Papers
19
Total Citations
254
H-Index
11
About
Ching-An Cheng is a robotics and machine learning researcher whose work spans motion planning, robot learning, and human-robot interaction. His research sits at the intersection of geometric control theory, kernel methods, and reinforcement learning, with a consistent focus on making robotic systems safer, more capable, and more adaptable in complex real-world environments. Among his most influential contributions is the development of RMPflow, a geometric framework grounded in Riemannian geometry that enables robots to generate fluid, multitask motion policies in dynamic environments — a work that has garnered 29 citations and represents a significant advance in reactive motion planning. His early work on learning inverse dynamics using structured reproducing kernel Hilbert spaces (27 citations) and modeling Lagrangian systems (17 citations) demonstrated rigorous, principled approaches to data-driven robot dynamics modeling. His research on virtual impedance control for safe human-robot interaction (40 citations, his most-cited work) reflects a sustained commitment to physical safety in collaborative robotics. Cheng has also contributed to soft robotics sensing, underactuated robot control, probabilistic motion planning, and heuristic-guided reinforcement learning, showcasing impressive breadth. With over 175 cumulative citations across his top works, his research continues to shape how robots learn, move, and interact safely alongside humans.
Research Focus
Key Achievements
Top Papers
- 1Virtual Impedance Control for Safe Human-Robot Interaction40 citations · 2015
- 2
- 3
- 4
- 5
- 6
- 7Heuristic-Guided Reinforcement Learning14 citations · 2021
- 8Stable, Concurrent Controller Composition for Multi-Objective Robotic Tasks13 citations · 2019
- 9
- 10