Virtual Inspection of Advanced Manufacturing via Process-Scale Digital Twins (Abbreviated Report)
Inspection and certification comprise the most significant bottlenecks in advanced manufacturing for NNSA applications, often requiring far more time and resources than the fabrication of the parts themselves. Traditional methods, such as manual review and X-ray computed tomography, are not only slow and costly, but also struggle to provide a clear connection between manufacturing instructions and the final performance of critical components. This gap limits both the agility and assurance needed to support the modernization and safety of the United States nuclear stockpile. In response, our Strategic Initiative established a digital twin framework that integrates realtime process monitoring, automated data analysis, and immersive virtual reality collaboration into a unified inspection pipeline. By leveraging data from sensors, machine instructions, and imaging, we created high-fidelity virtual models of manufactured parts that could be rapidly analyzed and certified. This approach was first demonstrated with Direct Ink Write, and then extended to other manufacturing settings, including conventional (or “subtractive”) manufacturing and to predict the end of life performance of parts per the aging and lifetimes programs. The result is a transformational capability: inspection times have been reduced by a factor of 120,000 without loss of accuracy and while simultaneously improving traceability and confidence in part quality. This framework not only streamlines certification for critical applications, but also positions the national security enterprise to respond more flexibly to emerging challenges, supporting agile manufacturing and digital engineering practices across a broad range of mission-relevant domains.