Xiaofeng Mao

Pennsylvania State University, University of Edinburgh, Peking University

Papers

Total Citations

H-Index

About

Xiaofeng Mao is a pioneering roboticist whose research spans multi-robot coordination and dexterous manipulation, bridging the gap between swarm intelligence and tactile-driven autonomy. His early work on multi-robot hunting in dynamic environments introduced the BCSLA approach, enabling mobile robots to intelligently capture evaders through state-based strategies like dispersion-random-search and prediction—a foundational contribution to cooperative robotics. More recently, Mao has revolutionized imitation learning for robotic hands. His framework, DexSkills, leverages haptic data to segment and reuse skills for long-horizon tasks, dramatically reducing the need for extensive training data. In parallel, his work on fine pinch-grasp skills demonstrates how rich tactile sensing from just a few real-world examples can achieve bimanual dexterity, a breakthrough for practical robot learning. With over 17 citations across his most-cited works, Mao’s impact is evident in both foundational multi-robot simulation platforms (CASIA-MR) and cutting-edge tactile manipulation. His research not only advances autonomous systems but also makes robotic learning more accessible, offering a blueprint for data-efficient, skill-based control in complex environments.