Digitization of Existing Plants AI Should Automatically Create Digital Twin

Related Vendors

Fraunhofer-Institut für Fabrikbetrieb und -automatisierung IFF

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Industrial plants of the past are moving into the future: With the KI-sy Twin project, Fraunhofer Institutes, together with industry partners, are developing new ways to make analog machines digitally usable.

In Germany, many industrial plants have been operating for decades. Their components function reliably, but there is often a standstill when it comes to digitalization. Cable clutter, evolved IT structures, and lack of networking hinder modern process control. This is exactly where the research project KI-sy Twin comes in.

Retrofit Networking Instead of Rebuilding

Most machines in the German industry are legacy systems, known as brown field plants. Their digitalization is considered complex, expensive, and time-consuming. Consequently, many companies forgo the transition. The KI-sy Twin project takes a different approach: existing components such as pumps or valves are to be digitally integrated without major modifications. The basis for this is the Asset Administration Shell (AAS), a standardized digital representation of physical objects.

In the project, the Fraunhofer Institutes IFF and IOSB-INA are cooperating with over a dozen industry partners, including ABB, Bayer, Festo, Siemens, and ZVEI. The goal is to develop simple tools that allow companies to digitize their existing plants without deep IT expertise.

From Nameplate to Administration Shell

A concrete component is a software that scans the nameplates of machines. Instead of manually capturing device information, a photo suffices. The software extracts the data, links it with technical documents, and automatically generates a passive asset administration shell. By specifying interfaces, this is further developed into an active AAS. This can be embedded into a data structure, which in turn is equipped with an EDC connector, an interface for secure data exchange in digital ecosystems.

The Fraunhofer IFF provides a demonstrator for the project, which allows the automated generation of the AAS. This enables the integration of more complex plant structures such as hierarchies and process flow diagrams.

More Transparency, Less Effort

The project partners rely on interoperability through standardization. Many existing plants consist of components from various manufacturers that operate according to different standards. The administration shell aims to break these isolated solutions. The goal is a seamless data space that enables condition monitoring, predictive maintenance, and transparent plant documentation.

In the long term, companies could not only work more efficiently but also deploy personnel more strategically. Instead of investing time in manually maintaining documentation, skilled workers could focus on maintenance and optimization. The KI-sy Twin project runs until the end of November 2025 and is supported by the Fraunhofer Big Data and AI Alliance as well as the Cluster of Excellence Cognitive Internet Technologies. (mc)

Machine Authentication

Machine Operator Authentication - Why knowing "who" matters