Papers
9
Total Citations
43
H-Index
4
About
Gaotian Wang is a robotics researcher whose work spans soft robotics, reinforcement learning-based control, and robotic manipulation, with a growing body of contributions that bridges theoretical modeling and practical robot autonomy. His most significant work focuses on developing intelligent control strategies for soft robotic arms, systems celebrated for their passive compliance and inherent safety but notoriously difficult to model and control. Wang has advanced reinforcement learning methods tailored for soft robots, proposing Q-learning frameworks that leverage data from imperfect simulators to overcome poor sample efficiency—a persistent barrier in the field—and designing constraint-aware motion controllers for hyper-redundant arms, work that has collectively accumulated over 20 citations. Beyond soft robotics, Wang has tackled fundamental challenges in robotic manipulation, including unseen object segmentation through interactive perception, nonprehensile pushing under uncertainty, and collision-inclusive planning for occluded grasping. His recent work on "caging in time" introduces a novel framework for robust manipulation under perception limitations. Across these areas, Wang consistently addresses real-world complexity—uncertainty, dynamic deformation, and environmental interaction—making his research particularly relevant for robotics applications in unstructured, human-centered environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9