Max Robotics — Global Service Robot Directory + US Certification

About

Peide Huang is a researcher working at the intersection of robust reinforcement learning, sim-to-real transfer, and robot learning, with a growing focus on integrating large language models into robotic systems. His most impactful contribution to date is his work on robust reinforcement learning framed as a Stackelberg game, where he introduced Adaptively-Regularized Adversarial Training to address fundamental limitations in existing Robust Adversarial Reinforcement Learning (RARL) frameworks — work that has garnered over 20 citations and advances the reliable deployment of RL agents under model errors and adversarial conditions. Huang has also tackled the persistent sim-to-real gap in robotics through a differentiable causal discovery framework, offering principled tools to diagnose and correct simulator inaccuracies that hinder real-world performance. More recently, his research has ventured into the creative application of large language models to enable robots to use tools in tasks requiring implicit physical reasoning and long-horizon planning. Collectively, his work reflects a coherent vision: building intelligent, robust robotic agents capable of operating reliably in complex, real-world environments — a contribution increasingly relevant as robotics research moves closer to practical deployment.