Papers
11
Total Citations
164
H-Index
6
About
Bowen Weng is a leading researcher in bipedal robotics, specializing in the intersection of reinforcement learning (RL) and model-based control for dynamic locomotion. His major contributions include pioneering hierarchical frameworks that combine RL-based high-level planners with model-based low-level controllers, enabling robust 3D bipedal walking on hardware like Agility Robotics' Digit robot. His work on integrating RL with Hybrid Zero Dynamics (HZD) has produced novel feedback control policies that achieve stable, agile locomotion without relying on reference trajectories. Weng's research has garnered over 160 citations, with his most influential paper on robust feedback motion policy design for Digit accumulating 67 citations. He has also advanced safety testing methodologies for legged robots, developing scenario-based algorithms to validate and characterize locomotion safety under uncertainty. Notable achievements include demonstrating the first successful implementation of RL-based cascade controllers on a full-scale 3D biped, and proposing standardized disturbance rejection testing protocols to accelerate commercialization. His work bridges theoretical control theory and practical deployment, offering students and researchers a clear path from algorithm design to real-world robotic locomotion.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Reinforcement Learning Meets Hybrid Zero Dynamics: A Case Study for RABBIT21 citations · 2019
- 4Template Model Inspired Task Space Learning for Robust Bipedal Locomotion20 citations · 2023
- 5
- 6
- 7
- 8
- 9
- 10