Project runtime: 01.02.2021 – 30.06.2024
2023-09-22_Vollversammlung_SmaDi
2022-03-17_Vollversammlung_SmaDi
2023-09-22_Vollversammlung_Poster_SmaDi
2022-11-03_Vollversammlung_Poster_SmaDi
2024-09.18_Vollversammlung_Poster_SmaDi
2024-09-18_Vollversammlung_Demonstrator_Pitch_SmaDi
Material and associated technological innovations are a key driving force for innovative products. This applies in particular to so-called "smart materials". These have the distinct property of reacting to external influences (such as electrical, magnetic and thermal stimuli), e.g. through deformation. In the joint project, the subclasses of smart materials piezoelectric ceramics (PC), thermal and magnetic shape memory materials (SMA and MSMA) and dielectric elastomers (DE) were considered as examples. However, due to the complex material behavior, which depends on the manufacturing process, the targeted development of smart materials requires an exact description of the properties and the necessary manufacturing processes. The joint project focused on a digital representation of the materials and their manufacturing processes, which should enable an accelerated development of materials and process technologies on an industrial scale in order to gain advantages in international competition and take advantage of market opportunities. Ontological access to material parameters and transducer characteristics enables users to make a simplified comparison and supports their selection for their applications. From an information technology perspective, this requires the development of ontological artifacts with which access to material science findings, which are stored in data structures and models at different scale levels, can be designed uniformly and efficiently. In order to guarantee the practical usability of the ontological artifacts, they were developed on the basis of ontology-based data access (OBDA), which enables access to heterogeneous databases, among other things. This theoretically sound and industrially proven paradigm of information retrieval was extended in the research project by mathematical model equations for ontology-based data and model access (OBDMA).
At the beginning of the project, representative sample inquiries were first created in order to derive basic requirements for the OBDMA system. On this basis, an initial prototype system was developed, which was discussed with industrial partners in terms of strengths and weaknesses. For the further development, the expert knowledge on all material part classes that was developed in the network was prepared as data and mathematical models using the Markdown markup language, which served as an interface for the transfer of knowledge between materials science and computer science. In particular, relevant use cases were defined in this context. While the first half of the project focused on the development of an OBDMA system valid for all subclasses of smart materials, the second half of the project was concerned with integrating the material behavior dependent on the manufacturing process. Tandems were formed for the four subclasses of smart materials, each comprising one partner from the material sciences for material characterization and one from the engineering sciences for material modelling up to component level. The development from an information technology perspective was carried out by the University of Lübeck (UzL). In addition to the industrial partners for the subclasses of smart materials, the project was also supported by a cross-materials end user team.
Based on the existing OBDA paradigm, a so-called OBDMA system was developed for the smart materials, which enables combined access to both material data and transducer characteristics via models as user-defined functions (UDFs). It turned out that the necessary database size can be reduced and the data processing is easy to understand and flexible. With the OBDMA system, an efficient structure was established that can be easily supplemented and transferred to other materials. In order to develop a sustainable, modular solution that is compatible with other systems and platforms, proven W3C standards and open source software were used to describe the necessary ontological artifacts in which the material knowledge is mapped. The main requirements of the industrial partners were ease of use and the possibility of system validation. As the SPARQL language used for the query is hardly familiar to users, a simple user interface was created in addition to the application. For system validation, a test option was developed using the SHACL language. This involves formulating conditions that are used to check the system. The project results were presented at various conferences and published in journals. The developed OBDMA system is freely accessible as a demonstrator via GitHub.
