Isaac ROS Physical AI#
Overview#
Isaac ROS Physical AI provides application-level bringup packages for deploying whole-body control and teleoperation on humanoid robots.
The current release supports the Unitree G1 with:
AGILE locomotion: Whole-body control policy for balance and walking
Bimanual inverse kinematics: Real-time IK for both arms using cuMotion
Finger control: Hand tracking from an XR headset
XR teleoperation: End-effector and locomotion control using an XR headset with Isaac Teleop
Both MuJoCo simulation and real hardware are supported.
For a fine-tuned GR00T policy deployed end-to-end — teleoperate, record, convert to LeRobot, fine-tune GR00T N, export via LEAPP, deploy — see the Isaac for Physical AI reference workflow.
Packages#
Supported Platforms#
This package is designed and tested to be compatible with ROS 2 Jazzy running on Jetson or an x86_64 system with an NVIDIA GPU.
Platform |
Hardware |
Software |
Storage |
Notes |
|---|---|---|---|---|
Jetson |
128+ GB NVMe SSD |
Simulation and real robot deployment both tested and qualified. For best performance, ensure that power settings are configured appropriately. |
||
x86_64 |
|
32+ GB disk space available |
Isaac ROS Environment#
To simplify development, we strongly recommend leveraging the Isaac ROS CLI by following these steps. This streamlines your development environment setup with the correct versions of dependencies on all supported platforms.
Note
All Isaac ROS Quickstarts, tutorials, and examples have been designed with the Isaac ROS CLI-managed environment as a prerequisite.
Customize your Dev Environment#
To customize your development environment, refer to this guide.
Updates#
Date |
Changes |
|---|---|
2026-07-06 |
Support CloudXR runtime without Docker |
2026-04-30 |
Initial release |