Demonstration and Transfer Network AI in Production (ProKI-Netz)

Key Info

Basic Information

01.11.2022 to 31.10.2024
Organizational Unit:
Chair of Manufacturing Technology, Grinding Technology
Federal Ministry of Education and Research (BMBF)

Research partner

    • Chairs of Laboratory for Machine Tools and Production Engineering (WZL) of RWTH Aachen University
    • Chair for AI Methodology (AIM) of RWTH Aachen University
    • Institute for Data Science in Mechanical Engineering (DSME) of RWTH Aachen University
    • Institute of Industrial Engineering and Ergonomics (IAW) of RWTH Aachen University
    • Surface Engineering Institute (IOT) of RWTH Aachen University



Eike Reuter

Research Assistant


+49 241 80 25388



Artificial intelligence (AI) algorithms are increasingly finding their way into industrial production. The potential for optimizing processes and business models along the entire value chain, from logistics to manufacturing and assembly processes, has already been recognized. Nevertheless, small and medium-sized enterprises (SMEs) in particular shy away from the (supposedly) high effort and the investment risk, or do not have the human resources with the necessary expertise.

The Demonstration and Transfer Network AI in Production explores new possibilities for applications of artificial intelligence in manufacturing and specifically promotes the implementation of applications in industrial practice. In external cylindrical grinding, the acquisition of high-frequency sensor and machine control data and their evaluation by means of time series analyses and machine learning allows the prediction of process result variables and the identification of disturbance variables. This enables a reduction of rejects and opens up the potential for semi-automated process control.

Within the project, the external cylindrical grinding process is used as part of a manufacturing sequence for the production of gear shafts in order to uncover potential for assistance systems in grinding technology and to reveal cross-process correlations by creating a digital twin.