Introduction to Multiphysics in Engineering: Multi-physics phenomena are central in many industrial applications and production lines, being directly connected to the efficiency and safety of the processes or the integrity of the produced parts. Embedded in process chains and engineering design tasks, multiphysics simulations successively become subject to tremendous automation and optimizations. This way, they are parts of Digital Twins, i.e., virtual, functional versions of systems that contain their properties and represent the life cycle of a system in the form of data and meta-data.
Using AI and ML in Engineering: However, this data must be combined and exploited correctly, where artificial intelligence (AI) and machine learning (ML) carry vast potential. For example, fast-running data-based surrogate models or automated evaluation of heterogeneous data and comparison/validation procedures, as well as clever calibration of uncertain simulation quantities, are possible through AI. For this, it is essential to be able to merge and reuse data of various kinds from various sources. Therefore, the digital twins can only be formed based on a solid data-interoperability backbone. The latter includes a machine-interpretable semantic meaning of the data and meta-data and it includes interfaces and data formats, as well as access control and the necessary processing software. Only when expertise and solutions from AI, semantics, data management, and engineering come together, the full potential of digitization and AI-enhanced digital twins can show advantage.
This joint laboratory project between Fraunhofer SCAI, the University of Applied Sciences Bonn-Rhein-Sieg, and the Dr. Reinold Hagen Foundation falls precisely in the center of this topic. It aims to create new digital-twin-related insights and turn them into benefits for small and medium-sized companies, which experience an exceptionally high digitalization pressure. These companies are often regional, e.g. experts in niche manufacturing techniques. It is particularly challenging for such companies to gather sufficient expertise in as many disciplines as required to these ends.
Therefore, the lab offers support to transform engineering and production into smart system developments and stay in leading positions using new methods. The many years of joint project experience of the lab partners bring together complementary capabilities, from theoretical, computational, to experimental knowledge. Central to the laboratory are new methods for integrating heterogeneous data, its use to form data-based machine learning models and hybrid data- and rule-based models, and the consequential improvements of the engineering processes.
R&D project partners: This project supports innovation in software research of Fraunhofer SCAI, engineering research and education of the HS BRS, and product development workflows of the Hagen Foundation. We explore customized multiphysics digital twins from all perspectives to integrate new aspects piece by piece and consequently mature concepts with less risk and on shorter time scales as in large industrial projects.
Students: Digital twins are likely to coin future engineering work environments in which many analysis systems, simulators, databases, and machines are interconnected. Professionals require new skillsets with good orientation on their specific tasks within larger, multi-disciplinary topics, knowing the difficulties and challenges of digital twins. Considering respective experience as a form of human intelligence, it is crucial to support education in this sector. Therefore, we continuously offer student positions and bachelor/master theses.
Head of Business Area Multiphysics
TREE Scientific Director
+49 2241 865-9678
Dr. Reinold Hagen Stiftung
Head of R&D Department
+49 228 9769-315