-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathtutorial.html
More file actions
357 lines (251 loc) · 15.8 KB
/
Copy pathtutorial.html
File metadata and controls
357 lines (251 loc) · 15.8 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
<!DOCTYPE html>
<!--[if lt IE 9 ]><html class="no-js oldie" lang="en"> <![endif]-->
<!--[if IE 9 ]><html class="no-js oldie ie9" lang="en"> <![endif]-->
<!--[if (gte IE 9)|!(IE)]><!-->
<html class="no-js" lang="en">
<!--<![endif]-->
<head>
<!--- basic page needs
================================================== -->
<meta charset="utf-8">
<title>Bridging the Sim-to-Real Gap in Traffic Engineering — Tutorial @ ITSC 2026</title>
<meta name="description" content="Bridging the Sim-to-Real Gap in Traffic Engineering: Integrating Traffic Signal Control and Reinforcement Learning. Half-day tutorial at IEEE ITSC 2026.">
<meta name="author" content="">
<!-- mobile specific metas
================================================== -->
<meta name="viewport" content="width=device-width, initial-scale=1">
<!-- CSS
================================================== -->
<link rel="stylesheet" href="css/base.css">
<link rel="stylesheet" href="css/vendor.css">
<link rel="stylesheet" href="css/main.css">
<link rel="stylesheet" href="css/site.css">
<!-- script
================================================== -->
<script src="js/modernizr.js"></script>
<script src="js/pace.min.js"></script>
<!-- favicons
================================================== -->
<link rel="shortcut icon" href="favicon.ico" type="image/x-icon">
<link rel="icon" href="favicon.ico" type="image/x-icon">
</head>
<body id="top">
<!-- header
================================================== -->
<header id="header" class="row">
<div class="header-logo">
<a href="tutorial.html">ITSC 2026 Tutorial</a>
</div>
<nav id="header-nav-wrap">
<ul class="header-main-nav">
<li class="current"><a class="smoothscroll" href="#home" title="home">Home</a></li>
<li><a class="smoothscroll" href="#about" title="about">Aim</a></li>
<li><a class="smoothscroll" href="#agenda" title="agenda">Agenda</a></li>
<li><a class="smoothscroll" href="#organizers" title="organizers">Organizers</a></li>
<li><a href="index.html" title="sim2signal">Sim2Signal</a></li>
</ul>
<a href="https://github.com/Red-Pheonix/Sim2RealTSCBenchMark" title="github" class="button button-primary cta" target="_blank">GitHub</a>
</nav>
<a class="header-menu-toggle" href="#"><span>Menu</span></a>
</header> <!-- /header -->
<!-- home
================================================== -->
<section id="home" data-parallax="scroll" data-image-src="images/car-trail.png" data-natural-width=3000 data-natural-height=2000>
<div class="overlay"></div>
<div class="home-content">
<div class="row contents">
<div class="home-content-left wide">
<h3 data-aos="fade-up">Friday, 18 September 2026 · 14:00–16:00 <br> Naples, Italy</h3>
<h1 data-aos="fade-up">
Bridging the Sim-to-Real Gap in Traffic Engineering
</h1>
<h3 data-aos="fade-up">
29th IEEE International Conference on Intelligent Transportation Systems (ITSC 2026)
</h3>
<div class="buttons" data-aos="fade-up">
<a href="#about" class="smoothscroll button stroke">
<span class="icon-circle-down" aria-hidden="true"></span>
Learn More
</a>
<a href="#agenda" class="smoothscroll button stroke">
<span class="icon-file" aria-hidden="true"></span>
Agenda
</a>
</div>
</div>
</div>
</div> <!-- end home-content -->
<div class="home-scrolldown">
<a href="#about" class="scroll-icon smoothscroll">
<span>Scroll Down</span>
<i class="icon-arrow-right" aria-hidden="true"></i>
</a>
</div>
</section> <!-- end home -->
<!-- about
================================================== -->
<section id="about">
<div class="row about-intro">
<div class="col-four">
<h1 class="intro-header" data-aos="fade-up">Aim of the Tutorial</h1>
</div>
<div class="col-eight">
<p class="lead" data-aos="fade-up">
Reinforcement learning has produced traffic signal controllers that perform very well in simulation. Almost none of them have been deployed on real signal hardware. The reason is the Sim-to-Real gap: the simulator a policy learns in differs from the street it is meant to run on, and the policy does not survive the move.
