Who Builds Unmanned Surface Vessels for the US Navy

The Companies Building Uncrewed Surface Vessels

The clearest public list came from the US Navy in May 2026, when it picked seven designs for at-sea Medium Unmanned Surface Vessel (MUSV) demonstrations. In the guides we publish here, a government shortlist beats a market-research ranking every time, because somebody with money at stake made the cut.

The seven were Sea Machines, Leidos, Saronic, Galliano Marine Services, PacMar Technologies, Birdon, and Huntington Ingalls Industries. Each receives $15 million for the demonstration phase, which ran from June to October 2026. Designs that complete the trials become eligible for follow-on production. Only two entries have public vessel names: the Saronic Marauder, and the Romulus family that Huntington Ingalls Industries is reported to have entered.

Two large names are missing from that list and still matter. Saildrone builds wind and solar survey vessels for ocean science and maritime surveillance rather than for the MUSV requirement. Anduril entered the surface market in 2026 without designing a hull at all. It partnered with Kraken Technology Group to build the K5 Kraken and K7 Sabre in US facilities. It also partnered with HD Hyundai on a medium vessel for the Navy's Modular Attack Surface Craft (MASC) line.

Saronic Owns Its Yards and Anduril Owns the Software

Saronic and Anduril are the two names people compare most, and they are running opposite strategies. Saronic designs its own hulls and buys shipyards to build them. It acquired a yard in Franklin, Louisiana in 2025 and announced Port Alpha in Brownsville, Texas. Anduril builds autonomy and payloads, then partners with people who already know how to build boats.

That difference decides what each one is actually selling. A Saronic contract buys vessels from a company that controls its own production line, so it can be held to a build rate. An Anduril maritime contract buys the Lattice autonomy layer and the integration work. The hull comes from Kraken Technology Group at the small end, or HD Hyundai at the medium end.

Neither approach is obviously better, and the risks sit in different places. Owning yards ties up capital in steel and creates a fixed cost that has to be filled with orders. Partnering gets you in faster and leaves the supply chain under somebody else's control. That second point is live rather than theoretical: one of those partners assembles prototypes in South Korea.

How to Tell These Companies Apart

Three questions sort them faster than any specification sheet. Does the company own the yard that builds the hull? Does it own the autonomy running on it? Is it selling you a vessel or a finished mission?

Yard ownership predicts whether a build rate is a promise or a plan. Saronic and Huntington Ingalls Industries build in yards they control, so their production claims can be checked against hulls in the water. Companies that partner for hulls move faster at the start and control less when a partner's schedule slips.

Selling a mission and selling a vessel are different businesses. Saildrone largely operates its own fleet and delivers the data. A customer never handles a boat, never staffs a shore team, and never owns a depreciating hull. If you want seabed maps rather than a maritime robotics program, that is a different purchase entirely. Comparing its vessel specifications against a Navy hull answers a question you were not asking.

• Owns its yard: Saronic, Huntington Ingalls Industries, Birdon
• Owns the autonomy and partners for hulls: Anduril, Sea Machines
• Sells the mission rather than the vessel: Saildrone
• Builds hulls for someone else's autonomy: Kraken Technology Group, HD Hyundai

What the Funding Does Not Buy

Private funding in this category has outrun delivered production. Saronic raised $600 million in early 2025 and a $1.75 billion Series D in March 2026, led by Kleiner Perkins at a $9.25 billion valuation. That is a large number for a company whose largest ship first touched water in 2026.

Capital buys yards and hiring, and it does not buy a delivered fleet. The MUSV demonstration awards were $15 million per company. That is a rounding error against a venture round, and it is still the only money tied to a vessel actually performing at sea. Read the venture figure as a bet on production capacity, and the Navy figure as the current size of the proven market.

For the listed companies the arithmetic runs the other way. Huntington Ingalls Industries and Leidos have decades of naval revenue, and their uncrewed work sits inside businesses large enough that a USV program barely moves the financials. That makes them slower to bet the company on a new class of ship, and much harder to knock out of a competition.

Who Each Company Sells To

These companies look like competitors on a list and rarely meet in a bidding room. The buyers are separate, and so are the contract shapes.

Navy programs are the largest market and the slowest. A demonstration contract can precede a production decision by years, and some designs never reach one. Budgets shift between phases, requirements get rewritten, and a company can spend three years inside a program that produces no hull. That is normal in naval procurement, and it is why funding rounds and yard purchases are read as commitment signals. MUSV and the Modular Attack Surface Craft line it came from both run demonstration phases before production. That is why seven companies can all be described as Navy USV builders while none has a production fleet. Allied navies buy on similar timelines, usually after the US program has picked a direction.

The commercial and scientific market moves faster and pays less per vessel. Ocean survey, offshore energy inspection and environmental monitoring are the buyers there, and they mostly want data rather than hardware. Saildrone built its business on that side, which is why its vessels are wind-powered and slow: nobody surveying the seabed needs 35 knots.

What to Ask Before Shortlisting a USV Builder

Five questions do more work than a specification comparison, because the published specifications in this category are thin and inconsistently disclosed.

The delivery question catches the most people out. A company that has never delivered a production run can still be the right choice. Just know that is the bet you are making, rather than discovering it later.

The sea-state question is the one most often waved away. Range and speed figures are quoted in conditions the manufacturer chooses. A boat rated for 1,000 nautical miles in a calm sea is a different vessel in a swell. The Navy writes Sea State 4 into its requirement precisely because a number without a condition attached is not a specification.

The recovery question changes budgets. Nobody prices the crewed ship that has to go and fetch a broken hull, and on a long-range mission that recovery can cost more than the vessel saved. Ask early, in writing, and ask who owns the vessel while it is adrift.

• Do you own the yard that builds this hull, or does a partner?
• What sea state are the published range and speed figures valid in?
• What does the vessel decide on its own when the satellite link drops?
• How is the vessel recovered when it fails far from shore, and who pays for that?
• How many hulls of this class have you delivered, as opposed to launched?

Who This Comparison Does Not Serve

Anyone looking to buy a boat should stop here. Every company here sells to governments, navies, or survey and energy customers under contract. None publishes a price or a purchase route for an individual. If you want an autonomous vessel you can own, the market you are looking for is recreational autopilots and small research craft, which is a different set of vendors.

Anyone evaluating underwater vehicles should look elsewhere. Anduril's best-known maritime products, including the Ghost Shark and Dive-XL programs, are large underwater vehicles rather than surface vessels. Judging its surface position by those programs gets the answer wrong in both directions.

One event would change the verdict here: a delivered production contract. The MUSV trials ran to October 2026. The first company to turn a demonstration into a funded production line will have an argument none of the others can make yet. Until then, every claim above about who leads rests on yard capacity, funding, and one Navy shortlist.