The Trustworthy Autonomous Mobility System (TAMS) Lab at UCLA develops embodied AI agents that can safely and reliably operate in human-centered environments. Our research lies at the intersection of control, robotics, and learning. We are dedicated to exploring principled approaches that integrate data-driven methods (deep learning, generative models, reinforcement learning, imitation learning) with control theory, explainable AI, and causality. Our work focuses on applications in autonomous driving and robot navigation. Our long-term vision is to facilitate trustworthy and human-centered autonomy, aiming to expedite the integration of autonomous systems into everyday life to yield substantial societal benefits.