-12 21:5</div> 38637 </body> 38638</html>
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.