I work in the City Science group on last-meter delivery robots — how legged machines should move, navigate, and behave in cities shared with people.
Control & Autonomy
Learned and implemented advanced robotic control methods — reinforcement learning, LQR, MPC, and PID — on custom bipedal (6-DOF) and quadruped (16-DOF) delivery robot platforms. Deployed RL control policies on real-world hardware, demoing live to 20+ industry partners.
Built path-planning and low-level control software in ROS, integrating AprilTag localization, SLAM, and a novel multi-robot docking mechanism for cooperative navigation over complex urban obstacles and energy-optimized routing.
Human-Robot Interaction
Worked under Masters and PhD candidates on published research into the legibility and predictability of multi-platform robotic motion — see Do Robots Need Body Language? (ACM/IEEE HRI ’26). Currently exploring societal acceptance of robots in human-shared urban areas through social deference and apology stratagems.
Ongoing
Invited to stay on as the lab's only high-school researcher on a semipermanent basis (2028+), working between City Science Cambridge and Taipei.