Recording materials-related information along typical process chains e.g. for manufacturing steel and copper components, typically poses a severe data management challenge for domain experts. During the acquisition of data from the supplier, the manufacturing process(es) and material testing procedures, engineers and scientists face a complex household of files, database entries, sometimes even non-digital documents and other scattered data sources. Potential technologies fitting the special needs in manufacturing are rare, sometimes expensive, and in many cases not scalable nor interoperable enough to fit into the ecosystem of already established data processing tools and data sources. Domain experts without a special background in computer science tend to struggle under applications with advanced technologies, since the design of dynamic data spaces following FAIR principles and the provision of user-friendly interfaces at the same time are two major challenges which are often imbalanced. Moreover, domain experts with different skills have different requirements with respect to the interaction level, such as the management and definition of semantics, the registration of new entities and items, the upload of data, the exploration of data, reporting new actions in the process chain of manufacturing, etc.  

With the Dataspace Management System (DSMS) developed at Fraunhofer IWM, this contribution, we demonstrate how data provenance in manufacturing processes using semantic technologies can be orchestrated. The capabilities for data enrichment is demonstrated via the concept of Semantic Material Card. A Semantic Material Card is a material card based on a vocabulary and data schema which are represented via semantic technologies. A material card itself contain data for the material parameters of a material model, typically identified via experiments, to feed a simulations software as the basis for the numerical analyses. By coupling data pipelines, exploratory search mechanisms and additional Python-interfaces, experts with background in different disciplines are able to establish a documentation of the testing of the material, the analysis and the characterization, and the simulation through a software engine of choice. This is achieved using knowledge graphs in combination with classical relational databases, which can be administrated in different levels of the technology stack – independently of whether the expert is aware of ontologies, object-oriented programming or material science.