Skip to content

Latest commit

 

History

History
99 lines (82 loc) · 2.95 KB

README.md

File metadata and controls

99 lines (82 loc) · 2.95 KB

Policy Deployment for Go1 Robots

Ubuntu 20.04 ROS Noetic Python

  • This repository provides deployment code for Go1 policies trained in IsaacLab, such as LocoTouch.
  • The codebase is structured to support easy customization and minor modifications for new policies.

Table of Contents

  1. Installation
  2. Go1 Policy Deployment

Installation

Basic Installation

  1. Create a conda environment with python 3.10: 
    cd Go1-Policy-Deployment
conda create -n policy_deploy python=3.10
  2. Install the dependencies: 
    conda activate policy_deploy
pip install -e .
  3. Install ROS Neotic (we only test the code on Ubuntu 20.04).

Go1 SDK Installation

  1. Download the SDK: 
    git clone https://github.com/unitreerobotics/unitree_legged_sdk.git
  2. Make sure the required packages are installed, following Unitree's guide. Most notably, please make sure to install Boost and LCM: 
    sudo apt install libboost-all-dev liblcm-dev
pip install empy catkin_pkg
  3. Then, go to the unitree_legged_sdk directory and build the libraries: 
    cd unitree_legged_sdk
mkdir build && cd build
cmake -DPYTHON_BUILD=TRUE ..
make
cd ../../

Go1 Policy Deployment

For onboard setup, please refer to MacMini-for-Onboard-Robotics

Common Threads:

  1. Launch ROS: 
    roscore
  2. Start the joystick interface (see locomotion_cfg.py for configuration details): 
    python teleoperation/joystick.py

Locomotion:

python deploy/locomotion.py

State-Based Object Transport (Teacher Policy):

  1. Run the OptiTrack motion capture system: 
    python mocap/run_optitrack.py
  2. Run the teacher policy: 
    python deploy/transport_teacher.py

Tactile-Aware Object Transport (Student Policy):

  1. Run the tactile sensor: 
    python tactile_sensing/run_tactile_sensing.py
  2. Run the student policy: 
    python deploy/transport_student.py

Reference

@article{lin2025locotouch,
  title={LocoTouch: Learning Dynamic Quadrupedal Transport with Tactile Sensing},
  author={Lin, Changyi and Song, Yuxin Ray and Huo, Boda and Yu, Mingyang and Wang, Yikai and Liu, Shiqi and Yang, Yuxiang and Yu, Wenhao and Zhang, Tingnan and Tan, Jie and others},
  journal={arXiv preprint arXiv:2505.23175},
  year={2025}
}