Data Augmentation for Accelerating Materials and Processing Development (DAAMP’D)

Many of the radical advances in the field of materials science can be directly linked to the impressive advances in digital materials characterization, physics-based simulation, and instrumented manufacturing processes. The field of materials science is currently making a transition from a data limited field to a data driven one and has not yet developed standards for the collection, processing, and archiving of these data and metadata. Current explosive growth of machine learning (ML) and artificial intelligence (AI) offers opportunities for confronting the complexity of additive manufacturing (AM) and high entropy alloys (HEA) material development. Machine learning requires access to data in efficient and standardized formats, as well as a record of the data provenance in the form of associated metadata. The proposed effort establishes a framework that receives the myriad data sources pertaining to the AM and HEA technologies areas and transforms the data to facilitate state-of-the-art analytics via ML/AI. The framework also addresses interoperation with remote and cloud-based data stores, performing analysis queries for improved materials and processes development. The framework demonstrates learning improved processes for AM, and learning new thermodynamic models for developing new HEA.