The “Skipro” research project entered its operational phase with a kick-off meeting on July 7 and 8. The project is coordinated by the Chair of Information, Quality, and Sensor Systems in Production (WZL-IQS) at the Machine Tool Laboratory (WZL) of RWTH Aachen University under the direction of Prof. Robert H. Schmitt.

The consortium, comprising research institutions, industrial companies, and metrology experts, is developing software technologies designed to make heterogeneous production data usable for generative AI systems. The goal is to preserve process context, traceability, and industrial reliability requirements.

Linking Production Data from Different Sources

In production environments, information is often scattered across machine controllers, automation systems, measurement and testing equipment, engineering tools, as well as quality documents and the employees’ practical knowledge.

According to the project partners, this fragmentation makes it difficult to deploy generative AI applications such as large language models (LLMs) or multi-agent systems, as the necessary process context is often missing.

Middleware and Multi-Agent Systems

"Skipro" is developing a software-based integration platform for this purpose. The platform will feature a cross-tier information architecture for data exchange, standardized middleware for the harmonization and semantic processing of heterogeneous data, and generative multi-agent workflows for the analysis and processing of complex production tasks.

“With ‘SKIPRO,’ we’re bridging the gap between visionary AI research and the real-world production floor,” explains Lucky Adam, project coordinator at the Chair of Information, Quality, and Sensor Systems in Production (WZL-IQS). “Our goal is not to build yet another isolated AI prototype, but to provide reliable, user-centered software tools that translate heterogeneous production data into a format suitable for industrial use while preserving the full process context.”

Research and Industry Involved

During the kick-off meeting, the project partners defined the work packages and initial milestones. The project results will then be tested in industrial applications.

In addition to the WZL-IQS at RWTH Aachen University, the consortium includes the Physikalisch-Technische Bundesanstalt (PTB), the Federal Institute for Materials Research and Testing (BAM), the Jülich Research Center, and the companies eurogard, INC Innovation Center, Peak Solution, NuCOS, and Siemens. As associated industry partners, Johnson Electric Aachen and Miltenyi Biotec are participating in the validation of the developed solutions.