Changhao Wang
Papers
19
Total Citations
408
H-Index
10
About
Changhao Wang is a robotics researcher whose work centers on robotic manipulation of deformable objects, legged locomotion, and robot learning. He is perhaps best known for pioneering frameworks that enable robots to reliably track, model, and control deformable linear objects (DLOs) such as cables and ropes — notoriously difficult problems due to their infinite-dimensional configuration spaces. His 2018 framework leveraging Coherent Point Drift (73 citations) laid important groundwork in this domain, followed by increasingly sophisticated approaches incorporating graph neural networks for offline-online deformation learning (62 citations) and spatial representations for cable routing (46 citations). Wang has also made notable contributions to legged locomotion, proposing online residual model learning to compensate for real-world dynamics mismatches in model-based controllers (52 citations). His more recent research expands into dual-arm manipulation in constrained 3D environments, primitive-based skill learning for robotic assembly, and SE(3)-equivariant geometric control for contact-rich tasks — reflecting a broadening toward generalizable, data-efficient robot learning. With over 370 cumulative citations across his published work, Wang has established himself as a significant voice in advancing robust, intelligent robotic manipulation systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Robotic Cable Routing with Spatial Representation46 citations · 2022
- 5Robust Deformation Model Approximation for Robotic Cable Manipulation41 citations · 2019
- 6
- 7
- 8
- 9
- 10