New ESF Plus Young Researchers Group investigates transferable knowledge-based process models for flexible manufacturing

01.Mar.2024   Reseach News

The ongoing development of Industry 4.0 and artificial intelligence (AI) is opening up new perspectives for modern industrial production. The possibility of digital twins for model-based process optimization and control is within reach. However, existing AI approaches are reaching their limits, as they require enormous amounts of data and often, particularly in flexible manufacturing environments and with heterogeneous product portfolios, only small and heterogeneous data sets are produced.

The aim of the interdisciplinary ESF Plus young researchers group WiProFlex is therefore to develop so-called knowledge-based process models that use informed machine learning to skillfully enrich existing data with domain knowledge such as physics-based simulations and expert knowledge. This allows the development of robust and reliable process models even with small amounts of data.

The models will be demonstrated using processes from the highly complex production of microelectronics. The result of the project is a demonstrator of a knowledge-based process model for the reference process of chemical-mechanical planarization (CMP) as well as recommendations for action in order to be able to transfer the procedure to other manufacturing processes across industries.

The project coordination lies at the ZfM with Prof. Dr. Karla Hiller as spokesperson and Linda Jäckel as project leader. The further involved partners are the Professorship Smart Systems Integration (Prof. Dr. Harald Kuhn), the Professorship Scientific Computing (Prof. Dr. Martin Stoll), the Professorship Production Systems and Processes (Prof. Dr. Martin Dix), the Professorship of Distributed and Self-organizing Systems (Prof. Dr. Martin Gaedke) and the Professorship of Circuit and System Design (Prof. Dr. Ulrich Heinkel).

This project is funded by the European Social Fund (ESF) and the Free State of Saxony.