Papers
16
Total Citations
135
H-Index
7
About
Zhaoyuan Gu is a robotics researcher whose work sits at the intersection of bipedal locomotion, motion planning, and machine learning, with a particular focus on making legged robots more capable, robust, and intelligent in real-world environments. His most influential contribution proposes a hierarchically integrated task and motion planning framework for bipedal robots navigating partially observable environments with dynamic obstacles, earning 29 citations and establishing him as a leading voice in safe autonomous locomotion. Gu has pioneered the application of signal temporal logic within model predictive control for bipedal systems — a first in the field — enabling robots to satisfy formal task guarantees while recovering from external perturbations. His early work on Adversarial A3C (20 citations) demonstrates a deep engagement with robust reinforcement learning, while more recent contributions span inverse reinforcement learning from demonstrations, whole-body humanoid loco-manipulation, and variable-stiffness robotic arms for safe human-robot interaction. His 2025 survey on humanoid locomotion and manipulation reflects his growing role as a synthesizer of the field. Collectively, Gu's research advances a vision of humanoid robots that are formally verifiable, physically resilient, and learning-capable — qualities essential for deploying legged robots beyond laboratory settings.
Research Focus
Key Achievements
Top Papers
- 1
- 2Adversary A3C for Robust Reinforcement Learning20 citations · 2019
- 3
- 4
- 5
- 6
- 7
- 8
- 9
- 10