Papers
7
Total Citations
113
H-Index
5
About
Shunyi Yao is a robotics and artificial intelligence researcher whose work centers on autonomous robot navigation, multi-agent collision avoidance, and human-robot interaction in crowded environments. His most significant contributions lie in applying deep reinforcement learning (DRL) to enable robots to navigate safely and efficiently through complex, real-world scenarios without relying on inter-robot communication. Yao's most-cited work (42 citations) tackles the challenge of crowd-aware robot navigation by developing map-based DRL frameworks that account for pedestrians employing diverse collision avoidance strategies — moving beyond the limiting assumption that all humans follow a single behavioral model like ORCA or the social force model. His complementary research on distributed multi-robot systems (35 citations) extends this approach to heterogeneous robot fleets operating in communication-free environments, a practically important constraint for real-world deployment. Beyond collision avoidance, Yao has explored the social dimensions of robot navigation, investigating how robots can use sound and interaction cues to engage pedestrians and improve navigational outcomes. His more recent work addresses robustness by training adversarial non-cooperative agents to expose vulnerabilities in navigation policies. Across roughly seven publications accumulating over 110 citations, Yao has established himself as a thoughtful contributor to safe, socially aware autonomous robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3
- 4Multi-Robot Collision Avoidance with Map-based Deep Reinforcement Learning10 citations · 2020
- 5
- 6
- 7