NVIDIA Launches Open-Source Physical AI Tools for Robotics and Digital Twins on June 1

NVIDIA unveiled a suite of open-source physical AI tools and agent skills on June 1, designed to simplify development of robotics, autonomous vehicles, computer vision, and industrial digital twin applications. The initiative reduces development complexity by enabling AI agents to access NVIDIA's software libraries, models, and frameworks for tasks including data generation, simulation, training, evaluation, and deployment.

The tools support multiple industries: robotics developers can automate synthetic data creation and robot deployment; autonomous vehicle teams can generate driving scenarios and expand training coverage; vision AI applications can leverage automated data labeling and model optimization; industrial software developers can streamline digital twin creation. Companies including TSMC, Foxconn, Li Auto, and Cadence have already adopted NVIDIA's physical AI technologies. The tools are now available through open repositories and integrated with cloud providers including Microsoft and CoreWeave.