The companies say they began co-developing and testing the applications in live production after announcing their factory collaboration at CES in January 2026.
For printed-circuit-board assemblies, Siemens uses digital twins and NVIDIA Cosmos world models to generate photorealistic training data for visual quality inspection before physical prototypes are available. Siemens says virtual production preparation can reduce engineering effort by up to 85% and time to market by up to 40%.
The factory also uses an AI agent for production planning. It generates scenarios when orders arrive, presents their effects against defined KPIs, and sends proposals for validation in a digital twin of the facility. Siemens says this workflow can enable cost savings of up to 20%. The Erlangen site has about 1,000 employees, more than 100 AI algorithms in use, and a planned EUR500 million transformation investment through 2030.
The release does not provide an independent evaluation, a calculation methodology for the claimed savings, product pricing, or general customer availability details.