At the Dexterous Manipulation and Soft Robotics lab, we study how hands — robotic, animated, and human — grasp, manipulate, and make sense of the physical world.
Our research spans the full depth of dexterous manipulation: from the geometry of contact and the biomechanics of the human hand, to algorithms for grasp planning, motion retargeting across radically different hand morphologies, and robot systems that recover gracefully from uncertainty.
Rather than treating manipulation as a solved problem of final-position control, we focus on the process — the real-time sensing, adjustment, and physical intelligence that makes skilled handling possible.
Our work bridges fundamental theory and working hardware, with results demonstrated on physical robot platforms and validated in collaboration with fields ranging from computer animation to biomechanics.
Underlying everything is a long-term conviction: that giving robots genuinely human-like dexterity requires understanding what human dexterity actually is.
