Papers
11
Total Citations
233
H-Index
10
About
Kaiyu Zheng is a robotics and AI researcher whose work spans robot navigation, probabilistic reasoning, and autonomous object search — areas that sit at the intersection of practical robotics engineering and principled decision-making under uncertainty. He is perhaps best known for his widely adopted "ROS Navigation Tuning Guide," which has accumulated over 80 citations and serves as an essential reference for roboticists worldwide seeking to optimize mobile robot navigation performance. This work reflects his dedication to making complex robotic systems more accessible and deployable in real-world settings. Beyond navigation, Zheng has made significant contributions to semantic mapping through frameworks like TopoNets and Graph-Structured Sum-Product Networks (GraphSPNs), enabling robots to build structured, probabilistic representations of large-scale environments. His research on multi-object search under uncertainty, particularly through POMDP-based planning in 3D environments, addresses one of robotics' most practically important challenges — helping robots efficiently locate objects in cluttered, real-world spaces. His open-source tool, **pomdp_py**, has further broadened the research community's ability to prototype and solve POMDP problems. More recently, his work incorporates spatial language understanding, allowing robots to interpret human descriptions of object locations, pushing toward more natural human-robot interaction.
Research Focus
Key Achievements
Top Papers
- 1ROS Navigation Tuning Guide83 citations · 2021
- 2ROS Navigation Tuning Guide23 citations · 2017
- 3
- 4Multi-Resolution POMDP Planning for Multi-Object Search in 3D21 citations · 2021
- 5
- 6Towards Optimal Correlational Object Search14 citations · 2022
- 7pomdp_py: A Framework to Build and Solve POMDP Problems13 citations · 2020
- 8A System for Generalized 3D Multi-Object Search12 citations · 2023
- 9
- 10