Physical AI is advancing from research labs to real-world applications like autonomous vehicles. OpenUSD’s Core Specification 1.0 standardizes data types and formats to scale autonomous systems. NVIDIA Omniverse libraries, powered by OpenUSD, create digital twins for simulation and testing. Advancements in synthetic data generation and multimodal datasets are converging with the NVIDIA Halos framework for AV safety. Bosch, Nuro, and Wayve are participating in the NVIDIA Halos AI Systems Inspection Lab to accelerate the deployment of robotaxi fleets. Mcity at the University of Michigan is enhancing its AV test facility with NVIDIA Omniverse technologies for safe testing of driving scenarios.

Read more at NVIDIA: OpenUSD and NVIDIA Halos Accelerate Safety for Robotaxis, Physical AI Systems