Zengxiang Lei

I am a postdoctoral associate at Purdue University working on transportation AI, simulation, and cyber-physical systems. My research builds scalable models and decision systems for ride-hailing, shared autonomous electric fleets, evacuation, and transportation security.

I completed my Ph.D. in Transportation and Infrastructure Systems Engineering at Purdue University, where my dissertation focused on efficient and secure ride-hailing services. I also served as student lead for the U.S. Department of Energy METS-R project and helped build METS-R SIM, an open-source city-scale multimodal electric mobility simulator.

My current research is organized around three connected directions:

• Transportation AI: reinforcement learning for pricing, relocation, and control in mobility-on-demand systems, including my Transportation Research Part B work on dynamic pricing in ride-hailing systems.
• Network-level simulation: city-scale simulation for multimodal electric mobility and service operations, centered on METS-R SIM.
• CPS operation and security: simulation-supported analysis of cyber-physical transportation systems, including network-level risks from adversarial booking attacks.

Recent work includes Stable GFlowNets with TV Monitoring and Probabilistic Guarantees and Controllable Road Marking Generation, both accepted to NeurIPS 2026. Our work on worst-case hidden-vehicle trajectory search received the Best Paper Award and was accepted for an oral presentation at the ECCV 2026 Workshop on Safe and Defensive Autonomous Driving (SDAD).