Papers
11
Total Citations
228
H-Index
8
About
Helei Duan is a leading researcher in bipedal locomotion and humanoid robotics, with a focus on using reinforcement learning (RL) to achieve agile, real-world performance. His work bridges the gap between simulation and reality, enabling robots to walk, run, and manipulate objects in complex environments. Duan’s major contributions include developing learning-based controllers that operate in task space rather than joint space, allowing for more natural and efficient movement. He has pioneered vision-based locomotion, allowing bipedal robots to anticipate and adapt to challenging terrain, and has optimized running gaits for the Cassie robot, achieving speeds comparable to human sprinters. His research also addresses practical challenges like handling unsensed dynamic loads and performing loco-manipulation, such as lifting boxes while maintaining balance. With over 200 citations across his most-cited works, Duan’s impact is evident in advancing the robustness and versatility of humanoid robots. Notable achievements include his work on sample-efficient optimization for human-in-the-loop systems, which has implications for prosthetics and exoskeletons, and his recent studies on reward design for robust standing and walking. Duan’s research is paving the way for bipedal robots that can operate safely and effectively in human environments.
Research Focus
Key Achievements
Top Papers
- 1Learning Task Space Actions for Bipedal Locomotion45 citations · 2021
- 2Learning Vision-Based Bipedal Locomotion for Challenging Terrain32 citations · 2024
- 3
- 4Sim-to-Real Learning for Bipedal Locomotion Under Unsensed Dynamic Loads29 citations · 2022
- 5
- 6Sim-to-Real Learning for Humanoid Box Loco-Manipulation25 citations · 2024
- 7
- 8Learning Dynamic Bipedal Walking Across Stepping Stones15 citations · 2022
- 9Learning Task Space Actions for Bipedal Locomotion3 citations · 2020
- 10Learning Vision-Based Bipedal Locomotion for Challenging Terrain2 citations · 2023