Chaoyi Pan
Papers
2
Total Citations
22
H-Index
2
About
Chaoyi Pan is a robotics researcher advancing the frontiers of legged locomotion and bimanual manipulation through novel control and learning frameworks. His work tackles two fundamental challenges in robotics: real-time optimal control for high-dimensional systems and efficient multi-arm coordination. In his highly cited 2025 paper on “Full-Order Sampling-Based MPC for Torque-Level Locomotion Control via Diffusion-Style Annealing,” Pan introduces a sampling-based model predictive control approach that overcomes the non-convexity and dimensionality barriers that have traditionally limited legged robot controllers to reduced-order models or local approximations. This work, already garnering 12 citations, opens the door to full-order dynamics in real-time locomotion. Equally impactful is his 2023 paper on “Efficient Bimanual Handover and Rearrangement via Symmetry-Aware Actor-Critic Learning,” which addresses the complexity of multi-object bimanual tasks. By leveraging symmetry in reinforcement learning, Pan’s method enables dual-arm robots to rearrange objects with unprecedented speed and coordination, earning 10 citations. Together, these contributions demonstrate Pan’s ability to blend theoretical rigor with practical robotic systems, making him a rising figure in both model-based and learning-based robot control.
Research Focus
Key Achievements
Top Papers
- 1
- 2