Papers
21
Total Citations
787
H-Index
11
About
Xingye Da is a robotics researcher specializing in legged locomotion control, bipedal and quadrupedal robot dynamics, and the integration of machine learning with model-based control. His work sits at the productive intersection of control theory, optimization, and reinforcement learning, tackling some of the most challenging problems in robotic movement. Da's most influential contribution — his 2019 paper on feedback control for the Cassie bipedal robot (216 citations) — established foundational methods for walking, standing, and dynamic balancing using virtual constraints and gait libraries, helping make Cassie a benchmark platform for the robotics community. His earlier research systematically bridged the gap between planar robot models and real-world 3D implementation, enabling stable walking with speed tracking (110 citations). He has also pioneered the use of supervised learning to build robust locomotion policies tested in challenging outdoor environments, and developed elegant approaches to dynamic walking over unpredictable discrete terrain using control barrier functions. More recently, Da has expanded into quadrupedal locomotion, examining simulation-to-reality transfer through dynamics randomization and developing generalizable centroidal models. His hierarchical frameworks combining model-based control with reinforcement learning reflect a sophisticated understanding of practical deployment challenges. With over 700 cumulative citations, his work meaningfully advances the field's ability to produce reliable, adaptable legged robots.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3Dynamics Randomization Revisited: A Case Study for Quadrupedal Locomotion75 citations · 2021
- 4
- 5
- 6Dynamic Walking on Randomly-Varying Discrete Terrain with One-step Preview51 citations · 2017
- 7
- 8
- 9
- 10