ORNL and ARC launch AI Foundry to compress defense part qualification from years to months

Oak Ridge National Laboratory and New York-based startup Autonomous Resource Corporation (ARC) have signed a Memorandum of Understanding to create the Exascale Foundry, a closed-loop system combining AI-enabled manufacturing qualification with autonomous production for U.S. national security applications. ARC will deploy advanced manufacturing equipment across seven production nodes linked via its secure ARCNet platform, while ORNL contributes high-performance computing, its Peregrine AI software (which has analyzed over 1.9 million AM layers), and technologies from its Manufacturing Demonstration Facility. The initial focus is metal binder jetting of high-temperature nickel superalloy turbine components for autonomous air vehicle engines, targeting a critical defense supply chain bottleneck. ARC CEO Bryan Wisk stated the goal is compressing manufacturing and qualification timelines from years to months.

This partnership directly addresses the aerospace qualification grind pattern, where proving a part often takes longer than printing it. By embedding ORNL's Peregrine AI for real-time adaptive control and quality assurance directly into ARC's production nodes, the Exascale Foundry aims to collapse the qualification cycle that has historically kept AM from scaling in defense and aerospace serial production. The deal also plugs into DOE's Genesis Mission, a national program to build the world's most powerful scientific platform for national security, giving ARC a structural market redefinition advantage: access to federally backed qualification infrastructure that most AM service bureaus cannot replicate. For the broader AM industry, this represents a deliberate institutional bridge between national lab research capability and production-scale defense manufacturing, a gap that has persisted despite years of government AM investment.

For ARC, the practical challenge is execution: deploying seven production nodes with integrated Peregrini AI, securing the nickel superalloy binder jetting process for turbine components, and demonstrating that the compressed qualification timeline holds under real defense procurement scrutiny. ORNL's credibility and computational resources provide a strong foundation, but the partnership's success will be measured by whether ARC can convert lab-proven technologies into repeatable, certified production at the volumes the warfighter needs.