Reinforcement Learning for Orbital Transfers at the 2026 AI Winter School (Brown University)

I recently gave a 2.5-hour, hands-on workshop on using reinforcement learning (RL) to solve a simplified orbital transfer problem at the 2026 AI Winter School, hosted by the Center for the Fundamental Physics of the Universe at Brown University.

We used the classic Hohmann transfer as an analytic baseline, then built a small 2D, two-body simulator and trained PPO agents (both discrete and continuous thrust control) to learn transfer policies in simulation. Along the way, we focused on practical workflow: environment design, reward shaping, debugging training runs, and interpreting failure modes.