NVIDIA Expands Cosmos Data Factory for Robotics and Physical AI
NVIDIA expanded Cosmos with five cloud-native data blueprints and new tools for training and testing world models used in robotics and industrial AI.
NVIDIA expanded its Cosmos platform October 21 with five cloud-native Data Factory blueprints for autonomous-vehicle datasets, image attributes, event video, defect imagery and video augmentation. The tooling is available through GitHub and cloud partners including Microsoft Azure, Nebius and CoreWeave.
The platform targets the data bottleneck in physical AI: labeling and curating multisensory inputs, generating synthetic edge cases and evaluating world models before deployment. NVIDIA cited Porsche, ANYbotics and Ultralytics as early users; performance improvements and test reductions in the announcement are vendor or customer claims.
For infrastructure operators, the update links cloud compute with factories, warehouses, vehicles and inspection robots. Adoption still requires validated site data, model governance and real-world safety testing; synthetic results do not replace field verification.
Official source · Development date: October 21, 2026; source reviewed October 22, 2026.