Skip to main content


We study robot dynamics, control, and design. We use techniques from the fields of applied mechanics, nonlinear dynamics, optimal control, machine learning, and computational design. Our approach is multidisciplinary and heavily utilizes theoretical and numerical investigations. We aim to develop robots – machines equipped with physical intelligence (adaptability) and control intelligence (decision making capability) – that parallel biological systems.


Mechanically Adaptive but Energetically Passive Robots: We develop robots that can assist humans without using external energy.
Robots Teaching Robots: We develop a new control paradigm to teach artificial systems.
Real-time Data-Driven Optimal Control of Robots: We develop algorithms to bypass the computational limitation and the simulation reality gap hindering the use of optimal control in applications.


NSF CAREER Award: Energy-efficient mechanically-adaptive robotics could lead to new-generation industrial robots, transportation systems, and devices that can assist humans
NSF (CMMI/DCSD) Award: What if robots could teach robots the same way parents teach their children?
Our paper “How to Run 50% Faster without External Energy”
was published in Science Advances in March 2020
Our paper “How to Run 50% Faster without External Energy”
was publicized in “The Guardian”
Our theoretical work on robot exoskeletons is featured in “The Conversation”