NVIDIA developers demonstrated applications combining frontier AI models with NVIDIA Omniverse simulation libraries, according to an October 29 blog post. The projects show developers directing AI agents through natural-language instructions to build simulation workflows.
David Wehr from NVIDIA’s simulation technology team used GPT-6 Astra to develop a workflow for reconstructing building models from site scans. The work addresses a warehouse automation planning challenge: shelving and machinery can obscure building structure in scans. Wehr guided Astra to interpret panoramic photos, create building geometry and align it with point-cloud measurements. The workflow produces an editable, simulation-ready building model stored in OpenUSD format, with Omniverse’s ovrtx library rendering results for inspection. Developers can then add or reposition equipment to evaluate warehouse layouts before making physical changes.
Pradeep Kaushik from NVIDIA’s Omniverse developer relations team directed Astra to build a what-if analysis application around a simulated gas turbine and generator. The application lets users adjust operating conditions including load, ambient temperature and fuel composition, then compare power output, efficiency, fuel consumption and equipment response. Six test procedures explore scenarios including load changes, according to the post.
Omniverse libraries provide GPU-accelerated physics, rendering and sensor simulation capabilities that the AI agents access through developer-supplied instructions. NVIDIA published agent skills documentation to help agents work with Omniverse libraries.