abizovnuralem/go2_omniverse

Unitree Go2, Unitree G1 support for Nvidia Isaac Lab (Isaac Gym / Isaac Sim)

What it solves

This project provides high-fidelity digital twins for the Unitree Go2 and G1 robots within the Nvidia Isaac Sim (Orbit/IsaacLab) environment. It allows researchers and developers to test robotics algorithms, simulate sensor data, and synchronize virtual robots with physical hardware without needing to risk expensive physical equipment in every test.

How it works

The project integrates Unitree robot models into Isaac Sim, leveraging the RTX renderer for realistic visuals and the ROS 2 bridge for communication. It features a "twinbot" mode that drives a simulated robot's joints in real-time using the /lowstate stream from a physical robot, ensuring the simulation mirrors the physical robot's movements via IMU-aligned kinematic playback. It also supports various sensor streams, including RTX Lidar, camera feeds, and foot force data, all routed through ROS 2.

Who it’s for

Robotics researchers and developers working with Unitree Go2 or G1 robots who need a high-fidelity simulation environment for algorithm development, RL training, and hardware-in-the-loop testing.

Highlights

  • Real-Robot Synchronization: Mirrors physical robot state into the simulation in real-time.
  • Cros-Platform Support: Compatible with both ROS 2 Humble (Ubuntu 22.04) and ROS 2 Jazzy (Ubuntu 24.04).
  • Comprehensive Sensor Simulation: Includes simulated RTX Lidar, camera streams, and foot force data.
  • RL Ready: Includes PPO balancing algorithms and various RL environments for custom training.
  • High-Fidelity Rendering: Uses studio HDRI lighting and RTX path-tracing for cinematic quality visuals.
  • VR Integration: Supports VR mode via SteamVR and ALVR for immersive robot monitoring.

Related

  • Project
    isaac-sim/IsaacLab

    A GPU-accelerated open-source framework built on NVIDIA Isaac Sim for simplifying robotics research in reinforcement learning, imitation learning, and motion planning.

  • Project
    gazebosim/gz-sim

    An open-source robotics simulator providing high-fidelity physics, rendering, and sensor models for testing and developing robotic systems.

  • Project
    umfieldrobotics/OceanSim

    A GPU-accelerated underwater robot perception simulation framework that uses physics-based rendering to model visual and acoustic sensors and reduce the simulation-to-real gap.

  • Project
    gazebosim/ros_gz

    A suite of integration packages that bridge ROS and the Gazebo simulator, enabling bidirectional communication and sensor data transport for robotics simulation.

  • Project
    NVIDIA-ISAAC-ROS/isaac_ros_visual_slam

    A ROS 2 package that provides GPU‑accelerated stereo visual‑inertial SLAM for NVIDIA Jetson (and x86_64) platforms, delivering high‑rate odometry and loop‑closed maps for robots and drones.