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 5 taining machine-readable information from tribal/artisan knowledge sources be significantly reduced. • Ensuring that data is reusable and accessible requires that the generator of data be incentivized to collect, organize, and curate it in a manner suitable for reuse. Note, the data generator has all the burden and little to gain while the data user has none of the burden with everything to gain. • The principal impediment to data accessibility is the political, economic, and social resistance to data sharing. This is especially true for proprietary, confidential, and classified data. • Proprietary data formats impede the findability and accessibility of data. • The interoperability, reusability, and findability of data is hindered because of the lack of common data/metadata formats, metadata definitions, and missing data. PATH FORWARD Based upon the results of the work- shop, the critical elements of a strategic path forward, and toward a realization of Materials 4.0, were identified. See Fig. 2. Overcoming the PEST challenges ards development work. Dr. Brandon Ribic, America Makes, and Doug Hall, Battelle Memorial Institute, discussed the AM data management strategy of their non-profit organizations. Participants were divided into four working groups aligned to the FAIR Data Management Principles, i.e., 1. Findable, 2. Accessible, 3. Interoperable, and 4. Reusable. Each working group was co-led by NIST and industry personnel. Working groups identified and prioritized the challenges associated with achieving the workshop goals and recommended approaches to overcome those challenges. An abbreviated summary of the salient observations of the workshop are as follows: • There are political, economic, social, and technological (PEST) impediments to effective data management. The technological challenges were viewed as tractable; the political, economic, and social challenges will require cross-agency and government/private sector collaborative efforts. • Data scientists spend 80% of their time finding, filtering, reformatting, and integrating data, leaving only 20% of their time for data analysis. This 80/20 ratio (time to prepare data/time using data) needs to be changed to 20/80. • The current work to establish an AM common data dictionary (CDD) was highlighted. • The need for a common data exchange format (CDEF) for AM was validated. • Continued work is required to build and expand community consensus around the FAIR principles. • Utilization of WWW consortium standards was an implicit theme. Similarly, the use of JavaScript object notation (JSON), representational state transfer API (REST API), and web ontology language (OWL2) was implicit to the participants’ thinking. • Establishing a “Data Commons” for the facile, cost effective exchange of AM data, models, and tools deserves to be explored. The workshop identified challenges associated with implementing FAIR principles, which were: • Enhanced interoperability requires the establishment of a common data model and formal knowledge representation. • The ability to find and access data requires that the manual labor, difficulty, and cost associated with obFig. 2 — Critical elements of an additive manufacturing data management plan: A PEST problem.
RkJQdWJsaXNoZXIy MTYyMzk3NQ==