Autonomous Resource Corporation partners with ORNL on Exascale Foundry for AI-enabled defense AM

Autonomous Resource Corporation (ARC), a New York-headquartered defense manufacturing startup, has signed a Memorandum of Understanding with Oak Ridge National Laboratory (ORNL) to establish the Exascale Foundry, a public-private partnership aimed at accelerating AI-enabled, on-demand production of qualified defense components. The collaboration will integrate ORNL's exascale supercomputing resources and Peregrine AI software—which has analyzed over 1.9 million additive manufacturing layers—with ARC's ARCNet distributed manufacturing platform. Initial focus is on high-temperature nickel superalloy turbine components for autonomous air vehicle engines, produced via metal binder jetting. ARC will deploy seven production nodes connected to ORNL via secure ARCNet infrastructure, leveraging ORNL's Manufacturing Demonstration Facility (MDF) for simulation-driven materials characterization and qualification.

This partnership directly addresses the defense vertical's most persistent bottleneck: the qualification grind for mission-critical metal parts. By combining ORNL's HPC-driven materials simulation and Peregrine's real-time adaptive control with ARC's distributed production model, the Exascale Foundry aims to compress qualification timelines from years to months—a structural market redefinition that aligns with the NDAA §849 domestic sourcing mandate effective December 2026. The focus on binder jetting for nickel superalloy turbine components targets a demonstrated production gap in U.S. defense supply chains, where traditional casting and forging lead times remain a strategic vulnerability. This is not a speculative R&D project; it is a deliberate attempt to create a closed-loop system where AI-driven qualification and autonomous production operate as a single pipeline, directly competing with incumbent defense suppliers who rely on conventional manufacturing.

For ARC, the partnership provides access to ORNL's computational and materials science infrastructure that would otherwise take years and hundreds of millions to replicate. The company's immediate execution challenge is demonstrating that ARCNet can maintain qualification integrity across distributed nodes while meeting defense-relevant production volumes. For the broader AM industry, this represents the most concrete integration of exascale computing with production-scale additive manufacturing to date—moving beyond simulation-as-a-service into real-time, AI-governed production control. The practical test will be whether the Exascale Foundry can deliver qualified turbine components at scale within 18 months, a timeline that would reset expectations for defense AM adoption.