SPEE3D Printed at Sea, But the Capability Lives in a Tennessee Armory

Eight Parts Afloat, Twenty-Four From Knoxville

On an open deck aboard MV Asterix, a Canadian support vessel working the Hawaiian operating area, operators from CAMRE and the Tennessee Army National Guard sprayed metal powder into the shape of an aluminium bronze hatch knob while the ship moved under them. That part is the photograph. The ratio behind it is the story.

Across RIMPAC 2026, SPEE3D systems produced 32 mission-critical metal parts. Twenty-four came through a reachback capability at the Knoxville Armory in Tennessee; eight were built at sea aboard Asterix, according to MarineLink's report of the CAMRE and SPEE3D figures. Of the 32, 29 were taken through the complete manufacturing process, including heat treatment and machining. The at-sea printer - the thing the capability is usually named after - accounted for a quarter of the output.

That is not a criticism of the demonstration. It is a description of what was actually demonstrated. CAMRE, working with the US Army DEVCOM Army Research Laboratory and partners including the US Marine Corps, FLEETWERX and the University of Tennessee Knoxville's Defense Development and Applied Research Center, deployed more than 50 advanced manufacturing technologies during the exercise. What the SPEE3D numbers describe inside that is a distributed production network: a shore node doing the volume, a finishing cell doing the work that turns deposits into components, and a container on a deck for the parts that could not wait for either.

Cold Spray Makes a Blank; the SPEE3Dcell Makes a Part

Cold spray manufacturing accelerates metal powder to supersonic velocity and bonds it by plastic deformation rather than melting. No melt pool means no inert gas envelope, no optical train to keep aligned, and far less sensitivity to humidity, vibration and ambient temperature. That is exactly why this process, rather than powder bed fusion, is the one that ends up on an open deck in a seaway.

The trade sits on the back end. As-deposited cold spray material is dense but heavily work-hardened, and its useful mechanical behaviour is largely set by what happens after deposition; near-net shape also means mating surfaces still have to be cut. SPEE3D's answer is the Expeditionary Manufacturing Unit - the containerised XSPEE3D printer paired with a SPEE3Dcell post-processing and machining unit with integrated CNC, so a component can move from powder to finished part without leaving the ship.

The 29-of-32 figure is the clearest available read on how binding that back end is. Nearly every part needed the second container. A vessel that receives an XSPEE3D without a finishing cell and a trained operator has not acquired a spare-parts capability; it has acquired a deposition capability and a queue.

Note also what "complete" means in this disclosure. It means heat treated and machined. It does not mean inspected, and it does not mean accepted.

Essex Printed Metal at Sea First; Bataan Bolted It to a Hull

At-sea metal printing itself dates to 2022. Xerox's Elem Additive business put its ElemX liquid-metal machine aboard USS Essex that July and printed the first metal part at sea using a wire-fed aluminium process. Months later, a Phillips-integrated Meltio laser wire deposition head under Haas CNC control went aboard USS Bataan as a permanent print-and-mill cell tied to a single hull.

RIMPAC 2026 tests a third architecture: transportable containers allocated to an exercise rather than to a ship, and placed here on a coalition auxiliary rather than a US combatant - a considerably larger base of available decks if the model survives contact with sustainment reality. The narrow and accurate claim is the one MarineLink made, that this was the first at-sea deployment of a SPEE3D cold spray system.

Measured against CAMRE's own previous run, the trajectory is throughput and workflow completeness rather than novelty. RIMPAC 2024 saw the same deployable system print 11 metal parts, all ashore in Hawaii. Two years on the count is 32, eight of them underway, with a documented finishing step that the 2024 coverage did not report at all.

Our company index carries SPEE3D as a Melbourne-based hardware company founded in 2014, and our news pipeline logged 4 items matching the name in the last 180 days against 2 in the 180 days before that - matched on company name only, but the shape is a firm whose public visibility still arrives in exercise cycles.

Keyport's NAVSEA Signature Is the Milestone RIMPAC Did Not Reach

Three weeks before the RIMPAC results were disclosed, Naval Undersea Warfare Center Division Keyport became the first Warfare Center to receive official approval from Naval Sea Systems Command for a metal powder bed fusion procedure, validating production of tactical- and shipboard-level parts in 17-4PH stainless steel on an EOS M290, per the command's own release. Same customer, same end of the value chain, opposite track. Keyport can now make one alloy on one machine that the Navy will permit sailors to install. CAMRE and SPEE3D made 32 parts, none of which is reported as installed, entered into the supply system, or inspected against a published protocol.

That gap is the counter-signal, and it is not the only one. No NDT, mechanical testing or acceptance criteria accompanied the disclosure, and the three parts that did not complete the workflow are unexplained. "Heavy seas" is unquantified - no sea state, wave height or platform motion data was released, so the environmental envelope is a claim rather than a measurement. The reachback leg, meanwhile, assumes a data and logistics chain from a ship to Tennessee that a scheduled exercise on a permissive auxiliary supplies for free and a contested Pacific would not.

Singapore's Propeller Programme Went In Through Class Rules Instead

The closest comparison outside the US military track sits in Singapore, where Mencast Marine took wire-arc deposition toward shipboard hardware by building the certification case first - a propeller programme run with the American Bureau of Shipping explicitly against classification rules, with Lloyd's Register and ABS approvals now reported by the company. Same segment, opposite entry point: one route starts with an exercise and looks for an acceptance path afterwards, the other starts with the rulebook and looks for parts to run through it. A class certificate is not a NAVSEA stamp. But both are acceptance artefacts, and the RIMPAC parts hold neither.

Read as a procurement signal, RIMPAC 2026 argues for buying a network instead of printers-per-hull: armory-class shore nodes with university engineering support, finishing capacity sized to print capacity, and the communications chain that lets a ship task a node an ocean away. Read as a qualification signal, it argues for nothing yet. The event worth waiting for is a cold spray part with a stock number, not a larger part count at RIMPAC 2028.