Han Ruihua (韩瑞华)

I am currently a Postdoctoral Researcher in the School of Computing and Data Science at The University of Hong Kong (HKU), supervised by Prof. Hengshuang Zhao. I received my Ph.D. in Computer Science from The University of Hong Kong (HKU), where I was supervised by Prof. Jia Pan and Prof. Qi Hao. Before that, I obtained my M.Eng. in Mechanical Systems from Xiamen University (XMU) and my B.Eng. in Mechatronic Engineering from Wuhan University of Technology (WHUT). I also worked as a Research Assistant in Computer Science at the Southern University of Science and Technology (SUSTech).

My research focused on the intersection of motion planning, robot learning, embodied AI, and optimal control. My work aims to advance the ability of robot systems to navigate and operate safely and intelligently in complex, real-world settings. I integrate model-based optimization theory with data scale driven learning based method for robotic motion and control to achieve both theoretically guaranteed and intelligent enough assist human centric tasks.

Beyond my research, I actively contribute to the robotics community by developing and sharing open-source research projects. My projects have been widely used in both academia and industry, collectively receiving over 2.9K stars on GitHub. Representative repositories include NeuPAN Planner (T-RO 2025), RDA Planner (RA-L 2023), RL-RVO-NAV (RA-L 2022), and IR-SIM (Rank #1 2D robotics simulator)

Research Interests: Robot Learning; Autonomous Navigation; Embodied AI; Motion Planning; Reinforcement Learning; Multi-Robot Systems; Field Robotics; Foundation Models for Robotics

Research Skills: Python, C++, ROS, Gazebo, CARLA, PyTorch, MATLAB, SolidWorks, LaTeX, Open-source Development, Real-world Robot Deployment

Selected Publications

T-RO

NeuPAN: Direct Point Robot Navigation with End-to-End Model-based Learning

Ruihua Han , Shuai Wang , Shuaijun Wang , Zeqing Zhang , Jianjun Chen , Shijie Lin , Chengyang Li , Chengzhong Xu , Yonina C Eldar , Qi Hao , and Jia Pan

IEEE Transactions on Robotics (Popular Paper), 2025

Bib PDF video Website Code

@article{han2024neupan,
  title = {NeuPAN: Direct Point Robot Navigation with End-to-End Model-based Learning},
  author = {Han, Ruihua and Wang, Shuai and Wang, Shuaijun and Zhang, Zeqing and Chen, Jianjun and Lin, Shijie and Li, Chengyang and Xu, Chengzhong and Eldar, Yonina C and Hao, Qi and Pan, Jia},
  journal = {IEEE Transactions on Robotics (Popular Paper)},
  year = {2025},
}

T-MECH

Edge Accelerated Robot Navigation With Collaborative Motion Planning

Guoliang Li* , Ruihua Han* , Shuai Wang , Fei Gao , Yonina C Eldar , and Chengzhong Xu

IEEE/ASME Transactions on Mechatronics, 2024

Bib PDF Code

@article{li2024edge,
  title = {Edge Accelerated Robot Navigation With Collaborative Motion Planning},
  author = {Li*, Guoliang and Han*, Ruihua and Wang, Shuai and Gao, Fei and Eldar, Yonina C and Xu, Chengzhong},
  journal = {IEEE/ASME Transactions on Mechatronics},
  year = {2024},
  publisher = {IEEE},
}

RA-L

Rda: An accelerated collision free motion planner for autonomous navigation in cluttered environments

Ruihua Han , Shuai Wang , Shuaijun Wang , Zeqing Zhang , Qianru Zhang , Yonina C Eldar , Qi Hao , and Jia Pan

IEEE Robotics and Automation Letters (RA-L), 2023

Bib PDF video Code

@article{han2023rda,
  title = {Rda: An accelerated collision free motion planner for autonomous navigation in cluttered environments},
  author = {Han, Ruihua and Wang, Shuai and Wang, Shuaijun and Zhang, Zeqing and Zhang, Qianru and Eldar, Yonina C and Hao, Qi and Pan, Jia},
  journal = {IEEE Robotics and Automation Letters (RA-L)},
  volume = {8},
  number = {3},
  pages = {1715--1722},
  year = {2023},
  publisher = {IEEE},
}

RA-L

Reinforcement learned distributed multi-robot navigation with reciprocal velocity obstacle shaped rewards

Ruihua Han , Shengduo Chen , Shuaijun Wang , Zeqing Zhang , Rui Gao , Qi Hao , and Jia Pan

IEEE Robotics and Automation Letters, 2022

Bib PDF video Code

@article{han2022reinforcement,
  title = {Reinforcement learned distributed multi-robot navigation with reciprocal velocity obstacle shaped rewards},
  author = {Han, Ruihua and Chen, Shengduo and Wang, Shuaijun and Zhang, Zeqing and Gao, Rui and Hao, Qi and Pan, Jia},
  journal = {IEEE Robotics and Automation Letters},
  volume = {7},
  number = {3},
  pages = {5896--5903},
  year = {2022},
  publisher = {IEEE},
}

ICRA

Cooperative multi-robot navigation in dynamic environment with deep reinforcement learning

Ruihua Han , Shengduo Chen , and Qi Hao

In 2020 IEEE International Conference on Robotics and Automation (ICRA) , 2020

Bib PDF video

@inproceedings{han2020cooperative,
  title = {Cooperative multi-robot navigation in dynamic environment with deep reinforcement learning},
  author = {Han, Ruihua and Chen, Shengduo and Hao, Qi},
  booktitle = {2020 IEEE International Conference on Robotics and Automation (ICRA)},
  pages = {448--454},
  year = {2020},
  organization = {IEEE},
}

IROS

A distributed range-only collision avoidance approach for low-cost large-scale multi-robot systems

Ruihua Han , Shengduo Chen , and Qi Hao

In 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2020

Bib PDF video

@inproceedings{han2020distributed,
  title = {A distributed range-only collision avoidance approach for low-cost large-scale multi-robot systems},
  author = {Han, Ruihua and Chen, Shengduo and Hao, Qi},
  booktitle = {2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages = {8020--8026},
  year = {2020},
  organization = {IEEE},
}

Open Source Projects:

IR-SIM 🥇 Rank #1 in 2D robotics simulators: A lightweight, extensible Python simulator supporting diverse robot kinematics (differential, Ackermann, omnidirectional), customizable sensors (LiDAR, odometry), and YAML-based configuration. Designed for rapid prototyping of motion planning and reinforcement learning algorithms. Adopted by HKU (COMP3356 Robotics) and SUSTech (Intelligent Robot Course) for teaching.
neupan_ros : ROS deployment of the NeuPAN planner—a neural-augmented MPC framework that unifies learning-based prediction with optimization-based control. Features Gazebo integration with ready-to-use navigation and imitation learning demos.
rda_ros : ROS wrapper for the RDA planner—an ADMM-based MPC enabling parallel collision avoidance computation for arbitrary convex shapes. Supports multiple kinematics and includes demos for Gazebo (dynamic obstacles) and CARLA (autonomous driving).
rvo_ros : ROS plugin implementing the ORCA velocity obstacle algorithm for multi-agent collision avoidance. Provides a Gazebo plugin for decentralized coordination in multi-robot systems.

Service

Reviewer: IEEE Transactions on Robotics (T-RO); IEEE Robotics and Automation Letters (RA-L); IEEE International Conference on Robotics and Automation (ICRA); IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS); American Control Conference (ACC).