Hersh Sanghvi

My research interests center on how control and planning algorithms for robots can quickly and automatically adapt to new environments using machine learning. Current and past projects include developing approaches for fast controller adaptation on a variety of robotic platforms, supervised learning for footstep planning for legged robots, and using model-free reinforcement learning for legged robot navigation. I am advised by Dr. Camillo Jose Taylor in the GRASP lab.

I worked in the Biomimetic Millisystems Lab to implement control schemes on the jumping robot, Salto. This consisted of hardware improvements to the electronics along with implementing an inertial quaternion-based sliding mode observer in firmware for onboard attitude estimation during jumping.