Attacking the Drydock Problem: Robotics and the Compression of Repair Time

There is a phrase that keeps surfacing in conversations with senior Navy leaders and industry executives working the sustainment side of the fleet: the drydock problem. It is shorthand for something larger than a scheduling headache. A ship sitting in dry dock, or tied up waiting its turn to get into one, is a ship that is not deterring anyone. In an interview on fleet readiness, Barrett put the point bluntly: the drydock backlog is not a maintenance inconvenience. It is the enemy of deterrence.

That framing deserves to anchor how we think about robotics in naval maintenance. It is easy to discuss robotic welding, hull-scanning drones, and AI-enabled inspection as narrow technical improvements, nice-to-haves that shave a few percentage points off cost or a few days off a work package. But if the strategic problem is that ships spend too long out of the fight because they are stuck in the industrial pipeline, then robotics applied directly to ship repair is not a side conversation. It is one of the most direct available levers against the drydock problem itself.

Compressing Time, Not Just Cost

The Navy’s own recent moves point in this direction. Gecko Robotics has been awarded a five-year Indefinite Delivery, Indefinite Quantity contract to deploy hull-scaling robots and AI-driven inspection across destroyers, amphibious ships, and littoral combat ships, beginning with 18 vessels in the Pacific Fleet. The company’s pitch is not primarily about labor savings; it is about speed and certainty. Gecko’s CEO Jake Loosararian has described the technology as detecting repair needs up to fifty times faster than traditional inspection methods, replacing an analog, manpower-intensive process with a digital record of hull, deck, and weld condition that can be interrogated on demand.

That distinction matters. The old model of ship maintenance answers the question “what is wrong with this ship?” slowly, often only after the vessel is already in the yard and workers have opened up compartments to look. A faster, more complete answer to that question, generated before or immediately upon arrival, is what allows a yard to plan work packages precisely rather than discover scope creep once the ship is already occupying a dock. Every day a ship spends having problems diagnosed rather than having problems fixed is a day subtracted from the deterrent value Barrett is pointing to.

Fairbanks Morse Defense’s work on diesel engine overhauls illustrates the same logic from the other end of the process. For two decades, engine maintenance operated on what FMD’s Jim Kenny calls an “open and inspect and time fills” model, durations that should have taken weeks routinely stretched into months, because the Navy was effectively paying for time rather than outcomes. FMD’s demonstration that a complete diesel overhaul can be delivered to OEM specification in 30 days or less, using robotic welding for the most repetitive and physically punishing repair tasks, is a direct assault on that old paradigm. The robotic welding cell that lets a machine execute a thousand consistent weld passes in an engine block, rather than a human welder fatiguing over three weeks in a 95-degree oil sump, is not an efficiency footnote. It is measured in weeks of hull availability returned to the fleet.

The Foundry as the Limiting Factor

Adm. Daryl Caudle’s Foundry–Fleet–Fight construct gives this argument its proper strategic frame. The Foundry, the shore infrastructure, depots, and industrial base that sustains the Fleet, is not backdrop to the fight; it is a warfighting system whose output is readiness itself. Judged against that standard, the drydock problem is a Foundry problem, and robotics aimed at ship repair is a direct test of whether the Foundry is being modernized fast enough to matter.

The three diagnostic questions worth asking of any robotic repair program are the same ones that apply to the drydock backlog generally.

• Is the technology moving from demonstration cells into standard practice at scale, or is it confined to a handful of showcase ships?
• Is the data these systems generate, weld parameters, corrosion maps, hull condition scans, actually flowing into the maintenance and scheduling systems that determine dock time, or is it being treated as disposable byproduct?
• And is the industrial culture around these tools, from certification regimes to workforce acceptance, evolving as fast as the hardware?

Where the answer is “not yet,” the bottleneck is not the robot. It is an industrial base still configured for artisanal, one-ship-at-a-time work rather than the high-throughput, data-rich production that a fleet under sustained pressure requires.

Why This Is the First Move, Not the Only One

Framing robotics-in-repair as the first aspect of the drydock problem is deliberate. It is the most tractable piece: the technology exists, contracts are being signed, and the return on investment is measured directly in days of hull availability. But it is only the opening move.

The deeper questions, whether ship classes are being designed with robotic access in mind, whether data-sharing arrangements between the Navy and its industrial partners are structured to reward speed rather than protect proprietary silos, and whether a forward-deployed layer of repair capability can push some of this compression out of CONUS yards and toward the fleet itself, will determine whether the drydock problem is actually solved or merely made marginally less painful.

Barrett’s framing should be taken at face value: if deterrence depends on ships being where they need to be rather than where the yard schedule says they must be, then every technology and every process change that shortens the interval between “this ship needs repair” and “this ship is back on station” is not a maintenance story.

It is a warfighting story.

Fighting with the Fleet You Have and Building Availability into the Autonomous Systems Age