Blog

The latest from Google Research

Exploring Nature-Inspired Robot Agility

Friday, April 3, 2020

Posted by Xue Bin (Jason) Peng, Student Researcher and Sehoon Ha, Research Scientist, Robotics at Googlereinforcement learningLearning Agile Robotic Locomotion Skills by Imitating AnimalsLearning Agile Robotic Locomotion Skills by Imitating AnimalsMotion Imitation

Reinforcement learning is used to train a simulated robot to imitate the reference motions from a dog. All simulations are performed using PyBullet.

a priori

Comparison of policies before and after adaptation on the real robot. Before adaptation, the robot is prone to falling. But after adaptation, the policies are able to more consistently execute the desired skills.

Results

Robot imitating various skills from a dog.

Skills learned by imitating artist-animated keyframe motions: side-steps, turn, and hop-turn.

Learning to Walk in the Real World with Minimal Human EffortQT-Opt

An automated learning system for legged robots must resolve safety and automation challenges.

Learning to Walk in the Real World with Minimal Human Effort

We solve automation and safety challenges with multi-task learning, a safety-constrained SAC algorithm, and an automatic reset controller.

Resultsany human intervention

Snapshots of the training process on the flat surface with zero human resets.

We train locomotion policies to walk in four directions, which allow us to interactively control the robot with a game controller.

Learned locomotion gaits on challenging terrains.

Conclusionskills from videosAcknowledgmentsWe would like to thank our coauthors, Erwin Coumans, Tingnan Zhang, Tsang-Wei Lee, Jie Tan, Sergey Levine, Peng Xu and Zhenyu Tan. We would also like to thank Julian Ibarz, Byron David, Thinh Nguyen, Gus Kouretas, Krista Reymann, and Bonny Ho for their support and contributions to this work.