Max Robotics — Global Service Robot Directory + US Certification

About

Lingheng Meng is a researcher specializing in deep reinforcement learning, human-robot interaction, and autonomous decision-making under uncertainty. His most significant contribution lies in advancing reinforcement learning for partially observable environments — a critical challenge in real-world applications where agents cannot access complete state information. His work on memory-based deep reinforcement learning for Partially Observable Markov Decision Processes (POMDPs), which has accumulated over 76 citations, proposes end-to-end learning frameworks that equip agents with memory mechanisms to handle incomplete observability, bringing reinforcement learning meaningfully closer to practical deployment. Beyond theoretical contributions, Meng has demonstrated a strong commitment to applied research, conducting field studies in real-world settings such as public museums to evaluate how autonomous systems can generate engaging, life-like behavior in dynamic, multi-person environments. His work on individualized affective human-machine interaction further highlights his interest in personalizing autonomous behavior to individual human preferences using reinforcement learning. Collectively, his research bridges the gap between controlled laboratory settings and the messy complexity of real-world human-machine interaction, making him a notable contributor to the fields of autonomous systems and socially intelligent robotics.