Conference Sessions

All (123)
Keynote Ses­sion (8)
Reg­u­lar Ses­sion (63)
Light­ning Ses­sion (43)
Mon­day, June 21 (32)
Tues­day, June 22 (31)
Wednes­day, June 23 (29)
Thurs­day, June 24 (31)
Behav­ior (9)
Behav­ior & Demand (8)
Con­nect­ed & Auto­mat­ed Vehi­cles (4)
Data (16)
Data-Informed Deci­sion Mak­ing (4)
Elec­tri­fi­ca­tion (4)
Emerg­ing Mobil­i­ty (13)
Freight (4)
Impli­ca­tion of Auto­mat­ed Vehi­cles (8)
Mod­el­ing, Sim­u­la­tion & Opti­miza­tion (10)
Shared Mobil­i­ty (9)
Traf­fic Con­trol & Man­ag­ment (4)
Traf­fic Oper­a­tions (9)
Trans­porta­tion Net­work Mod­el­ing (4)
PRESENTERS (109)
Abdul­la, Bahrul­la (1)
Ahamed, Tan­vir (1)
Ali, Rafaqat (1)
Alshu­rafa, Ahmed (1)
Ansari, Reza (1)
Axhausen, Kay (1)
Azeve­­do-Sa, Hebert (1)
Bal­ac, Milos (1)
Bal­lare, Sud­heer (1)
Bayen, Alexan­dre (1)
Bhat, Chan­dra (1)
Bramich, Daniel (1)
Bur­sa, Bar­tosz (1)
Cai, Xiaolin (1)
Calderon, Fran­cis­co (1)
Caros, Nicholas (1)
Chang, Yohan (1)
Chen, Rong­sheng (1)
Chen, Xiang­dong (1)
Daus, Matthew (1)
Dean, Matthew (1)
Dong, Jiqian (1)
Du, Lili (1)
Dug­gal, Mausam (1)
Eftekhar, Zahra (1)
Fakhrmoosavi, Fate­meh (1)
Fer­nan­do, Cel­so (1)
Fil­ipovs­ka, Moni­ka (2)
Flan­na­gan, Car­ol (1)
Fourati, Walid (1)
Geroli­m­in­is, Niko­las (1)
Gong, Feng­min (1)
Gong, Yun­hai (1)
Gopalakr­ish­nan, Ragaven­dran (1)
Guan, Xiangyang (2)
Guo, Hao (1)
Guo, Xiao­tong (1)
Guo, Yi (1)
Hale, David (1)
Hu, Zijian (1)
Jayara­man, Suresh Kumaar (1)
Kaneko, Noriko (1)
Kavia­n­ipour, Moham­madreza (1)
Kawasa­ki, Yosuke (1)
Ke, Jin­tao (2)
Kleiber, Mar­cel (1)
Kon­tou, Eleft­he­ria (1)
Koushik, Gun­takan­ti Sai (1)
Ladi­no, Andres (1)
Lee, Tony (Yoon-Dong) (1)
Levin, Michael (1)
Li, Ang (1)
Li, Can (1)
Li, Qian­wen (1)
Li, Xiaopeng (1)
Liu, Hen­ry (1)
Liu, Xiao­hui (1)
Liu, Zhao­cai (1)
Lorente, Ester (1)
Lou, Yingyan (2)
Luo, Zhix­iong (1)
Ma, Jiaqi (1)
Ma, Mingy­ou (1)
Mah­mas­sani, Hani (1)
Mar­tinez, Irene (1)
Miah, Md Mintu (1)
Miller, Eric (1)
Mintsis, Evan­ge­los (1)
Mirali­naghi, Moham­mad (1)
Moham­ma­di­an, Abol­fa­zl (Kouros) (1)
Mol­nar, Tamas (1)
Nakan­ishi, Wataru (1)
Nam, Daisik (1)
Okuhara, Rui (1)
Rahi­mi, Ehsan (1)
Ros-Roca, Xavier (1)
Sal­lard, Aurore (1)
Sayed, Md Abu (1)
Seo, Toru (1)
Shen, Hui (2)
Song, Zhanguo (1)
Su, Qida (1)
Tafreshi­an, Amirmah­di (1)
Tang, Xin­di (1)
Tay, Tim­o­thy (1)
Tian, Qiong (1)
Tsub­o­ta, Takahi­ro (1)
Ume­da, Shogo (1)
Vacek, Lukas (1)
Wang, Jingx­ing (1)
Wang, Mengx­in (1)
Wang, Shen­hao (1)
Wang, Yineng (1)
Wang, Yiyang (1)
Wei, Bangyang (1)
Xie, Tingt­ing (1)
Xu, Min (1)
Xu, Zhengt­ian (1)
Yan, Huimin (1)
Yang, Chen (1)
Yang, Di (1)
Yang, Hai (1)
Yang, Jie (1)
Zhang, Guo­qing (1)
Zhang, Ke (1)
Zhang, Kenan (1)
Zhang, Wen­wen (1)
Zheng, Zhengfei (1)
Zock­aie, Ali (1)

