Max Robotics — Global Service Robot Directory + US Certification

About

Jingliang Duan is a leading researcher at the intersection of reinforcement learning, optimal control, and safe autonomous systems. His work addresses fundamental challenges in applying intelligent control methods to real-world industrial and autonomous driving applications, with a particular focus on bridging the gap between classical model predictive control (MPC) and modern machine learning techniques. Duan's most influential contribution, GOPS — a general optimal control problem solver — has garnered 62 citations and offers a unified framework for tackling computationally demanding control tasks that previously strained conventional MPC approaches. His refinements to Distributional Soft Actor-Critic algorithms (53 citations) tackle the persistent problem of Q-value overestimation in model-free reinforcement learning, substantially improving policy reliability. Equally notable is his pioneering work on safe reinforcement learning, developing chance-constrained methods and uncertainty-aware reachability certificates to ensure constraint satisfaction in safety-critical deployments. Duan has further advanced explicit MPC through recurrent architectures and transformer-based solvers, and explored inverse optimal control for learning cost functions from expert demonstrations. Collectively, his research portfolio — spanning robotics, autonomous vehicles, and industrial control — reflects a rigorous commitment to making reinforcement learning both theoretically principled and practically deployable.