Zejiang Wang was born in Xi’an, Shaanxi, China in April 1992. He joined the Oak Ridge National Laboratory in July 2022, as a Postdoctoral Research Associate, and was early promoted to R&D Associate Staff in Feb 2023. He received his Ph.D. degree in Mechanical Engineering from The University of Texas at Austin in May 2022, the Dipl. Ing. degree from ENSTA ParisTech, France, the double M.S. degree in Design, Modeling, and Architecture of Complex Industrial Systems from Ecole Polytechnique, University of Paris-Saclay, France, both in 2017, and the B.E. degree (Hons.) in Mechanical Engineering and Automation from Southeast University, Nanjing, China, in 2014. He was a research intern at the French Institute of Petroleum (IFPen) in Rueil-Malmaison, France, the Mitsubishi Electric Research Laboratories in Cambridge, MA, USA, and a visiting scholar at the French National Centre for Scientific Research (CNRS). He received the IEEE ITSS Best Dissertation-First Prize from IEEE, Professional Development Awards and George J. Heuer, Jr. Ph.D. Endowed Graduate Fellowship from the University of Texas at Austin, the ATS Technical Committee Best Paper Award from the American Society of Mechanical Engineers (ASME), and the National Scholarship from the Ministry of Education of China. His research interests include vehicle dynamics and control, intelligent transportation systems, and cyber-physical systems.
Driving safety enhancement is an eternal topic since traffic casualties and the consequent economic loss remain salient. For instance, there were 12,107 serious injury crashes in Texas in 2020, with 14,656 people sustaining a severe injury. The estimated economic loss of all motor vehicle crashes reached more than 43 billion dollars, doubling the counterpart in 2003. Fortunately, recent progress on sensing, real-time computing, and vehicle-to-vehicle (V2V) communication has bestowed ground vehicles with unprecedented levels of situation awareness, computational capacity, and connectivity, which, if harnessed appropriately, could yield a society with zero crashes. Indeed, various advanced driving assistance systems (ADAS) and automated driving systems (ADS) have been invented to help a driver cope with challenging situations. However, the competition among control algorithms for the limited onboard implementation resources, e.g., the processor time and memory space, intensifies. Reducing the implementation resources consumption while guaranteeing the control performance of ADAS/ADS, or equivalently, enhancing the control performance of ADAS/ADS with the same amount of implementation resources consumption, constitutes the first research subject of this dissertation. With both the ADAS and the human driver concurrently controlling a vehicle, ADAS should consider individuals' driving characteristics and preferences to mitigate the human-machine conflict and enhance the overall performance of the driver-vehicle system. Moreover, driving assistance systems should detect driver behavior variation to adapt their intervention in real-time. Designing human-centric ADAS becomes the second research subject of the dissertation. Finally, V2V communication can improve drivers' situation awareness and reduce traffic-related casualties. However, the Federal Communication Commission in the United States mandated that only one fixed channel, among the seven available channels within the Dedicated Short-Range Communication (DSRC) spectrum, could be employed to transmit the safety messages. This restriction can entail severe package transmission delay and impair the safety benefits of V2V communication. To reduce the transmission delay and package missing rate, we invented and experimentally validated a dynamic channel selection algorithm, forming the dissertation's third research subject. The three research subjects, focusing on, respectively, embedded computing unit level, human-vehicle system level, and vehicular network level, complement each other and answer the question of driving safety enhancement from a human-cyber-physical perspective.