This work aims to push the limits of agility for bipedal robots
by enabling a torque-controlled bipedal robot to perform robust
and versatile dynamic jumps in the real world. We present a
multi-task reinforcement learning framework to train the robot
to accomplish a large variety of jumping tasks, such as jumping
to different locations and directions. To improve performance
on these challenging tasks, we develop a new policy structure
that encodes the robot's long-term input/output (I/O) history
while also providing direct access to its short-term I/O
history. In order to train a versatile multi-task policy, we
utilize a multi-stage training scheme that includes different
training stages for different objectives. After multi-stage
training, the multi-task policy can be directly transferred
to Cassie, a physical bipedal robot. Training on different
tasks and exploring more diverse scenarios leads to highly
robust policies that can exploit the diverse set of learned
skills to recover from perturbations or poor landings during
real-world deployment. Such robustness in the proposed
multi-task policy enables Cassie to succeed in completing a
variety of challenging jump tasks in the real world, such as
standing long jumps, jumping onto elevated platforms, and
multi-axis jumps.
@article{
CassieJumpLi2023,
title={Robust and Versatile Bipedal Jumping Control through Multi-Task Reinforcement Learning},
author={Zhongyu Li and Xue Bin Peng and Pieter Abbeel and Sergey Levine and Glen Berseth and Koushil Sreenath},
year={2023},
eprint={2302.09450},
archivePrefix={arXiv},
primaryClass={cs.RO}
}
