For industrial decision-makers, the question is less whether to use digital twins and more how consistently they can be integrated into existing organisational, data and decision-making structures. Those who master this step will lay the foundation for more resilient, flexible and economical production systems.

The strategic advancement of the digital twin concept presented by Siemens and Nvidia goes far beyond classic 3D models and engineering simulations. At its core is a new, AI-supported approach in which digital twins not only map, but also actively prepare, evaluate and secure industrial decisions. For industry, the announcement marks an important step towards digitally controlled production and operating models.

Large-scale industrial metaverse environments

Siemens presented the Digital Twin Composer, a new software within the Xcelerator portfolio, at CES. The Composer brings together product, production and process twins in a shared, photorealistic 3D environment and connects them with engineering, automation and operational data. Technologically, the solution is based on NVIDIA Omniverse libraries, but remains deeply integrated into the Siemens world of PLM, TIA, simulation and automation.

Permanent, consistent reference to the real system

The key difference to previous approaches lies in the claim: the digital twin is no longer understood as a project-related model, but as a permanent, consistent reference to the real system. ‘From the most comprehensive digital twin and AI-supported hardware to co-pilots in manufacturing, we are scaling intelligence in the physical world,’ announced Roland Busch, CEO of Siemens AG, in Las Vegas, predicting that this unique technological change will lead to measurable results: ‘In this way, companies can achieve speed, quality and efficiency at the same time.’ In concrete terms, this will mean that changes to layouts, material flows, cycle times or automation concepts can be virtually tested, evaluated and prioritised before they are physically implemented. Siemens speaks of a significant reduction in unexpected problems and measurable effects on throughput, investment security and time-to-change.

Nvidia provides physics-based simulation layer for industrial applications

For Nvidia, the partnership with Siemens is a central component of its own ‘Physical AI’ strategy. With Omniverse, the chip manufacturer is positioning itself not only as a hardware and AI platform provider, but also as a physics-based simulation layer for industrial applications. In combination with Siemens software, this creates a continuous model from engineering to operation – including AI agents that can automatically generate, evaluate and optimise scenarios.

Vision of an ‘industrial AI operating system’

Both companies explicitly refer to the vision of an ‘industrial AI operating system,’ i.e., a platform on which industrial processes are not only visualised but also actively controlled, optimised and automated. For industrial users, this means a new quality of data-based decisions – provided they also have issues such as data model governance, IT/OT integration and cybersecurity under control.

Digital twins on the way to becoming an industrial standard

The momentum surrounding Siemens and Nvidia is part of a broader market trend. In recent weeks, several companies have demonstrated that digital twins are increasingly penetrating the operational core of industrial value creation. WuXi Biologics presented PatroLab, a digital twin platform for the bioprocess industry that more closely integrates development and production phases and reduces risks during scale-up. In the process industry, Omega Simulation's OmegaLand V4.1 relies on dynamic digital twins that are directly linked to control systems, enabling closed-loop optimisation.

Digital twins are even being used in highly complex fields

The topic is also becoming more broadly anchored strategically: Dassault Systèmes is investing in a new Centre of Excellence for Virtual Twin and AI, while digital twins are even being used in highly complex fields such as fusion research and pharmaceutical series production to shorten development cycles and reduce investment risks. These latest developments show that the digital twin is being promoted from an engineering tool to a central element of industrial control.