Max Robotics — Global Service Robot Directory + US Certification

About

Yan Duan is a robotics and machine learning researcher whose work spans motion planning, imitation learning, and visuomotor control — areas that lie at the intersection of classical robotics and modern deep learning. His most influential contribution, "Motion Planning with Sequential Convex Optimization and Convex Collision Checking" (2014), introduced a principled optimization-based framework for generating collision-free robot trajectories, earning over 840 citations and establishing itself as a foundational reference in the motion planning community. Duan also made significant strides in data-efficient robot learning through his work on "One-Shot Imitation Learning" (2017), which demonstrated that robots could generalize to new tasks from a single demonstration — a compelling step toward flexible, generalizable robot intelligence, with over 229 citations. His research on Gaussian belief space planning addressed the challenges of uncertainty and sensing discontinuities in robot manipulation, particularly for low-cost, imprecise robots. Additionally, his work on deep spatial autoencoders explored how robots can autonomously learn compact visual representations to support reinforcement learning in unstructured environments. Together, Duan's contributions reflect a sustained effort to make robots more capable, adaptable, and practical in real-world settings.