Dr. Eunhyek Joa

Software Engineer | Zoox USA

Co-Evolving Motion Control in Software-Defined Vehicles : Robustness Across Environments, Retargeting Across Vehicles, and Reliable Trajectory Tracking

In software-defined vehicles (SDVs), driving behavior is no longer fixed by hardware or static control logic, but continuously shaped by evolving perception, planning, and decision-making software. As a result, motion control must co-evolve with upstream autonomy modules while remaining grounded in vehicle physics and safety constraints. This talk discusses three topics for motion control in SDVs. First, achieving robustness across diverse environments, including varying road conditions, friction levels, and environmental uncertainties, where motion control serves as the final safeguard for stability and safety. Second, enabling retargeting across vehicles and platforms, where learned behaviors, trajectories, and policies must be adapted to different actuation capabilities and dynamic limits. Third, ensuring reliable trajectory tracking, where planned trajectories with inherent uncertainty must be translated into feasible, safe, and comfortable vehicle motions on highly overactuated platforms.

Bio

Eunhyek Joa received the B.S. (2014) and M.S. (2016) degrees in Mechanical Engineering from Seoul National University, and the Ph.D. degree in Mechanical Engineering from the University of California, Berkeley, in 2024, where his research focused on data-driven model predictive control and its applications to connected vehicles. He subsequently spent a year at Zoox, working on autonomy behaviors for its robotaxi. He will join the Department of Mechanical Engineering at Seoul National University as an Assistant Professor in 2026. His research interests include autonomous driving with a focus on autonomy behaviors, vehicle motion control, vehicle connectivity, and machine learning for safety-critical systems.