Mat-O-Lab describes material science human and machine readable, proving the use from laboratory to product.
Mat-O-Lab stands for the digitization of materials, parts and components along their entire life cycle. For that purpose, concrete practical and applicable solutions for pilot cases centered about specific materials are developed and provide for public use. Key results (datasets, ontologies, tools for data structuring and data analysis) from parallel research projects are consolidated and constantly improved. Mat-O-Lab also stands for a new, open and agile collaboration between the institutes of the Fraunhofer Group Materials and Components (MATERIALS), the Bundesanstalt für Materialforschung und -prüfung (BAM) and interested partners from industry and academia.
- We use the Basic Formal Ontology (BFO) as top level ontology (https://basic-formal-ontology.org/)
- We focus on using Common Core Ontology Stack https://github.com/CommonCoreOntology/CommonCoreOntologies, therefore we migrated the BWMD Ontology and provide it as Material Science an Engineering Ontology (MSEO) https://github.com/Mat-O-Lab/MSEO
- Tools and Ontologies created in Mat-O-Lab should be open source and publicly available
- We aim to create meaningful metadata for primary experiment data, therefore raw data in files (Ascii, Images, ....) is referenced by an URI but all metadata is converted to sematic data. It also includes information on the structure of the raw data file, to enable digestion of the raw data itself . By combining it with a method knowledge representation, a graph. The tabular data is untouched and will not be converted.
- Semantic data is to be shared in a decentral dataspace. Data sovereignty and security are primary aspects to be addressed in a practical solution. Mat-O-Lab considers a IDS based solution https://internationaldataspaces.org/ foremost.
- Method knowledge graphs must be constructed by domain experts; therefore, we must provide tools for them to do so. mapping between data and method graphs must be done by domain experts, inherent knowledge about the specific method is necessary to do so.!
We try to actively connect to
- PMD: https://www.materialdigital.de/
- IOF: https://www.industrialontologies.org/
- NFDI-MatWerk: https://nfdi-matwerk.de/
and all willing partners