Papers
4
Total Citations
133
H-Index
3
About
Hongyi Zhou is a leading researcher in intelligent robotics, with a focus on multi-agent navigation, path planning, and imitation learning. His most impactful contribution is the MAPPER framework (113 citations), a decentralized multi-agent path planning method that combines evolutionary reinforcement learning to enable large-scale robot fleets to navigate safely in mixed dynamic environments—a critical advance for industrial deployment. Zhou also developed HIRO, a heuristics-informed online path planning system that separates static and dynamic environmental elements to efficiently compute collision-free trajectories for robots in real-time. His recent work, MaIL, introduces a novel imitation learning architecture that leverages Mamba state-space models as a computationally efficient alternative to Transformer-based policies, achieving selective attention to key data features. This innovation addresses scalability challenges in learning from demonstration. Zhou’s research consistently bridges theoretical advances with practical robotics applications, from warehouse automation to autonomous navigation, making him a key figure in the evolution of intelligent, adaptive robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4MaIL: Improving Imitation Learning with Mamba3 citations · 2024