On September 28, 2022, from 19:00 to 21:00 Beijing time, the seventh and eighth special lectures of the 2022 Transportation Series were successfully held by the Transportation Research Institute of the University of Leeds and Beijing Jiaotong University. The seventh special lecture was presented by Professor Stephane Hess from the Institute of Transportation at the University of Leeds, titled "Modeling choices: the conflict between realities, ease of implementation and prediction performance". Professor Liu Zhiyuan from Southeast University and Professor Dong Chunjiao from Beijing Jiaotong University were invited to comment. The title of the eighth special lecture was "Human Factors Challenges and Opportunities of Automated Vehicle: How we study driver and pedestrian interactions", presented by Professor Natasha Merat from the Institute of Transportation at the University of Leeds. Professor Feng Zhongxiang from Hefei University of Technology and Professor Ma Lu from Beijing Jiaotong University were invited to comment. The two lectures in this issue were hosted by Associate Professor Zhou Huiyu from the School of Transportation, Beijing Jiaotong University. They were broadcast live through Zoom online conferences, continuing to build academic dialogue bridges for Chinese and foreign transportation scholars.

Beijing Jiaotong University - Institute of Transportation, University of Leeds, UK

The 7th and 8th lectures of the 2022 Transportation Series will be held

Stephane Hess, the guest speaker of the seventh special lecture, is a professor at the Institute of Transportation at the University of Leeds and serves as the director of the Selection Model Center. Professor Stephane Hess is skilled in using advanced discrete choice models to analyze human decision-making, and his research covers multiple fields such as transportation, health, and environmental economics. He is the founding editor-in-chief of the Journal of Choice Modeling journal, as well as the founder and chairman of the steering committee of the International Conference on Choice Modeling. At the same time, he served as a member of the Editorial Advisory Committee for Transportation Research Part B and Transportation, and served on the academic committees of major international conferences in the field of transportation.

In this lecture, Professor Stephane Hess shared the latest research on selection modeling. He first introduced the process and objectives of behavioral modeling, noting that the main methodologies for choice decision analysis include choice modeling, mathematical psychology, and machine learning, each with its own emphasis and applicable in specific research contexts. Selection modeling focuses on why individuals make specific choices, emphasizing the heterogeneity of user preferences (and decision-making processes). Mathematical psychology focuses on how individuals make choices, emphasizing the process of making choices. Machine learning focuses on the results of individual selection processes and can fit estimation data well, but the individual selection process is often in a "black box" state. Then, Professor Stephane Hess used an empirical study as an example to introduce the processes of establishing three models: selection model, machine learning model, and mathematical psychology model, as well as the applicability of different methods. He then compared the accuracy of different models' predictions. Finally, Professor Stephane Hess pointed out the differences between these three models and emphasized the synergistic effects between different methods. Professor Stephane Hess specifically pointed out that in the modeling process, one should not only focus on fitting, but also pay attention to preserving the type of samples.

Professor Stephane Hess from the Institute of Transportation at the University of Leeds shares the latest research findings

Professor Liu Zhiyuan from Southeast University and Professor Dong Chunjiao from Beijing Jiaotong University, as guest reviewers for the seventh special lecture, gave in-depth comments on Professor Stephane Hess' academic report. Professor Liu Zhiyuan agreed with Professor Stephane Hess' proposed goal of model selection. He believed that machine learning models can improve the accuracy of traffic modeling. In addition, Professor Liu Zhiyuan and Professor Stephane Hess unanimously stated that human behavior is complex, and high-quality traffic models cannot ignore the connection with behavior.

Highlights of Professor Liu Zhiyuan from Southeast University

Professor Dong Chunjiao agreed with Professor Stephane Hess' viewpoint that fit does not necessarily mean everything. She believes that applying selection models can better assist policy making, but in contrast, the "black box" features of machine learning cannot explain why individuals make such decisions. In addition, Professor Dong Chunjiao maintains that the choice of travel mode and time is subject to many practical influential factors and further discussed with Professor Stephanehess about the measurement of information such as time value and personal preferences in the process of behavioral modeling.

Professor Dong Chunjiao's Exciting Comments from Beijing Jiaotong University

Following Professor Stephane Hess' excellent speech, there was the eighth special lecture, where the guest speaker, Professor Natasha Merat, shared the latest research on the impact mechanism of autonomous driving on human consciousness and behavior. Professor Natasha Merat is an experimental psychologist and the head of the Human Factors and Safety Center at the University of Leeds Transportation Research Institute. Her main research direction is to understand the interaction between road users and new technologies, with a focus on studying the factors that affect driver distraction and driver safety. She is an internationally recognized expert in the study of human factors in highly autonomous vehicles.

Professor Natasha Merat first introduced human factors research in autonomous driving from the driver's perspective in his lecture, including the amount and transfer of driver attention, the level of trust in autonomous driving, as well as the driver's own skills and age. Then, Professor Natasha Merat conducted field experiments simulating autonomous driving scenarios to study the effects of different interference environments on driver visual attention and behavioral responses during the autonomous driving process. The results indicate that the less visual information a driver receives during autonomous driving, the longer it takes to respond. Next, Professor Natasha Merat further explored the interaction between pedestrians and autonomous driving. Focus group interview research showed that pedestrians are most concerned about whether autonomous driving will detect pedestrians in a timely manner and avoid them. She further introduced the impact mechanism of the eHMI system, which interacted with pedestrians based on flash information, on pedestrians crossing the street. Finally, Professor Natasha Merat looked forward to the development direction of autonomous driving in the next 5 to 10 years.

Professor Natasha Merat from the Institute of Transportation at the University of Leeds shares the latest research findings

Professor Feng Zhongxiang from Hefei University of Technology and Professor Ma Lu from Beijing Jiaotong University, as the guest reviewers of the eighth special lecture, gave excellent comments on Professor Natasha Merat's academic report. Professor Feng Zhongxiang believes that the research methods introduced by Professor Natasha Merat based on pedestrian simulators and VR technology are very advanced, and the eHMI system can become an excellent carrier for interactive information between pedestrians and autonomous vehicles in road traffic. In future research, it is possible to gain a more detailed understanding of how eHMI affects road users other than pedestrians. Professor Feng Zhongxiang stated that many Chinese scholars have also conducted a series of research in the field of autonomous driving and look forward to in-depth cooperation with Professor Natasha Merat.

Professor Feng Zhongxiang from Hefei University of Technology's Exciting Comments

Professor Ma Lu believed that Professor Natasha Merat's speech on autonomous driving was informative and rich. He stated that safety is a great concern for both autonomous driving and current road traffic, and more laws and regulations are needed to promote the development of autonomous driving. Finally, Professor Ma pointed out that the situation of pedestrians is a key consideration for the future development of autonomous driving, and further in-depth research is urgently needed.

Highlights of Professor Ma Lu from Beijing Jiaotong University

The 7th and 8th symposiums were the last two sessions of the 2022 transportation lecture series jointly organized by the Transportation Institute of Beijing Jiaotong University and the University of Leeds, UK. They continued the academic style and depth of academic dialogue of previous symposiums, providing participants from home and abroad, researchers, teachers and students with excellent ideas, rigorous research paradigms, useful research results and profound insights. This has not only enriched the research on transportation theory and practice at home and abroad, but also provided useful academic support for the current implementation of the transportation power strategy in China.