Haoxiang Liu

State Key Laboratory of Industrial Control Technology, The University of Tokyo

Papers

Total Citations

H-Index

About

Haoxiang Liu is a robotics researcher whose work focuses on enabling robots to operate effectively in challenging, unstructured environments. His primary research areas include spherical robotics, robot teleoperation, and reinforcement learning for human-robot interaction. Liu’s most notable contribution is the development of a novel spherical robot designed for exploration in harsh, unknown environments. This robot, driven by a heavy pendulum and enclosed in a fully sealed spherical shell, is amphibious, highly protected, and capable of traversing rugged terrain—a significant advancement for search-and-rescue and planetary exploration. His most-cited paper on this topic (2023, 14 citations) has already established a foundation for robust, all-terrain robotic platforms. Beyond hardware, Liu has made impactful strides in improving robot teleoperation. He pioneered a reinforcement learning-based method for selecting optimal camera viewpoints, enabling operators to perform faster and more accurate remote control of robot arms. His 2022 and 2023 works on viewpoint selection (totaling 7 citations) address a critical bottleneck in teleoperation: providing operators with the best visual perspective without requiring extensive subject experiments. Liu’s research elegantly bridges mechanical design and intelligent control, offering practical solutions for deploying robots in inaccessible or hazardous spaces.