NVIDIA FOX Blueprint Orchestrates Специализированные агенты промышленного ИИ

Manufacturers moving from isolated automation to plant-wide intelligence need AI systems that can connect live machine signals, quality systems, and operational alerts into a single decision layer. The NVIDIA Factory Operations Blueprint (FOX) provides a reference design for building an autonomous factory manager agent that continuously monitors and reasons across real-time data to orchestrate a fleet of specialized industrial AI agents.

Connecting Factory Systems and Specialized Agents

FOX integrates with industrial data sources, machines, applications, and robot fleets, connecting specialized agents from leading software developers through standard APIs and agent skills. Built with NVIDIA NemoClaw, AI-Q Blueprint, and NVIDIA Nemotron open models, the blueprint provides a customizable foundation for automating model development and running intelligent operations at scale. The system is optimized to run on the NVIDIA DGX Station, which features the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, delivering 20 petaflops of FP4 performance and 748GB of coherent memory.

Real-World Deployment and Efficiency Gains

Foxconn is using the FOX blueprint and NemoClaw to build MoMClaw, a manufacturing operations multi-agent system that connects sensors and machine signals with hundreds of specialized agents. Foxconn projects an 80% improvement in root cause analysis time, a 15% increase in labor productivity, and a 10% decrease in machine failure rates. Pegatron is utilizing the blueprint to orchestrate robot utilization more efficiently, with an estimated 15% reduction in asset redundancy costs.

Specialized Agents for Quality and Safety

Beyond orchestration, specialized agents are being developed for specific tasks like visual inspection and process compliance. Spingence is using NVIDIA Cosmos and the TAO Toolkit to develop a factory manager agent for Cooler Master that achieves 99.6% defect recall and reduces defect escapes by 78%. Overview AI is using synthetic defect data to deploy visual inspection AI models 12x faster, reducing time to first inference to under 30 minutes across more than 300 products. These deployments demonstrate the potential for autonomous factory management to significantly improve first-pass yield and reduce production waste.

Source: NVIDIA Factory Operations Blueprint Gives Factories a New AI Brain
Domain: blogs.nvidia.com