Latest News
After many years of work, I'm very excited to release our preprint on Maximum Diffusion Reinforcement Learning. MaxDiff RL is a novel decision-making framework built from the ground up to take into account the physical properties of embodied agents during learning and control. MaxDiff RL agents are provably robust and capable of learning in single-shot deployments. Separately, I recently published a perspective piece in Advanced Intelligent Systems on the future of soft robot design, learning, and control across scales in collaboration with Dr. Ryan Truby. Check out one of our MaxDiff RL paper videos:
Last year, I published an article in Nature Communications. I was also interviewed by the Northwestern News team about the work. In this collaboration with chemical engineers at the Strano Research Group in MIT, I make use of the physics of self-organization to enable sophisticated capabilities in a simple collective of active colloidal particles. Making use of my previous work on rattling theory, I discovered a thermodynamic mechanism for asymmetry-induced order in strongly-interacting systems. We exploited this mechanism to generate alternating currents directly on-board particles, which allowed us to power state-of-the-art microrobotic arms. Check out the video overview:
I am also proud to be a recipient of the Presidential Fellowship! The highest honor bestowed to a graduate student by Northwestern University.
About Me
I am a Ph.D. candidate and Presidential Fellow at Northwestern University's Center for Robotics and Biosystems. I am a member of the Interactive and Emergent Autonomy Laboratory, and am advised by Dr. Todd D. Murphey. I received my B.S. in Engineering from Harvey Mudd College in 2017 and my M.S. in Mechanical Engineering from Northwestern University in 2020.
I am an interdisciplinary scientist and engineer interested in exploring and exploiting the consequences of agent embodiment in robot learning and control across scales. My work combines insights from artificial intelligence, statistical physics, chemical engineering, and materials science to make engineered systems more life-like by exploiting their physical intelligence and emergent capabilities.
Mastodon