Keynote Session 1 — Hai Yang 

Title: Smart Mobil­i­ty Man­age­ment in the Era of Smart Trans­porta­tion
Speak­er: Hai Yang
Abstract: The cur­rent rev­o­lu­tions of shar­ing, automa­tion and elec­tri­fi­ca­tion are reshap­ing the way we trav­el, with broad impli­ca­tions for future mobil­i­ty man­age­ment. While much uncer­tain­ty remains about how these dis­rup­tive tech­nolo­gies would exact­ly impact demand for future mobil­i­ty and enhance­ment of trans­porta­tion sup­ply, it is clear that Inno­v­a­tive demand man­age­ment is equal­ly impor­tant as smart sup­ply tech­nol­o­gy devel­op­ment in solv­ing wors­en­ing traf­fic prob­lems in big cities. In this talk, I will dis­cuss the oppor­tu­ni­ties and chal­lenges of smart mobil­i­ty man­age­ment in the era of smart trans­porta­tion. Inno­v­a­tive ways of trav­el demand man­age­ment are described, includ­ing trad­able trav­el cred­it scheme for road con­ges­tion mit­i­ga­tion, rev­enue-pre­serv­ing and Pare­to-improv­ing strate­gies for peak-hour tran­sit demand man­age­ment con­ges­tion, and a nov­el reward scheme inte­grat­ed with surge pric­ing in a ride-sourc­ing market.

Keynote Session 2 — Hani Mahmassani 

Title: Oper­a­tional Strate­gies for Urban Air Mobil­i­ty and 4D Sys­tem Fun­da­men­tal Dia­grams
Speak­er: Hani Mah­mas­sani
Abstract: We take urban mobil­i­ty to the next lev­el by con­sid­er­ing shared mobil­i­ty ser­vices offered through auto­mat­ed elec­tric ver­ti­cal take-off and land­ing (eVTOL) vehi­cles (“fly­ing taxis”), enabled by new gen­er­a­tion of eVTOL air­craft. We present var­i­ous con­cepts for ser­vice oper­a­tions at urban/regional lev­els, along with algo­rithms adapt­ed for the real-time oper­a­tion of shared air mobil­i­ty fleets. We also exam­ine the con­gesta­bil­i­ty of urban air space through a micro­scop­ic sim­u­la­tion and illus­trate the emer­gence of sys­tem fun­da­men­tal dia­gram (for prop­er­ly defined aver­ages tak­en over four-dimen­sion­al space) com­pa­ra­ble in shape to urban road traf­fic networks.

