Gaoyuan Liu
Papers
3
Total Citations
14
H-Index
1
About
Gaoyuan Liu is a robotics researcher whose work bridges the critical gap between high-level task planning and low-level physical interaction, with a focus on autonomous manipulation in complex, unstructured environments. His primary research areas include task and motion planning (TAMP), non-prehensile manipulation, and multi-agent relative localization. Liu’s most notable contribution is his work on integrating reinforcement learning with TAMP to enable robots to perform non-prehensile actions—such as pushing or sliding objects—alongside traditional grasping in cluttered settings. This synergistic approach, detailed in his 2023 paper (12 citations), offers a fast, generalizable solution for multi-modal manipulation, advancing the practicality of robots in real-world scenarios like warehouses and homes. He has also tackled the challenge of relative pose estimation for multi-robot teams using Ultra-Wideband (UWB) and visual-inertial odometry, proposing an unscented particle filter for robust performance even during unobservable motion. Most recently, Liu has applied his behavior planning expertise to agricultural robotics, developing automated pruning strategies for fruit trees using redundant manipulators. His work demonstrates a clear trajectory from foundational planning algorithms to impactful, real-world applications in agriculture and collaborative robotics.
Research Focus
Key Achievements
Top Papers
- 1
- 2
- 3