Reinforcement learning implementation of the quadruped get-up task in IsaacLab. It includes different robots, with scripts for sim-to-sim and sim-to-real transfer.
Features:
- Concurrent State Estimator
- Rapid Motor Adaptation
- Identification of robot parameters for sim2real using pace via our repo sim2real-robot-identification
- Sim-to-Sim in Mujoco
- Sim-to-Real using ROS2
Real-world deployment via:
- unitree-ros2-dls for unitree robot communication
A list of robots and environments available is described below:
| Robot Model | Environment Name Pattern |
|---|---|
| Aliengo, Go2, B2 | GetUp-RobotModel-Flat-Blind GetUp-RobotModel-Rough-Blind |
If you want only to deploy a trained policy on your robot, continue on README_DEPLOY otherwise on README_TRAIN.
For the train, check first the compatibility with IsaacLab and rsl-rl at the top of this readme. They indicate the releases that we tested.
PRs are very welcome (search for TODO in the issue, or add what you like)!
This repository is maintained by Giulio Turrisi.