Haoru Xue
Papers
2
Total Citations
18
H-Index
2
About
Haoru Xue is pioneering the frontier of agile, whole-body robot control, with a focus on legged locomotion and universal mobility. His key research areas include sampling-based model predictive control (MPC) and learning-based dynamics models for high-performance robotics. Xue’s major contribution is the development of "Full-Order Sampling-Based MPC for Torque-Level Locomotion Control via Diffusion-Style Annealing," a breakthrough that tackles the long-standing challenge of real-time optimal control using full-order, non-convex dynamics for legged robots—a feat previously limited to reduced-order models. This work has already garnered 12 citations since its 2025 publication, signaling its rapid impact. Additionally, his paper "AnyCar to Anywhere: To Anywhere: Learning Universal Dynamics Model for Agile and Adaptive Mobility" (6 citations) introduces a generalist model capable of controlling diverse robot embodiments for agile tasks, pushing beyond static navigation into dynamic performance. Xue’s work is notable for bridging the gap between theoretical optimal control and practical, torque-level execution, promising more adaptive and resilient robots. His innovative use of diffusion-style annealing in MPC marks a significant step toward truly autonomous, high-speed locomotion in complex environments.
Research Focus
Key Achievements
Top Papers
- 1
- 2