Keynote Session 3 — Nikolas Geroliminis 

Title: On the Inef­fi­cien­cy and Man­age­ment of Ride-Sourc­ing Ser­vices towards Urban Con­ges­tion
Speak­er: Niko­las Geroli­m­in­is
Abstract: Human mobil­i­ty in con­gest­ed city cen­ters is a com­plex dynam­i­cal sys­tem with high den­si­ty of pop­u­la­tion, many trans­port modes to com­pete for lim­it­ed avail­able space and many oper­a­tors that try to effi­cient­ly man­age dif­fer­ent parts of this sys­tem. New emerg­ing modes of trans­porta­tion, such as ride-hail­ing and on-demand ser­vices cre­ate addi­tion­al oppor­tu­ni­ties, but also more com­plex­i­ty. Lit­tle is known about to what degree its oper­a­tions can inter­fere in traf­fic con­di­tions, while replac­ing oth­er trans­porta­tion modes, or when a large num­ber of idle vehi­cles is cruis­ing for pas­sen­gers. We exper­i­men­tal­ly ana­lyze the effi­cien­cy of TNCs using taxi trip data from a Chi­nese megac­i­ty and an agent-based sim­u­la­tion with a trip-based MFD mod­el for deter­min­ing the speed. We inves­ti­gate the effect of expand­ing fleet sizes for TNCs, pas­sen­gers’ incli­na­tion towards shar­ing rides, and strate­gies to alle­vi­ate urban con­ges­tion. We observe that, although a larg­er fleet size reduces wait­ing time, it also inten­si­fies con­ges­tion, which, in turn, pro­longs the total trav­el time. Such con­ges­tion effect is so sig­nif­i­cant that it is near­ly insen­si­tive to pas­sen­gers’ will­ing­ness to share and flex­i­ble sup­ply. Final­ly, park­ing man­age­ment strate­gies can pre­vent idle vehi­cles from cruis­ing with­out assigned pas­sen­gers, mit­i­gat­ing the neg­a­tive impacts of ride-sourc­ing over con­ges­tion, and improv­ing the ser­vice qual­i­ty. We are also devel­op­ing dif­fer­ent type of con­trol strate­gies, such as relo­ca­tion of emp­ty vehi­cles, park­ing man­age­ment and pric­ing incen­tives to alle­vi­ate the neg­a­tive effects.

Keynote Session 4 — Alexandre Bayen 

Title: Lagrangian Con­trol at Large and Local Scales in Mixed Auton­o­my Traf­fic Flow
Speak­er: Alexan­dre Bayen
Abstract: This talk inves­ti­gates Lagrangian (mobile) con­trol of traf­fic flow at local scale (vehic­u­lar lev­el). The ques­tion of how self-dri­ving vehi­cles will change traf­fic flow pat­terns is inves­ti­gat­ed. We describe approach­es based on deep rein­force­ment learn­ing pre­sent­ed in the con­text of enabling mixed-auton­o­my mobil­i­ty. The talk explores the grad­ual and com­plex inte­gra­tion of auto­mat­ed vehi­cles into the exist­ing traf­fic sys­tem. We present the poten­tial impact of a small frac­tion of auto­mat­ed vehi­cles on low-lev­el traf­fic flow dynam­ics, using nov­el tech­niques in mod­el-free deep rein­force­ment learn­ing, in which the auto­mat­ed vehi­cles act as mobile (Lagrangian) con­trollers to traf­fic flow. Illus­tra­tive exam­ples will be pre­sent­ed in the con­text of a new open-source com­pu­ta­tion­al plat­form called FLOW, which inte­grates state of the art microsim­u­la­tion tools with deep-RL libraries on AWS EC2. Inter­est­ing behav­ior of mixed auton­o­my traf­fic will be revealed in the con­text of emer­gent behav­ior of traf­fic: https://flow-project.github.io/

Keynote Session 5 — Henry Liu 

Title: Intel­li­gent Dri­ving Intel­li­gence Test for Autonomous Vehi­cles with Nat­u­ral­is­tic and Adver­sar­i­al Dri­ving Envi­ron­ment
Speak­er: Hen­ry Liu
Abstract: Dri­ving intel­li­gence tests are crit­i­cal to the devel­op­ment and deploy­ment of autonomous vehi­cles. The pre­vail­ing approach tests autonomous vehi­cles in life-like sim­u­la­tions of the nat­u­ral­is­tic dri­ving envi­ron­ment. How­ev­er, due to the high dimen­sion­al­i­ty of the envi­ron­ment and the rareness of safe­ty-crit­i­cal events, hun­dreds of mil­lions of miles would be required to demon­strate the safe­ty per­for­mance of autonomous vehi­cles, which is severe­ly inef­fi­cient. We dis­cov­er that sparse but adver­sar­i­al adjust­ments to the nat­u­ral­is­tic dri­ving envi­ron­ment, result­ing in the nat­u­ral­is­tic and adver­sar­i­al dri­ving envi­ron­ment, can sig­nif­i­cant­ly reduce the required test miles with­out loss of eval­u­a­tion unbi­ased­ness. By train­ing the back­ground vehi­cles to learn when to exe­cute what adver­sar­i­al maneu­ver, the pro­posed envi­ron­ment becomes an intel­li­gent envi­ron­ment for dri­ving intel­li­gence test­ing. We demon­strate the effec­tive­ness of the pro­posed envi­ron­ment in a high­way-dri­ving sim­u­la­tion. Com­par­ing with the nat­u­ral­is­tic dri­ving envi­ron­ment, the pro­posed envi­ron­ment can accel­er­ate the eval­u­a­tion process by mul­ti­ple orders of magnitude.

