Max Robotics — Global Service Robot Directory + US Certification

About

Wee Sun Lee is a researcher specializing in robot planning under uncertainty, reinforcement learning, and human-robot interaction, with a particular focus on making autonomous systems more robust and practical in real-world environments. His most influential contribution is the development of POMDP-lite, a computationally tractable subclass of Partially Observable Markov Decision Processes (POMDPs) that enables principled robot planning under uncertainty without the prohibitive computational costs of general POMDP solving — a paper that has garnered 48 citations and become a foundational reference in the field. Building on this framework, Lee extended these ideas to robotic grasping tasks, demonstrating how POMDP-based approaches can improve object manipulation under sensor uncertainty, work that has attracted 18 citations. More recently, his research has expanded into deep model-based reinforcement learning, where he introduced Contrastive Variational Reinforcement Learning (CVRL), addressing the significant challenge of decision-making from complex natural visual observations. He has also explored human-robot interaction through intention modeling. Across his body of work, Lee consistently bridges theoretical probabilistic frameworks with practical robotics challenges, making him a notable contributor to intelligent autonomous systems research.