Ontologies for FAIR Data in Additive Manufacturing: A Use Case-Based Evaluation

Abstract

The development of an ontology-based approach for generating Findable, Accessible, Interoperable, Reusable (FAIR) data for powder bed fusion, a representative additive manufacturing process, is explored. Addressing key aspects of part design, parameter selection, and processing history, the study identifies both the advantages and disadvantages of using ontologies to manage and utilize distributed and heterogeneous data from additive manufacturing effectively. Critical to this approach is the establishment of unique digital and physical identifiers for physical objects, which facilitate the creation of digital object records and enhance data findability, crucial for enabling digital twins. Despite the benefits of increased findability and domain expandability, challenges persist, such as the complexity of integrating diverse data sources and the high demand for specialized knowledge to navigate ontology-based systems, discussed by incorporating the basic formal ontology. The study also explores data integration techniques using Python, the application of reasoning to reduce manual input, and the implications on reusability. The research demonstrates the potential of FAIR data to transform additive manufacturing processes by enabling more efficient data utilization. Applications such as material property and process parameter selection, as well as the creation of digital part records, serve as exemplary implementations showcasing the practical benefits of this approach.