Max Robotics — Global Service Robot Directory + US Certification

About

Yajing Zang is a rising researcher in intelligent robotics, specializing in reinforcement learning, imitation learning, and motion planning for complex assembly tasks. Her work focuses on bridging the gap between human demonstration and autonomous robot skill acquisition, particularly in high-precision peg-in-hole operations. Zang’s major contributions include developing geometric-feature representation pre-training methods that significantly improve reinforcement learning efficiency for assembly tasks, reducing unnecessary learning steps while maintaining compatibility with mathematical optimization. Her 2023 paper on this topic has garnered 12 citations. She has also advanced imitation learning by introducing Probabilistic Movement Primitives (ProMPs) under task geometric representation, addressing the challenge of sparse and imperfect human demonstration data. In motion planning, Zang has proposed linearly constrained quadratic programming approaches for creating collision-free trajectories, earning 6 citations. Her most recent work (2024) introduces human skill knowledge-guided global trajectory policy reinforcement learning, enabling robots to adapt learned trajectory knowledge to new environments through environmental interaction—a key step toward more flexible and generalizable robotic manipulation. With a growing citation record and a clear trajectory toward integrating human expertise with autonomous learning, Zang is establishing herself as a thoughtful contributor to next-generation robotic assembly and skill transfer.