Yun-Hao Cao
Papers
1
Total Citations
12
H-Index
1
About
Yun-Hao Cao is a roboticist advancing the frontier of real-world reinforcement learning (RL) for legged locomotion. His research centers on bridging the gap between simulation-trained policies and practical deployment, with key contributions in sample-efficient, safe, and stable RL algorithms. In his highly cited work, "Grow Your Limits: Continuous Improvement with Real-World RL for Robotic Locomotion" (2024, 12 citations), Cao introduced APRL—a policy reuse framework that enables robots to continuously refine locomotion skills directly in the physical world without resetting to safe states. This breakthrough addresses critical constraints of efficiency, safety, and training stability, allowing quadrupedal robots to autonomously acquire and improve complex behaviors over extended periods. By demonstrating that robots can "grow their limits" through iterative real-world learning, Cao’s work has significant implications for deploying autonomous systems in unstructured environments. His achievements mark an important step toward practical, lifelong learning for robotic platforms, inspiring researchers and students working on the intersection of reinforcement learning, control, and embodied AI.
Research Focus
Key Achievements
Top Papers
- 1