</p>
<p class="lead" data-aos="fade-up">
This half-day tutorial connects two communities that have been working on the same problem under different names. Traffic engineering has fought this gap for decades under the name <em>calibration</em>. Reinforcement learning attacks it with grounding, domain randomization, domain adaptation, and reward shaping. We show that both are addressing the same four things — mismatches in observation, action, transition, and reward — and that the two halves fit together.
</p>
<p class="lead" data-aos="fade-up">
Attendees will leave knowing how a real network is calibrated and exported into a simulator, which RL transfer methods attach to which part of the gap, and how to tell a good transfer from a bad one by reading the evaluation. Two live demos anchor the material: a GMNS network editor for the calibration half, and a phase-time visualizer for the evaluation half.
</p>
<p class="lead" data-aos="fade-up">
The four-gap decomposition used throughout the tutorial comes from <a href="index.html">Sim2Signal</a>, our benchmark for measuring the Sim-to-Real gap in traffic signal control.
</p>
</div>
</div>
</section>
<!-- agenda
================================================== -->
<section id="agenda">
<div class="row about-how">
<h1 class="intro-header" data-aos="fade-up">Agenda and Materials</h1>
<div class="about-how-content" data-aos="fade-up">
<div class="about-how-steps block-1-2 block-tab-full group">
<div class="bgrid step" data-item="10 min">
<h3>Welcome and Opening</h3>
<p>What the Sim-to-Real gap is, and why it matters for traffic engineering.</p>
<p>The (S, A, T, R) decomposition: observation, action, transition, and reward gaps mapped onto the MDP tuple — the spine of the rest of the tutorial.</p>
<p><em>Hua Wei</em></p>
</div>
<div class="bgrid step" data-item="10 min">
<h3>The Sim-to-Real Gap in Traffic Engineering</h3>
<p>The gap predates RL: traffic engineering has fought it for decades under the name calibration.</p>
<p>Where the gap comes from in practice, and the state of real deployment — the surveyed RL-for-signal-control literature is simulation-only.</p>
<p><em>Hua Wei</em></p>
</div>
<div class="bgrid step" data-item="30 min">
<h3>Calibration for Traffic Signal Control</h3>
<p>Signal terminology for an RL audience: movements, phases and phase transitions, min/max green and clearance, pretimed vs. actuated control, NEMA dual-ring.</p>
<p>Data formats and simulators: GMNS as the hub, with UTDF, SUMO and CityFlow as spokes; the UTDF → GMNS → SUMO/CityFlow conversion pipeline and where it breaks.</p>
<p><strong>Live demo:</strong> the GMNS Editor — auto-join, merging split intersections, movements, NEMA timing, validate and export.</p>
<p><em>Hua Wei</em></p>
</div>
<div class="bgrid step" data-item="15 min">
<h3>Coffee Break</h3>
</div>
<div class="bgrid step" data-item="45 min">
<h3>RL Methodology and Evaluation</h3>
<p>The four families of RL solutions — grounding actions, domain randomization, domain adaptation, reward shaping — and where each attaches to the taxonomy.</p>
<p>A worked sim-to-sim example: train in CityFlow, transfer to SUMO, induce gaps and measure them.</p>
<p>The multi-agent extension: why multi-intersection control does not reduce to the single-intersection case, and the cascading invalidation effect.</p>
<p><strong>Live demo:</strong> the phase-time visualizer — reading space-time and phase-time diagrams to tell a good transfer from a bad one.</p>
<p><em>Ferdous Al Rafi</em></p>
</div>
<div class="bgrid step" data-item="10 min">
<h3>Closing and Open Problems</h3>
<p>Recap: four gaps, one MDP, calibration and RL as two sides of the same problem.</p>
<p>Open problems: grounding the interface rather than only the dynamics, delay- and constraint-aware MDPs, and the empty deployment cell.</p>
<p>Relevant ITSS Technical Committee: Decision and Control in Transportation Systems. Discussion and Q&A.</p>
</div>
</div>
</div> <!-- end about-how-content -->
<!-- TODO: add slide / video / code links per block after the tutorial is delivered,
following the ITSC 2023 page format:
<p><a href="..." target="_blank">Slides</a>, <a href="..." target="_blank">Video</a></p> -->
</div> <!-- end about-how -->
</section> <!-- end agenda -->
<!-- organizers
================================================== -->
<section id="organizers-section">
<div class="row about-how" id="organizers">
<h1 class="intro-header" data-aos="fade-up">Organizers</h1>
</div>
<div class="row about-features">
<div class="features-list block-1-2 block-m-1-2 block-mob-full group">
<div class="bgrid feature" data-aos="fade-up">
<div class="service-content">
<h3><a href="https://www.public.asu.edu/~hwei27/index.html" target="_blank">Hua Wei</a></h3>
<p>
Assistant Professor, School of Computing and Augmented Intelligence, Arizona State University. His research bridges reinforcement learning and transportation, with an emphasis on closing the sim-to-real gap for traffic signal control and multi-agent traffic systems. He has organized community tutorials on traffic-control RL at ITSC 2020 and ITSC 2023, and delivered a keynote at an ITSC 2024 workshop.
