Metal additive manufacturing has a well-known bottleneck that has nothing to do with printers themselves: proving, after the fact, that a part is actually as good as it looks. According to a report from VoxelMatters, America Makes and the National Center for Defense Manufacturing and Machining (NCDMM) have selected EWI, the Columbus, Ohio-based engineering and technology development organization, to lead a three-year, $13.2 million project aimed squarely at that problem. The project is called INSITE — Integrated System for In-situ Testing & Evaluation — and it's a bet that pairing real-time process monitoring during a build with AI-assisted nondestructive evaluation (NDE) afterward can finally give defense manufacturers a fast, trustworthy way to sign off on large, complex, dense metal parts.

If that sounds like a narrow technical fix, the partner list suggests otherwise. EWI isn't running this alone — the project brings in Lockheed Martin, Northrop Grumman, Boeing, GKN Aerospace GTC LLC, Norsk Titanium, Addiguru, FormAlloy, TRI Austin, and Computational Tools, according to VoxelMatters. That's three of the largest defense primes in the country sitting at the same table as a metal-AM machine builder (FormAlloy), a process-monitoring software company (Addiguru), and independent testing labs (TRI Austin). It's the kind of roster America Makes typically assembles when it wants a result that becomes a standard, not just a research paper.

Why Inspection, Not Printing, Is the Holdup

For desktop and prosumer printing, "quality control" usually means eyeballing a print for layer shifts or stringing. Metal additive manufacturing for aerospace and defense operates on an entirely different plane. A directed energy deposition (DED) build of a large metal structure can take many hours to days, involves complex thermal behavior that builds up across hundreds of layers, and — critically — has to be certified as free of internal defects before it's cleared to fly or fight. Right now, that certification typically leans on post-build nondestructive evaluation, which is slow, expensive, and doesn't scale well to the larger, denser parts the industry increasingly wants to print.

INSITE's pitch is to attack the problem from both ends at once. According to The Vindicator, the system is meant to track process data, thermal and visual images, and eddy-current array signals in real time during the build, flagging likely defects while the part is still on the machine rather than after it's been fully built. Layered on top of that, the project uses AI and machine learning to flag areas of post-build inspection data that may need a closer look — a way to interpret scan results faster and more consistently than a human reviewing scan after scan. Combined, the goal is a verification pipeline that doesn't force manufacturers to choose between speed and confidence.

According to VoxelMatters, Ben DiMarco, America Makes' Technology Transition Director, put the stakes plainly: "As additive manufacturing advances toward larger, more complex applications, the ability to reliably verify part quality is becoming increasingly critical." That's an institutional admission that the printing side of metal AM has, in some respects, outrun the inspection side — machines can now build bigger and denser than the current QA toolkit can efficiently certify.

A Defense-Certification Play, With the Air Force, Army, and Navy at the Table

A separate report from The Vindicator fills in the government side of the picture, confirming that EWI was selected by America Makes and NCDMM and noting that government collaborators on the project include the US Air Force, Army, and Navy. The Vindicator's report also specifies that INSITE targets directed energy deposition metal 3D printing specifically — a process that can replace castings and forgings — and that beyond the technology itself, the project is meant to produce industry and military standards developed alongside those Air Force, Army, and Navy partners, not just a one-off inspection technique for EWI's own use.

That framing matters. Standards work is slower and less visible than a flashy new printer launch, but it's often the actual gating factor for adoption. A defense prime can have a metal AM process that works beautifully in the lab and still be blocked from putting a printed part on an aircraft or ship because there's no accepted, repeatable way to certify it met spec. The Vindicator quotes DiMarco again on this point: "By strengthening confidence in AM technologies, this work will help accelerate their adoption across the defense industrial base." In other words, INSITE isn't trying to make printing better — it's trying to make trusting the printed result faster, which is the thing actually standing between prototype and fielded hardware for large metal AM structures.

What It Means for Makers

None of this touches a desktop FDM or resin printer directly — INSITE is scoped to directed energy deposition metal AM, a process that bears little resemblance to a Bambu Lab or Prusa sitting on a hobbyist's bench. There's no software, firmware, or slicer feature here that trickles down to consumer machines in the next product cycle.

But it's worth understanding what this signals about where the industry's money and attention are going. The $13.2 million isn't funding a new printer or a faster deposition head — it's funding the unglamorous, unsexy layer underneath: how do you actually know the part is good? That's the same question every additive manufacturing segment eventually has to answer as parts get load-bearing and safety-critical, whether that's a bracket on an aircraft or, eventually, a metal AM part in a car or a piece of industrial equipment. The techniques and standards that come out of a project like INSITE — in-situ monitoring correlated with post-build inspection results, AI models trained to flag areas needing a closer look — are well placed to migrate outward over time, first into other aerospace and defense supply chains that already work with primes like Lockheed Martin, Northrop Grumman, and Boeing, and later into commercial metal AM service bureaus that want to offer the same confidence without the same certification overhead.

For makers and small shops running metal AM equipment today — bound powder, DED, or otherwise — the practical takeaway is smaller but real: watch for the inspection and monitoring tooling that emerges from this consortium, including whatever Addiguru and FormAlloy contribute to the project, to show up as add-on products or software features well before any resulting military certification standard becomes public. The quality bottleneck this project targets is an industry-wide one; the fix, if it works, is well positioned to eventually reach beyond defense primes.

Sources