D. Abeyrathne
Advanced Structures & Composites Center, Maine, United States
Keywords: Data warehouse, Digital thread, Data Analysis, ML, Data traceability, Auditability
The Material Process Property Warehouse (MPPW) provides a process-agnostic, provenance-rich data infrastructure designed to unify and streamline advanced manufacturing workflows. Developed within the context of large-area additive manufacturing but adaptable to multiple processes, including subtractive, hybrid, casting, and compounding, the MPPW enables the capture, organization, and retrieval of diverse datasets such as CAD designs, sensor streams, operator logs, process parameters, test results, and material properties. This flexibility makes it a critical foundation for research and production environments where projects demand custom operation types, heterogeneous data integration, and traceable digital threads. By structuring data around projects, manufacturing operations, and material artifacts, the MPPW preserves geometric, process, and material-property relationships while providing secure, role-based access and auditability. Beyond storage, the system integrates a machine learning pipeline that leverages these datasets for advanced inferencing, predictive analytics, and adaptive feedback into manufacturing workflows. This capability is central to achieving “born-certified” parts, where traceability and performance validation are embedded throughout the lifecycle rather than post-process. Together, the MPPW’s process-agnostic design, AI/ML integration, and digital thread alignment position it as a robust framework that supports human-AI collaboration, accelerates data-driven discovery, and establishes a scalable pathway toward next-generation, adaptable, and certifiable manufacturing systems.