</p>
</div>
</div> <!-- /bgrid -->
<div class="bgrid feature" data-aos="fade-up">
<div class="service-content">
<h3>Xuesong Zhou</h3>
<p>
Professor of Transportation Systems, School of Sustainable Engineering and the Built Environment, Arizona State University, and Director of the ASU Transportation+AI Lab. He is the principal architect of several open-source packages including DTALite, NEXTA and OSM2GMNS, and serves as Executive Editor-in-Chief of Urban Rail Transit.
</p>
</div>
</div> <!-- /bgrid -->
<div class="bgrid feature" data-aos="fade-up">
<div class="service-content">
<h3>Longchao Da</h3>
<p>
Ph.D. student in Computer Science at Arizona State University. His research focuses on mitigating the sim-to-real gap in reinforcement learning through uncertainty-aware learning, grounding and transfer mechanisms, and more realistic simulation environments. He has organized hands-on tutorials at KDD 2025, ITSC 2023 and ICDM 2025.
</p>
</div>
</div> <!-- /bgrid -->
<div class="bgrid feature" data-aos="fade-up">
<div class="service-content">
<h3>Ferdous Al Rafi</h3>
<p>
Ph.D. student in Computer Science at Arizona State University, with a background in Civil Engineering (Transportation) from the Bangladesh University of Engineering and Technology. His research interests include multi-agent reinforcement learning, traffic signal control, and bridging the sim-to-real gap in transportation systems.
</p>
</div>
</div> <!-- /bgrid -->
</div> <!-- end features-list -->
</div> <!-- end about-features -->
</section> <!-- end organizers -->
<!-- footer
================================================== -->
<footer>
<div class="footer-main">
<div class="row">
<div class="col-four md-1-2 tab-full footer-info">
<div class="footer-logo">ITSC 2026 Tutorial</div>
<p>
This project is developed and maintained by DaRL Lab from <a href="https://scai.engineering.asu.edu/" target="_blank">the School of Computing and Augmented Intelligence</a>,
<a href="https://www.asu.edu/" target="_blank">Arizona State University (ASU).</a>
</p>
</div> <!-- end footer-info -->
<div class="col-four md-1-4 tab-full mob-full footer-contact">
<h4>Contact</h4>
<p>
Hua Wei — hua.wei@asu.edu<br>
Xuesong Zhou — xzhou74@asu.edu<br>
Longchao Da — longchao@asu.edu<br>
Ferdous Al Rafi — frafi1@asu.edu<br>
</p>
</div> <!-- end footer-contact -->
<div class="col-four md-1-2 tab-full footer-subscribe">
<h4>Call for Participation</h4>
<p>Slides, code and recordings will be posted on this page after the tutorial. If you would like to contribute material or have questions about the session, please get in touch.</p>
</div> <!-- end footer-subscribe -->
</div> <!-- /row -->
</div> <!-- end footer-main -->
<div class="footer-bottom">
<div class="row">
<div class="col-twelve">
<div class="copyright">
<span>© Copyright Sim2Signal 2026.</span>
<span>Design by <a href="http://www.styleshout.com/">styleshout</a></span>
</div>
<div id="go-top">
<a class="smoothscroll" title="Back to Top" href="#top"><i class="icon-arrow-up"></i></a>
</div>
</div>
</div> <!-- end footer-bottom -->
</div>
</footer>
<div id="preloader">
<div id="loader"></div>
</div>
<!-- Java Script
================================================== -->
<script src="js/jquery-2.1.3.min.js"></script>
<script src="js/plugins.js"></script>
<script src="js/main.js"></script>
</body>
</html>