Installation.md

Environment Setup

Recommended Environment

• Ubuntu 20.04, python 3.8
• Ubuntu 22.04, python 3.10

1. Clone the repository

git clone https://github.com/uiuckimlab/PAPRLE.git

2. Create a virtual environment or use conda

# Create a virtual environment
virtualenv <env_name> -p python3.8
virutalenv <env_name>/bin/activate

or

# Create a conda environment
conda create -n <env_name> python=3.8 # or python=3.10
conda activate <env_name>

3. Install

pip install -e .

4. Usage

You can run the teleoperation system with the following command:

python run_teleop.py -f <robot_name> -l <leader_name> -e <env_name>

For example:

python run_teleop.py -f papras_kitchen -l sliders -e mujoco # Teleoperate PAPARAS Kitchen configuration with Sliders in Mujoco

python run_teleop.py -f g1 -l keyboard -e isaacgym # Teleoperate Unitree G1 with keyboard in Isaacgym

More options can be found with:

python run_teleop.py --help

Hardware

You can run your own robot if the robot is based on ROS1/ROS2.

• We also provide a GUI to write config file for the new follower setup.

If you are interested in using PAPRAS and Pluggable Puppeteer, please refer to:

• PAPRAS Interface: https://github.com/uiuckimlab/PAPRAS-V0-Public/

• PAPRAS Models + PAPRLE Leaders: https://github.com/uiuckimlab/PAPRLE_hw

Also, we utilize moveit for initializing and terminating in data collection process.

• Please install pymoveit2: https://github.com/AndrejOrsula/pymoveit2/tree/main
• Place this package in your ROS workspace, and source the workspace before running teleop code.

Simulator

Isaacgym (Only for Python 3.8)

• Pre-installation is needed: Download IsaacGym