A D V A N C E D M A T E R I A L S & P R O C E S S E S | J U L Y / A U G U S T 2 0 2 1 1 6 is viewed as central to success. Quantifying the value proposition is required to overcome the activation energy associated with concerns (e.g., IP rights, ROI, security, etc.), and provides a business rationale for investment in FAIR data management. Standards are needed to facilitate interoperability. These must include a common data dictionary, a domain specific AM ontology, and metadata formats. The foundational elements include many data standards and architectures developed by the World Wide Web Consortium (W3C). In addition, the aspirational goal of the Semantic Web must be embraced, i.e., to format data in such a way as to make it both machine and human readable. There are critical imperatives associated with implementing effective FAIR data management principles and being productive in Industry 4.0. The work before us will require cross disciplinary efforts. The knowledge, skills, and abilities of the materials scientist and engineermust expand to include that of data science. Further, there is a need for an organization, perhaps a public-private consortium, to serve as a focal point for prototyping data management technologies and serving as the steward for persistent data identifiers. EXAMPLE DATA MANAGEMENT CONCEPTS Technology Stack: When considering the development of a viable data management system, it is perhaps useful to view it in terms of a technology stack, Fig. 3. This permits a systems-level perspective. In this table, the FAIR principles, Semantic Web, the World Wide Web (WWW), etc., are viewed as the foundation upon which an effective data management structure must be built. The applications and platforms listed at the top of the table illustrate how the data might be used by the consumer. The wide breadth of applications underscores the need for the use of FAIR data management principles and a well-defined knowledge management architecture. AM generates large amounts of data, and effective knowledge man- agement is crucial to the advancement of AM. A well-structured and community accepted domain ontology supports knowledge sharing, i.e., interoperability. A pragmatist may characterize an ontology as a model constructed to describe reality that consists of a taxonomy, lexicon, concepts, and defines interrelationships. Here, the goal of an ontology is to support knowledge sharing and the electronic management of scientific information. R. Arp et al. state, “Ontology is a top-down approach to the problem of electronically managing scientific information,” and go on to say “Definitions are perhaps the most important component of ontologies, since it is through definitions that an ontology draws its ability to support consistent use across multiple communities and disciplines, and to support computational reasoning”[15]. Hence, the hard work of establishing components (i.e., dictionary, thesaurus, taxonomy, and ontology) of effective AM data management is critical to achieving innovation and accelerated product deployment[16,17]. The means of data curation and management is also very important. Data must be extracted, transformed, and loaded into the curation site in such a way as to make it usable to data consumers. Several common architectures are considered data repositories. Whether the data is physically curated at a “brick and mortar” location or in the cloud, its purpose and functionality is largely determined by political, social, and economic consideration. For example, a data commons typically supports precompetitive R&D and provides the community with access to both data and computational tools. A data warehouse, however, is typically employed by corporations to deliver specific, actionable business information. Data hubs have characteristics that may make them well suited for agile team formation, secure data curation, and collaboration. However, the sustainability of any of these repositories requires a viable long termbusiness model. One study[18] of 48 repositories in 18 countries assessing their business case concluded, “Yet good data stewardship is costly and research budgets are limited. So, the development of sustainable business models for research data repositories needs to be a high priority in all countries.” EWI AM Qualification: EWI and partners have shown proof of concept that data-enabled approaches can change the approach to qualification Fig. 3 — FAIR AM data management technology stack.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==