NVIDIA has unveiled a major push to drive America’s reindustrialization through the integration of physical AI, robotics, and digital twin technologies at GTC Washington, D.C. The company announced that leading U.S. manufacturers, industrial software developers, and robotics firms are leveraging NVIDIA Omniverse to build intelligent, autonomous factories and collaborative robots designed to address labor shortages and boost national manufacturing competitiveness. Jensen Huang, NVIDIA’s founder and CEO, emphasized that AI is transforming factories into intelligent systems, marking the dawn of a new industrial revolution. He highlighted that NVIDIA’s ecosystem—including Omniverse, Isaac, and AI infrastructure—is serving as the operating system for this era, enabling the design, simulation, and deployment of advanced manufacturing systems. A key development is the expansion of the NVIDIA Omniverse Mega Blueprint, now supporting the creation of large-scale digital twins of entire factories. Siemens is the first to deliver software for this blueprint, integrating it into its Xcelerator platform. This new technology stack combines high-fidelity 3D models with real-time operational data, allowing engineers to simulate, optimize, and monitor factory performance in real time. Robot manufacturers including FANUC and Foxconn Fii are adopting OpenUSD-based digital twins of their robots, enabling seamless integration into digital factory environments. Foxconn is using these tools to design and optimize its 242,287-square-foot facility in Houston, Texas, dedicated to manufacturing NVIDIA’s AI infrastructure. In 2025, over $1.2 trillion in U.S. manufacturing investments have been announced, driven by semiconductor, pharmaceutical, and electronics companies. These firms are using Omniverse-based applications and AI tools to accelerate production. Belden is using Accenture’s Physical AI Orchestrator—powered by Omniverse, Metropolis, and agentic AI—to create virtual safety zones and real-time quality inspection systems. Caterpillar is applying digital twins for predictive maintenance, dynamic scheduling, and supply chain optimization using NVIDIA NIM microservices and cuOpt. Lucid Motors uses Omniverse for real-time factory planning and AI robot training, while Toyota leverages idealworks’ iw.sim to simulate automation at its Kentucky plant. TSMC is using Omniverse to speed up fabrication plant design and deploying NVIDIA Isaac for robotics in its Arizona facility. Wistron is applying NVIDIA AI and Omniverse tools to rigorously test and validate systems in its Texas operations. In robotics, companies are building next-generation fleets using NVIDIA’s three-computer architecture. Figure Robotics is advancing humanoid robots with the Helix vision-language-action model and the Isaac platform, aiming to create versatile robots for industrial and domestic use. Agility Robotics’ Digit humanoid uses Isaac Lab for reinforcement learning to improve stability and adaptability in dynamic environments, powered by the Jetson AGX Thor module. Amazon Robotics has cut development time for its BlueJay manipulator from years to under a year using simulation. Skild AI is developing a general-purpose robotics foundation model across different robot types using Isaac Lab and NVIDIA Cosmos world models. FieldAI is training robots for inspection tasks in construction and oil and gas using synthetic data and software-in-the-loop validation. NVIDIA’s infrastructure is also enabling this transformation at scale. The IGX Thor platform, built on the Blackwell architecture, is being adopted by companies like Diligent Robotics and Hitachi Rail for edge AI applications. Google Cloud now offers G4 instances with RTX PRO 6000 Blackwell Server Edition GPUs, and Microsoft will soon provide these GPUs in Azure public cloud and at the edge via Azure Local. Together, these efforts are accelerating the digital transformation of American industry, positioning the U.S. at the forefront of the physical AI revolution.