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 3 We are at the dawn of Materials 4.0, a critical component of the digitally driven, data enabled epoch of research and manufacturing, Industry 4.0[1-4]. In this new era, every aspect of a product’s life cycle (research, development, engineering, design, manufacturing, deployment, sustainment, and sunsetting) is interconnected and interdependent. Concomitantly, digitally intense manufacturing technologies of significant importance have emerged. Additive manufacturing (AM) is one such manufacturing technology area that has demonstrated its potential to enhance innovation, accelerate product deployment, and reduce cost. AM is not a single technology, but instead refers to several layer-by-layer processes that fall within its scope. These layer-by-layer processes are com- monly defined as a “a process of joining materials to make parts from 3D model data, usually layer upon layer, as opposed to subtractive manufacturing and formative manufacturing methodologies”[5]. The flexibility offered by AM gives it a certain advantage over many traditional processes, such as the potential to produce components where and when they are needed. Part-defining technical data packages can be sent electronically to any global manufacturing site. Long lead time, complex components, such as aircraft forgings and castings, can be produced in days or weeks compared to months or years. However, the Achilles’ heel of this vision remains the inability to rapidly and cost effectively qualify AM processes and certify components. The traditional means of process qualification involves optimizing the materials technology process, “freezing” it, and then developing statistically substantiated design allowables[6]. Unfortunately, AM does not lend itself to thismethodology as key process parameters may be part-specific and a function of material, geometry, orientation, and proprietary processing. Because of the multitude of factors affecting part quality, the qualification process is time consuming and very costly. A great deal of effort has been devoted to overcoming this challenge. AM standards are Makes/DoD, Crystal City, Va., June 18, 2019). • Additive Manufacturing for Maintenance Operations (AMMO), (America Makes/DoD, Crystal City, Va., June 23, 2020). A fundamental tenet emerged that the challenge could be addressed by building upon the FAIR principles of data management[8]. Simply stated, AM data must be captured, curated, and managed in a form that allows the characteristics of Findability, Accessibility, Interoperability, and Reusability, i.e., FAIR. Details of these principles may be found at https://www.go-fair.org/ fair-principles/ and are summarized in Table 1[9]. THE COST OF NOT BEING FAIR The EuropeanUnionhas published compelling evidence of the impact on the research community of not having FAIR data[10]. They estimated the annual cost of not having FAIR data to be a minimum of €10.2bn per year. An additional cost to innovation was estimated at €16.9bn per year. Further, they stated, “The actual cost is likely to be much higher due to unquantifiable elements such as the value of improved research quality and other indirect positive spillover effects of FAIR research data.” They went on to conclude that about 80% of the duplicative funded work could be avoided with FAIR. Furthermore, the need for FAIR data extends beyond the research community in AM, to practitioners and developers as well. As Fig. 1 indicates, those who work primarily with data spend 80% of their time finding, filtering, reformatting, and integrating data[11,12]. FAIR AM DATA MANAGEMENT WORKSHOP Under the backdrop of this community alignment, the FAIR Additive Manufacturing (AM) Data Management Workshop was held virtually on October 27-28, 2020[13]. The workshop was organized and executed by NIST, ASM International, and Pilgrim Consulting LLC. A workshop assessment[14], speaker briefs, agenda, etc., may be found being developed. Integrated computational materials engineering (ICME) tools are being developed. Modeling and simulations tools, as well as artificial intelligence (e.g., machine learning, neural networks, etc.) are being employed, and new testing methodologies are being adopted. Application of these physics-based and data analytical tools requires the capture, transformation, curation, and analysis of data from across the product’s life cycle. Further, given the complexity of the processes, vast amounts of data are required to achieve any correlations of significance. Few have the necessary resources to attain the required amount of data. To reduce cost, time, and duplicative work, government, academic, and corporate organizations must be able to share data easily across organizational lines. Unfortunately, there are many obstacles to the facilitation of data sharing. Today most data are stored in a range of diverse formats, e.g., paper files, PDFs, spreadsheets, relational databases, etc. They are stored in an equally diverse set of containers including in engineers’ desk drawers, desktop computers, product life cycle management systems, and in the cloud. FAIR AM DATA MANAGEMENT PRINCIPLES Emerging from a series of additive manufacturing workshops held by the U.S. Navy, National Institute of Standards and Technology (NIST), America Makes, and U.S. Department of Defense (DoD) has come a community consensus for a transformational shift in AM process and part qualification and the imperative to adopt FAIR AM data management practices[7]. The workshops were held at these locations and dates: • Navy Additive Manufacturing Technology Interchange (NAMTI) 2018, (Navy, Quantico, Va., November 27, 2018). • AM Materials Database and Data Analytics Workshop, (NIST, Gaithersburg, Md., May 7, 2019). • Additive Manufacturing for Maintenance Operations (AMMO), (America
RkJQdWJsaXNoZXIy MTYyMzk3NQ==