Zekai Sun
Papers
3
Total Citations
12
H-Index
2
About
Zekai Sun is a pioneering researcher at the intersection of robotics, autonomous navigation, and robust machine learning. His primary research areas include air-ground robot systems, deep reinforcement learning (DRL) for robotic control, and distributed training systems for the Internet of Robotic Things. Sun’s most impactful contribution is **AGRNav**, a novel autonomous navigation framework for air-ground robots operating in occlusion-prone environments like forests and large buildings. This work, which has garnered 8 citations since 2024, addresses the critical challenge of navigating unknown, obstructed spaces by enabling energy-efficient and safe traversal without pre-mapped obstacles. In parallel, Sun has advanced the reliability of DRL in real-world settings through his work on **state-aware perturbation optimization**, a method to harden robotic control policies against environmental disturbances. His earlier research on **ROG**, a high-performance distributed training system for robotic IoT teams, laid the groundwork for scalable machine learning in disaster response scenarios. By tackling the fundamental tension between robotic autonomy and real-world uncertainty, Zekai Sun is shaping the future of resilient, field-deployable robotic systems.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3