Community-Driven Development of Infrastructure Use Cases in NFDI-MatWerk

Digital transformation in Materials Science and Engineering (MSE) hinges on scalable, interoperable, and community-driven research data infrastructures. To achieve this, NFDI-MatWerk introduces Infrastructure Use Cases (IUCs). These IUCs stem from a requirement analysis of the MSE community where key challenges were identified: harmonization of data formats, documentation of contextual information, interoperable software tools, and implementation of FAIR principles. The IUCs serve as research-relevant profiles for the development and validation of a sustainable research data infrastructure. They are continuously refined through an iterative process involving community feedback, ensuring the infrastructure remains aligned with evolving user requirements. We present the results of four IUCs, selected for their community relevance and broad applicability across multiple research scenarios. We have developed demonstrators that showcase research data workflows. These demonstrators serve multiple purposes: validate the technical feasibility of our infrastructure, offer concrete examples of FAIR data principles in practice, and support knowledge transfer across the community by illustrating best practices in data harmonization and evaluation. - IUC02: Framework for curation and distribution of reference datasets. The use case describes reference data for creep testing, a method to determine time-dependent deformation of a material under constant load at elevated temperatures. The data must meet high standards of completeness, documentation, and measurement precision. In our demonstrator, these requirements are addressed through metadata schema, ontology, and SHACL shapes for validation, specifying mandatory entries and data types. The goal is to populate a local database with validated, searchable data compatible with other databases. - IUC04: Data Integration for the construction of Defect Phase Diagrams. The aim of the IUC and its key participant project CRC-1394 is to extend bulk phase diagrams to defect phase diagrams for predicting material properties. This requires managing large volumes of experimental and computational data. We showcase how data from diverse sources can be seamlessly collected using tools built around openBIS, creating a cohesive research data management ecosystem. Relevant metadata were identified through a community effort and are currently being extended. Automated metadata extraction supports data sharing across a large interdisciplinary collaboration. - IUC09: Reproducible data analyses workflows for atom probe tomography (APT). APT is widely used for nanoscale structure and composition characterization in materials science, geosciences, and biological sciences. Standardized workflows for analysis and post-processing are essential to enable interoperability across software tools from different research communities. In collaboration with FAIRmat, we demonstrate reproducible APT workflows that integrate different analysis tools. This is a key step toward automating FAIR data production and advancing reproducible scientific analysis in the field. - IUC17: Ontologies for crystal defects and integration with simulation workflows. In materials science, studying crystalline structures and defects is crucial for understanding material properties. However, simulation and experimental data often lack standardized annotation, making it difficult to compare, share, and reuse. The goal is to develop ontologies for crystallographic defects that can be applied consistently across length scales. The demonstrator focuses on atomistic simulations. It addresses the challenge of inconsistent data annotation through the development of atomRDF, an ontology-based metadata annotation tool. By semantically structuring simulation data, this approach improves data interoperability and reusability. These demonstrators not only address community-identified needs, but also promote shared understanding of digital workflows in MSE. Our iterative development process aims to ensure long-term alignment with scientific practices while also enabling future scalability of the research data infrastructure.