Papers
9
Total Citations
304
H-Index
6
About
Rongye Shi is a leading researcher at the intersection of autonomous systems and multi-agent reinforcement learning (MARL), with a focus on enabling intelligent, scalable coordination for robot teams and self-driving vehicles. His work bridges physics-based control and AI-guided policy learning, particularly in the challenging domain of mixed-autonomy traffic, where human-driven and autonomous vehicles share the road. His highly cited survey on autonomous vehicle control (224 citations) has become a key reference for researchers navigating this transition. Shi’s core contributions lie in advancing MARL through structural inductive biases: he has pioneered methods to exploit symmetry, partial symmetry, and hierarchical consensus in multi-agent systems, dramatically improving sample efficiency and generalization. His work on leveraging symmetry priors (ESP) and hierarchical symmetry has set new standards for data-efficient multi-robot cooperation. Shi also developed Air-M, a virtual reality platform for large-scale aerial swarm reinforcement learning, addressing the critical sim-to-real transfer gap. With recent publications in top venues and a growing citation impact, Shi is shaping the future of decentralized, scalable multi-agent intelligence for real-world robotic and autonomous systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Leveraging Partial Symmetry for Multi-Agent Reinforcement Learning15 citations · 2024
- 5ESP: Exploiting Symmetry Prior for Multi-Agent Reinforcement Learning10 citations · 2023
- 6
- 7
- 8
- 9Exploiting Hierarchical Symmetry in Multi-Agent Reinforcement Learning3 citations · 2024