Keynote Session 6 — Fengmin Gong 

Title: Bet­ter Jour­neys For All Through Impact, Inno­va­tion & Respon­si­bil­i­ty
Speak­er: Feng­min Gong
Abstract: Data sci­ence and AI are at the core of the “fourth indus­tri­al res­o­lu­tion”. While the sci­ence and tech­nol­o­gy com­mu­ni­ty are dili­gent­ly push­ing the fron­tier for the ben­e­fits of human­i­ty, some fear the neg­a­tive impact of the same. The crust of the mat­ter is that, Data sci­ence and AI are pow­er­ful tools with huge poten­tial, HOW we har­ness this pow­er is the most crit­i­cal fac­tor to suc­cess or dis­as­ter. In this talk, I will share three main guid­ing prin­ci­ples — impact, inno­va­tion, and respon­si­bil­i­ty, which should help us to do the right things the right way in apply­ing AI. DiDi has been at the fore­front in trans­form­ing trans­porta­tion through AI. To illus­trate these prin­ci­ples, I will use some exam­ples in rein­force­ment learn­ing for opti­miza­tion, NLP for safe rides, and use-case dri­ven sim­u­la­tion for AV.

Keynote Session 7 — Kay Axhausen 

Title: Think­ing about the Long-Term Impacts of the Pan­dem­ic
Speak­er: Kay Axhausen
Abstract: The pan­dem­ic has accel­er­at­ed a num­ber of trends with a big impact on the trans­port sys­tem: work­ing from home and e‑commerce. The pre­sen­ta­tion will out­line the behav­iour­al changes observed in the last year using a sub­stan­tial Swiss GPS track­ing pan­el. Based on these changes it will dis­cuss, if these are enough to address the dilem­ma of trans­port plan­ning between acces­si­bil­i­ty improve­ments and induced demand, espe­cial­ly giv­en our duty to reduce GHG emissions.

Keynote Session 8 — Chandra Bhat 

Title: What Can We Learn about Trav­el and Safe­ty Impli­ca­tions from Par­tial­ly Auto­mat­ed Vehi­cle Use?
Speak­er: Chan­dra Bhat
Abstract: Inves­ti­gat­ing the poten­tial activ­i­ty-trav­el behav­ior impacts of ful­ly autonomous vehi­cles (des­ig­nat­ed as Lev­el 5 automa­tion on the Soci­ety of Auto­mo­tive Engi­neers or SAE scale) can only be under­tak­en today through stat­ed pref­er­ence or SP sur­veys (that is, ask­ing indi­vid­u­als how they may change their mobil­i­ty pat­terns in a hypo­thet­i­cal envi­ron­ment with a Lev­el 5 vehi­cle). But indi­vid­u­als may not be in a posi­tion to pro­vide appro­pri­ate respons­es when thrust into a hypo­thet­i­cal envi­ron­ment that is dif­fi­cult to con­jure up. In this regard, SAE Lev­el 1 fea­tures (such as adap­tive cruise con­trol or park­ing assist fea­tures) are in most new vehi­cles today, while many high­er-end vehi­cles today also achieve Lev­el 2 automa­tion (such as vehi­cles with adap­tive cruise con­trol, hands-free lane chang­ing, and self-park­ing). The avail­abil­i­ty and use of these vehi­cles today, albeit with low­er lev­els of automa­tion, can pro­vide impor­tant and reli­able insights on how trav­el pat­terns may change with advanc­ing tech­nol­o­gy. In this paper, we pro­pose to exam­ine poten­tial mobil­i­ty changes due to tech­nol­o­gy fea­tures that exist today in vehi­cles. Impor­tant­ly, while some ear­li­er stud­ies have exam­ined con­sumer accep­tance of exist­ing vehi­cle tech­nol­o­gy, we go beyond con­sumer accep­tance to also exam­ine how indi­vid­u­als with and with­out automa­tion fea­tures in their vehi­cles dif­fer in their annu­al vehi­cle miles of trav­el (VMT). Poten­tial impli­ca­tions for road­way safe­ty due to VMT changes are also discussed.