Shamel Fahmi

I am a Research Scientist and Technical Lead in Reinforcement Learning at the Robotics and AI (RAI) Institute, formerly known as the Boston Dynamics AI Institute (BDAI). My research focuses on robot locomotion and learning-based control for a wide range of platforms, including quadrupeds, bipeds, humanoids, wheeled robots, and others.

At RAI, I am part of the UMV team where I develop RL behaviors for the UMV robots. Before RAI, I was at MIT, where I worked with Prof. Sangbae Kim on reinforcement learning and whole-body control for the Mini Cheetah and the MIT Humanoid. Prior to that, I was a researcher at the Italian Institute of Technology (IIT), where I worked on both optimization-based and learning-based locomotion controllers for the HyQ and HyQReal quadrupedal robots, as well as perception and state estimation.

Over the years, I have also worked at the German Aerospace Center (DLR), the University of Twente, and RWTH Aachen University on a variety of robotics and controls projects.

I have more than 10 years of experience in robotics and over 5 years of experience in Physical AI and reinforcement learning. My work sits at the intersection of robot learning, locomotion, control, and perception, with the goal of building agile, robust, and intelligent robots that can operate in the real world.

Outside of research, I enjoy racing triathlons and painting.