Max Robotics — Global Service Robot Directory + US Certification

About

Hongwei Ge’s research sits at the intersection of reinforcement learning, evolutionary computation, and robotics, with a focus on enabling more adaptive, efficient, and structurally intelligent systems. A central theme in their work is the integration of human feedback into learning agents, as demonstrated in their 2018 paper “Adaptively Shaping Reinforcement Learning Agents via Human Reward,” which has garnered 7 citations and explores how non-expert human input can guide agent behavior. Ge has also made notable contributions to multi-task optimization and quality diversity, including a 2022 study on multi-task MAP-Elites with knowledge transfer for robotic arm design—addressing the challenge of designing robotic arms under varying constraints. Their work on decentralized multiagent reinforcement learning using coordination graphs (2018) and decomposed deep reinforcement learning for robotic control (2020) tackles high-dimensional control problems by breaking them into manageable, interactive sub-problems. Most recently, Ge’s 2025 paper on evolutionary heterogeneous multitasking for quality diversity optimization pushes the boundaries of QD algorithms beyond single-task settings. With a growing citation footprint and a clear trajectory toward scalable, human-aware, and structurally decomposed robotic intelligence, Hongwei Ge is shaping the next generation of adaptive robotic systems.