Max Robotics — Global Service Robot Directory + US Certification

About

Hongxu Liu is a leading researcher in robot learning for contact-rich manipulation, with a focus on enabling robots to perform complex assembly and manufacturing tasks. His work bridges learning from demonstration and autonomous exploration, developing methods that allow torque-controlled robots to acquire dexterous skills without exhaustive manual programming. Liu’s most cited paper, “Proactive Action Visual Residual Reinforcement Learning for Contact-Rich Tasks Using a Torque-Controlled Robot” (13 citations), introduces a novel framework that combines visual perception with residual reinforcement learning to handle the multimodal, high-variance dynamics of physical contact. His complementary study, “Combining Learning from Demonstration with Learning by Exploration to Facilitate Contact-Rich Tasks” (12 citations), demonstrates how robots can efficiently learn constrained-space operations by bootstrapping from human examples and then refining through self-guided practice. This dual approach reduces programming time while improving task success in real-world settings. Liu’s contributions are particularly impactful for collaborative robotics, where rapid reconfiguration of assembly lines demands adaptable, safe, and sample-efficient learning. His work has been recognized for advancing practical robot autonomy in manufacturing, with cumulative citations reflecting growing influence in the field of robot manipulation and reinforcement learning.