Bohao Xu

Papers

Total Citations

194

H-Index

About

Bohao Xu is a leading researcher in intelligent robotic manipulation and advanced manufacturing, with a primary focus on reinforcement learning for complex assembly tasks and additive manufacturing process control. His most impactful work, "Feedback Deep Deterministic Policy Gradient With Fuzzy Reward for Robotic Multiple Peg-in-Hole Assembly Tasks" (2018, 179 citations), addresses the long-standing challenge of automating multi-step assembly without complex contact modeling. By formulating the task as a Markov decision process and introducing a model-driven deep reinforcement learning approach with fuzzy reward shaping, Xu’s method enables robots to learn precise insertion strategies autonomously, significantly reducing the need for hand-crafted controllers. This contribution has become a foundational reference for researchers applying deep RL to industrial robotics. In parallel, Xu has advanced wire and arc additive manufacturing (WAAM) through his work on "Shape-driven control of layer height in robotic WAAM" (2019, 15 citations), where he developed a novel control method to align uneven substrates—a critical step for achieving consistent layer deposition. His research bridges the gap between learning-based control and practical manufacturing constraints, offering scalable solutions for automation. Xu’s work continues to influence both the robotics and manufacturing communities, demonstrating how intelligent algorithms can solve real-world industrial challenges.