Videos

These videos showcase my research toward capable robotic systems with 3D vision, world models, robot learning, motion planning, optimal control and teleoperation. The goal is to enable robots to perceive their surroundings, reason about actions, and perform reliably on real hardware. My long-term vision is to build robotic foundation models that transfer across tasks, requiring less task-specific data while achieving stronger generalization to new objects, scenes, and environments. This work bridges academic research in perception, learning, and control with real-world robotic applications.

Pick and Place

Pick and place in dense clutter
Pick-and-place persistence under disturbance

Human-Robot Interaction and Safety

Adaptive human-robot interaction for fruit delivery
Safety behavior for collision avoidance

Contact-Rich Skills

Contact-rich insertion
Whole-body lifting with contact-rich coordination

Dexterous Hands

Dexterous hand grasping in clutter
Teleoperation