Xiaopeng Fan
Papers
2
Total Citations
8
H-Index
2
About
Xiaopeng Fan is a robotics researcher advancing the frontiers of autonomous navigation and multi-agent coordination. His work centers on two critical challenges: intelligent scheduling for multi-robot systems and robust place recognition for LiDAR-based localization. In his 2022 paper on multi-AGV scheduling, Fan introduced a hierarchical multi-agent reinforcement learning framework with intrinsic rewards, enabling fleets of automated guided vehicles to efficiently coordinate material delivery in warehouses and flexible manufacturing systems. This approach addresses the complex combinatorial optimization problem of collision-free, deadlock-free scheduling in dynamic environments. More recently, Fan developed CCTNet, a circular convolutional transformer network that dramatically improves LiDAR-based place recognition under occlusion from moving objects. By replacing standard single-column convolutions with a circular design, his network maintains feature invariance to column shifts in range images—a persistent challenge in SLAM loop closure detection and re-localization. With 5 and 3 citations respectively on these foundational works, Fan’s contributions are gaining traction in both industrial automation and robotics communities. His research bridges reinforcement learning and computer vision, offering practical solutions for real-world autonomous systems operating in cluttered, dynamic spaces.
Research Focus
Key Achievements
Top Papers
- 1
- 2