A startup 3D-printed thousands of parts on a Navy warship in 'the roughest water in five years'

An amphibious assault ship pitched through Pacific swells tall enough to qualify as Sea State 5, the maritime scale's marker for rough water that already pushes most small-boat operations off the deck. Below decks, in two 20-foot shipping containers bolted to USS Essex's well deck, a small team of engineers pulled freshly 3D-printed parts off a build plate. By the time the ship reached RIMPAC 2026, the biennial Rim of the Pacific exercise that brings roughly two dozen navies to Hawaiian waters, Firestorm Labs had produced what the company counts as thousands of unique parts aboard a warship in transit, without resupply from shore. The mechanical demo worked. The contested-logistics claim attached to it is still the company's argument, not the Navy's conclusion.

The printers were HP multi-jet fusion machines running on nylon composite, packaged inside Firestorm's xCell unit, a deployable pair of containers the company says can be set up almost anywhere with power and a hard surface. The workflow runs on a single commercial partnership: Firestorm holds an exclusive arrangement with HP for this print pipeline. The transit gave the team a real test, not a lab one. Sea State 5 seas kept the ship pitching for hours at a time, and the xCell kept running. Firestorm field operations manager Duane Blank told Defense One the conditions were "the roughest water off the California coast in the last five years," a company characterization rather than an official sea-state record.

Twelve of the parts were Squall first-person-view drones, small, fast, deliberately attritable aircraft meant to be treated as expendable. Once Essex reached RIMPAC, the exercise's counter-drone drill used those same drones as adversary aircraft to stress-test ship defenses. Other printed items were not exotic. Rotor-scrape stops, a small bracket that keeps an AH-64 Apache's rotor blades from chewing into the ship's non-skid deck coating, were printed below decks and used before the ship reached the exercise area. These are the unglamorous parts that, in a real Pacific operation, would be the first to run out and the hardest to reorder on a short timeline.

This is the "expeditionary manufacturing" pitch that Firestorm CEO Dan Magy has been building since the company formed around the idea. His argument, made on the record, is that a fleet operating under blockade or shore denial could keep producing its own consumable parts if it carried an xCell. He frames the system as a "contested logistics solution that is going to be scalable across all branches of the Department of War," and points to Ukraine, Iran, and the South China Sea, where, Magy says, 70% of the world's goods are produced, as the operational reason. That is company framing, not a Navy commitment, and one RIMPAC transit in heavy weather is not the same as sustaining production through a real Pacific blockade.

The single-vendor piece is the load-bearing constraint the demo does not address. Expeditionary manufacturing, in the form Firestorm is selling, depends on HP's multi-jet fusion resin supply, on HP's printer servicing pipeline, and on the commercial terms of an exclusivity arrangement. A wartime logistics layer built on that stack inherits each of those dependencies. If the resin line stalls, the well-deck container goes quiet. None of that shows up in the printed-part count.

RIMPAC 2026 has been framed by the Naval Postgraduate School as the largest advanced manufacturing demonstration in Defense Department history, a signal that the sea-based production idea is getting institutional attention beyond one company. The same exercise also ran a separate digitally crewed surface-vessel resupply drill under the Navy's CAMRE and FLEETWERX programs, evidence that the contested-logistics problem is being worked on more than one way at once. The next test for Firestorm's claim is whether the xCell runs at tempo in a real exercise, not a transit.