An IsaacLab DirectEnv for performing teleoperation with Unitree Go2 + AgileX Piper, with scripts for sim-to-sim and sim-to-real.
Features:
- Locomotion policy able to adjust pose and carry a manipulator
- Manipulation controller using whole-body Inverse Kinematics with reduced model using mink + feedback linearization
- 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 in ROS2 compatible with our public low-level robot's hal for Go2 unitree_ros2_dls and Agilex Piper arm piper-ros2-dls
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.
This work takes a lot of inspiration from our repo trash-collection-isaaclab. If you find it useful, please consider citing:
BinWalker: Development and Field Evaluation of a Quadruped Manipulator Platform for Sustainable Litter Collection
@article{turrisi26littercollection,
author = {Giulio Turrisi and Angelo Bratta and Giovanni Minelli and Gabriel Fischer Abati and Amir H. Rad and João Carlos Virgolino Soares and Claudio Semini},
title = {BinWalker: Development and Field Evaluation of a Quadruped Manipulator Platform for Sustainable Litter Collection},
journal = {arXiv},
year = {2026}
}
PRs are very welcome (search for TODO in the issue, or add what you like)!
This repository is maintained by Giulio Turrisi.


