NVIDIA announced updates to its Isaac platform for robotics, enabling to move from and into real-world robot skills. The system combines NVIDIA Isaac Lab for training, Newton for physics simulation, and Isaac Lab-Arena for evaluation, all runnable on NVIDIA DGX Station hardware.
The platform includes Agile ONE, a reference humanoid robot framework built on NVIDIA Isaac GR00T and powered by NVIDIA Jetson Thor chips. The workflow lets developers build training worlds, replay and learn from real-world motion, and tune physics using NVIDIA open libraries before deploying skills to physical robots.
This is an AI-written summary. The original reporting belongs to @NVIDIA. How we write these · Spot a mistake?