This Ship Does Not Have a Client

Week 45 | October 2026

In July I reported on LinkedIn that EUROGUARD’s sea trials with the Estonian Navy were coming this (European) autumn. Baltic Workboats told Janes last year it expected to launch the hull this month. As at publication, I haven’t found any report that it has.

Prototypes slip. A 45-metre, multinational demonstrator that appears to be running behind wouldn't normally earn a Brief. What earns it is what happened elsewhere while the hull sat at Nasva.

On 29 September the US Navy signed production agreements for 30 medium uncrewed surface vessels, ten each from Galliano Marine Services, HII and Saronic, at an average of about US$40 million a hull. First deliveries are due before the end of fiscal 2027. The same day, Brussels closed a €90 million European Defence Fund call for an "enhanced" medium semi-autonomous vessel. The call text says it aims to review the concepts and designs developed under the topic that funded EUROGUARD.

Washington bought boats. Brussels put €90 million on the table for the next prototype of a ship nobody has yet ordered.

The line that carried the July post came from Jüri Saska, a former commander of the Estonian Navy and now Baltic Workboats' project manager for EUROGUARD. He told Estonian public broadcaster ERR: "This ship does not have a client.”

I quoted him short. The sentence keeps going. In ERR's English version, published on 25 March, the ship has no client "apart from the framework document under which it is being built and the capabilities being developed." In the Estonian original he just calls it the paper, Saska didn’t say which document he meant. The regulation tells you what has to sit behind it.

That paper turns out to be the whole story.

EUROGUARD costs about €95 million, with the EU putting in up to €65 million through the European Defence Fund. The fund's regulation (EU 2021/697) attaches a condition to exactly this kind of work. Before EU money goes into prototyping or testing, Article 21(3) requires the consortium to show, in documents issued by national authorities, that at least two member states "intend to procure the final product or use the technology". Recital 30 says that intent can be expressed through a memorandum of understanding or a letter of intent.

So somewhere in a ministry filing cabinet there's paper from at least two governments saying they intend to want this ship. Which governments, I can't tell you. Those documents aren't public. What I can tell you is what a letter of intent costs. Nothing. It carries no appropriation and it creates no program of record. Some governments have probably put money into the prototype itself. Baltic Workboats says the €30 million outside the EU grant comes from participating countries, though the split with industry isn’t public. Chipping in for a demonstrator still leaves the ship without an order.

The regulation’s drafters knew the difference. Article 22 lets evaluators score a proposal higher where member states have actually undertaken to jointly use, own or maintain the final product. A real commitment earns points. It is never a condition of funding. Article 13 then sets the EU’s share of the prototyping stage at a base of 20%, which bonuses for PESCO projects and smaller firms can lift, while earlier design work can be funded in full. Recital 25 gives the reasoning: the prototype phase is where member states usually decide on their investment and start the acquisition process, and the share is kept low “to ensure the credibility of their commitment.” That's the theory. Nearly three years after kick-off, EUROGUARD has a hull and no announced acquisition.

Read that way, Saska was simply describing the instrument accurately. The EDF pays to develop capability against a stated intention, and the step from intention to order belongs to somebody else's budget: EDIRPA, EDIP (EDIRPA’s successor), the SAFE loans, or a national defence appropriation.

The US went the other way round. The Navy set the terms, tested seven vendors at sea, paid those that finished and then bought. Its Phase II call even rewrites the requirement, easing some Phase I demands while still defining what ‘medium’ means: hulls up to 90 metres, at least ten days of unmanned operation. The buyer wrote the requirement. Two excluded firms are suing over how the Navy applied it, which is a separate fight.

The Dutch did something similar on a much smaller scale. In August the Netherlands authorised the design of a 12-metre uncrewed vessel carrying a Thales Compact FLASH dipping sonar, one for each of its new anti-submarine frigates. Tiny next to EUROGUARD, and also tied to a named host ship, operated by a navy that already knows where it will live.

The UK’s Project Beehive complicates my argument, and you’d spot it anyway, so I’ll say it first. In March the UK MoD awarded Kraken Technology Group £12.3 million for 20 K3 Scout USVs with no fixed operational requirement behind them. The Royal Navy was explicit that it was buying a way to learn what it needs. That's building before the requirement too. What differs is who holds the boats afterwards: the Coastal Forces Squadron and 47 Commando, which own and operate them. Beehive has a user with an open question. EUROGUARD has a design and nobody lined up to sail it. Of those two gaps, I'd back the first to close faster.

There's a decent defence of all this, and the regulation makes it itself. European industry can't self-fund defence development at this scale, and no single Baltic navy was going to pay for a 45-metre modular prototype. Getting 23 partners across ten countries, Damen and Fincantieri among them, into one consortium took about a year and a half by Baltic Workboats’ own account. They have since agreed one hull and one modular design. That’s a real achievement. And a hull with sea miles on it is a much easier thing to sign for than a render.

Saska, as it happens, is no admirer of the buy-second habit. At the Smart Sea centre’s annual seminar, reported by TalTech in February, he asked how many times the Estonian state had been the first customer of an Estonian defence company. The answer, as reported: not once. If the state won't order Estonian-made boots before another country places a big order, he argued, the whole innovation chain slows. The man running the project knows exactly which problem he's working inside.

For primes and autonomy firms the lesson is practical. An EDF-funded demonstrator signals political backing for a capability and says nothing yet about who will pay for it. Treat the letter-of-intent signatories as a shortlist of possible customers and price the risk accordingly.

The European Court of Auditors got to a version of this in September, warning that EU capability planning runs bottom-up and risks fragmented investment and persistent gaps. EUROGUARD is what that sentence looks like in steel.

The test I flagged in July hasn't moved. It lands in 2027, when somebody has to sign. What has changed is the order of operations. With the €90 million successor call already closed, the EU may fund generation two before any government commits to generation one. And because that call covers prototyping, any bid for it would have needed fresh intent letters from at least two governments. Europe may have a second round of paper before the first has turned into an order. 

Week 40 asked whether you scale by owning the shipyard or owning the sonar. A third group of operators runs the fleet itself and sells the coverage. I’ll look at what that does to how navies and offshore operators buy.

Since you have been, thanks for reading.

Cheers,

Mick

Ocean Tech Intelligence provides informational analysis only. Nothing in this publication constitutes financial, investment, legal, or strategic advice. Readers act on this content at their own risk. For full details see our Disclaimer.