The Humanoid Robotics Lab in the Department of Computer Science at University College London explores how robots can move, act, and interact with the world as fluidly and adaptively as humans. Our research focuses on humanoid and legged robots, whole-body control, teleoperation, and embodied AI — combining control theory, reinforcement learning, and bio-inspired design to create systems that walk, balance, manipulate, and express themselves in complex environments.

Directed by Dr Chengxu Zhou, Associate Professor in Robotics and AI at UCL, the lab builds on over a decade of experience in humanoid robotics research across Europe and the UK. We are a collaborative, curiosity-driven group working at the intersection of engineering, AI, and human–robot interaction. Our projects range from advancing the fundamentals of robot locomotion and loco-manipulation to exploring expressive, human-centred movement in humanoids.

selected publications

Nat. Mach. Intell.

Learning to Adapt through Bio-Inspired Gait Strategies for Versatile Quadruped Locomotion

Joseph Humphreys and Chengxu Zhou

Nature Machine Intelligence, 2025

DOI Bib PDF Video Code

@article{humphreys2025nmi,
  title = {Learning to Adapt through Bio-Inspired Gait Strategies for Versatile Quadruped Locomotion},
  author = {Humphreys, Joseph and Zhou, Chengxu},
  journal = {Nature Machine Intelligence},
  volume = {7},
  pages = {1141–-1153},
  year = {2025},
  doi = {10.1038/s42256-025-01065-z},
}

Humanoids

Gait-Conditioned Reinforcement Learning with Multi-Phase Curriculum for Humanoid Locomotion

Tianhu Peng, Lingfan Bao, and Chengxu Zhou

In IEEE-RAS International Conference on Humanoid Robots, 2025

Bib PDF Video Website

@inproceedings{peng2025humanoids,
  title = {Gait-Conditioned Reinforcement Learning with Multi-Phase Curriculum for Humanoid Locomotion},
  author = {Peng, Tianhu and Bao, Lingfan and Zhou, Chengxu},
  booktitle = {IEEE-RAS International Conference on Humanoid Robots},
  year = {2025},
}