Installation

Installation#

Prerequisites#

  • Python 3.10+

  • Git with submodule support

  • (Optional) CUDA-compatible GPU for model evaluation

  • (Optional) Oculus Quest for VR teleoperation

System Requirements#

Operating Systems:

  • Linux (Ubuntu 20.04+ recommended)

  • macOS (with some limitations on rendering)

  • Windows (with WSL2)

Dependencies:

  • MuJoCo 3.1.5

  • Gymnasium

  • NumPy 1.26.x

  • PyTorch (for model evaluation)

Quick Installation#

  1. Clone the repository with submodules:

git clone --recurse-submodules https://github.com/helen9975/RoboEval.git
cd RoboEval

If you already cloned without submodules:

git submodule update --init --recursive

  1. Create and activate conda environment:

conda create -n roboeval python=3.10 -y
conda activate roboeval

  1. Install the package:

Choose the installation option that fits your needs:

# Core package only
pip install -e .

# With example scripts (recommended for first-time users)
pip install -e ".[examples]"

# With VR teleoperation support
pip install -e ".[vr]"

# Development installation (includes testing and linting tools)
pip install -e ".[dev]"

# All features
pip install -e ".[examples,vr,dev]"

Installation Options Explained#

Core Installation (pip install -e .)

Includes only the essential dependencies to run environments and replay demonstrations.

Examples (.[examples])

Adds dependencies for running example scripts including model evaluation (OpenVLA, ACT, etc.)

VR Support (.[vr])

Adds PyOpenXR and related dependencies for Oculus Quest teleoperation.

Development (.[dev])

Includes pytest, black, flake8, and other development tools.

Verify Installation#

Test your installation by running a simple demo replay:

python examples/1_data_replay.py

This script will:

  1. Download demonstration data (on first run)

  2. Initialize the environment

  3. Replay recorded demonstrations

You should see a MuJoCo viewer window showing a bimanual robot replaying a task.

Troubleshooting#

MuJoCo rendering issues

If you encounter OpenGL or rendering errors:

# Ubuntu/Debian
sudo apt-get install libgl1-mesa-dev libegl1-mesa-dev

# Also try setting:
export MUJOCO_GL=egl

Missing system libraries

# Ubuntu/Debian
sudo apt-get install build-essential libglfw3 libglfw3-dev

CUDA/GPU issues

For model evaluation, ensure CUDA is properly installed:

# Check CUDA availability
python -c "import torch; print(torch.cuda.is_available())"

Submodule issues

If you see import errors related to mujoco_menagerie or mojo:

git submodule update --init --recursive

Next Steps#

  • Follow the Quick Start guide to run your first environment

  • Explore Examples to learn about data collection and evaluation

  • Read about Environments to understand task variations

Contents