Distributed Learning for the Decentralized Control of Articulated Mobile Robots
The result of my first summer of research after my freshman year, where I utilized new distributed reinforcement learning algorithms to greatly improve the gait adaptation of a snake robot in challenging environments. Published in ICRA 2018.
Distributed Reinforcement Learning for Multi-Robot Decentralized Collective Construction
I worked to extend the A3C algorithm to enable multiple agents to learn a single policy, allowing them to work together toward a commn goal without explicitly interacting. Accepted to DARS 2018.