By Robbin Laird
When I spoke with Robert Dane, founder and CEO of Ocius, on September 30, the Bluebottle program had moved past the “will it work” phase. It is now in the harder phase of deciding how navies and border agencies should build a force around a system that works.
Dane does not use my vocabulary. He never mentions the hull gap or the re-imagining of capital ships. But the way he describes his company’s operating experience speaks to both problems, and this article takes up both: the program update first, then what it implies for the two problems I have been working on. I also draw on my earlier work on Australia’s Maritime Autonomous Systems Unit (MASU), whose flagship systems include the Bluebottle.
Where the Program Stands
Ocius has come a long way from its origins. Dane came from a Navy and shipping family but spent his early career as country doctor delivering babies and doing anesthetics in the seaside town of Ulladulla in the 1980s and 90s. In 1997, with a bunch of sailing and surfing friends from Ulladulla he entered a prototype ‘solar sailing’ boat in a solar boat race in Canberra and then in 2000 began a company called Solar Sailor, building hybrid solar- and wind-powered ferries. In 2014 he changed the name of the Company to Ocius and turned that experience toward defence and the problem of persistent, wide-area maritime awareness. The result was the Bluebottle, a renewable-powered, low-observable uncrewed surface vessel. The company says Bluebottles have been deployed continuously on real missions since 2024.
The production picture is now concrete. The Australian Defence Minister for Industry, Pat Conroy, said publicly at the Pacific Islands Forum that the government is acquiring 55 Bluebottles. Ocius has delivered the second vessel under its SEA1200 contract, BB717, which was christened in September by Sheryl Lutz, who heads Defence’s naval shipbuilding and sustainment group. Dane told me the first vessel of the new build is in test, the second has just been christened, and numbers three through five are close behind.
The new build is more than a bigger order of the same boat. Dane described a step change inside the vessel. Components must now be non-Chinese and of higher military grade, and the company is seeing better performance and detection as a result. The contract requires building 40 vessels to the same rigor while preserving the ability to swap subsystems later. Those two requirements, consistency and swappability, pull against each other. Holding both is the real manufacturing discipline here.
The payloads show the same logic. In August Ocius announced a contract with ThayerMahan for 40 Outpost acoustic payloads with TransparenSea processing software under SEA1200. It also selected Thales Australia’s BlueSentry thin-line arrays for Bluebottles supporting Maritime Border Command operations to the north. Ocius does not sell payloads. The Bluebottle is designed to be array-agnostic, carrying whichever sensor suits the mission and the customer.
Ocius has also grown physically. Dane said the company has expanded from two buildings to four at its Sydney industrial complex, with now 150 and growing number of employees including about 45 software, hardware, and AI engineers. Conroy is due to open a new operations center on November 16. Dane was clear that it will run Bluebottles worldwide.
The Diesel Hybrid
The most interesting near-term development is a variant Dane calls the Atlantis. It is the standard 24-foot Bluebottle, with the same winch, platform, trailer, and logistics, plus a diesel generator. The solar-and-sail boats work well north of Darwin where there is plenty of sun, but at high latitudes in January there is no sun. The hybrid can ride out a cyclone under sail, stay quiet while using its array, and then run the diesel at about seven knots to reposition. In his words, it arrives, turns off the diesel, and becomes a “silent picket.”
Two of these are in San Diego now for the Maritime Big Play exercise, which starts within a week or two. They will then go to Europe for winter trials in the Baltic with a customer there, and on to REPMUS, the NATO experimentation exercise, in September 2027. Dane has just returned from Europe describing the conversations there as very positive.
New Zealand and the Pacific
In New Zealand, the Royal New Zealand Navy operates two Bluebottles, TAHI and RUA, and deployed them to Fiji in August to demonstrate tactical employment to the Fijian government and navy. Ocius has taken possession of a maintenance facility in Auckland, which it says could later convert to production.
Fiji is the next step. At the Pacific Islands Forum, Conroy said the government is looking at acquiring further Bluebottles, starting with Fiji. The Fijian mission set is drug interdiction, fisheries, and illegal entry, and the model would mirror New Zealand: a small local facility, with Fijians recovering, cleaning, and servicing the boat at the ramp. Dane, when he was a doctor, worked in medical clinic on an island in Fiji for 2 weeks every year for ten years, made a point I found persuasive. A Fijian sailor does not want to spend 28 days on a big ship. He wants to go home to his family at night, and a system operated from the shore fits that.
The Service Model and the Hull Gap
My core argument on the hull gap is that the period before new frigates arrive is not dead time. Navies can use it to build the skills, concepts of operations, and organizational habits that let new hulls plug into a mesh from day one. The Mogami decision gives Australia new frigates, but the years before they arrive will shape how well the navy can use them.
Dane’s experience supports this in three ways.
First, the learning is happening now, and it is bidirectional. He described what Ocius is doing with Border Force in the north as “a team sport” that nobody has yet learned to play properly. Every time someone in uniform on a ship engages with Ocius, the result is a different concept of operations. A skipper given a Bluebottle as a loyal wingman or over-the-horizon asset works out quickly what it is best for. He also described the game as adversarial: as Ocius and Border Force change how they position the boats, the people they are trying to catch change how they move. He compared it to whack-a-mole, chess, and sport. The positioning logic comes from many forms of intelligence and assets working in a network.
That is what the hull gap is for. An older ship’s crew that works with these systems now builds exactly the judgment Dane describes. The ships are not optimized for it, but the skills carry over to the new hulls.
Second, the skills are hard to codify, and they have to be. If each skipper develops a different concept of operations, frigate crews will not know what to expect from the systems working alongside them. We are past the prototype stage, and the skipper on a frigate needs a shared baseline. Dane said Australia is setting up an organization to manage Ghost Shark, Bluebottle, and other uncrewed systems, which he described as a world leading initiative. That is the Maritime Autonomous Systems Unit, which I have written about as one of the clearest institutional expressions of mesh fleet thinking in any allied navy. Its flagship systems are Ghost Shark, the extra-large autonomous underwater vehicle; Speartooth, the large uncrewed underwater vehicle; and the Bluebottle. What matters most is its structure: a shore-based Uncrewed Systems Control Centre combined with deployable teams that can operate forward from major bases, austere ports, or embarked on RAN combatants. The operational unit is no longer the hull in the water but the combination of sensors, communications nodes, data flows, and human operators spread across a theatre.
MASU also carries a mandate for experimentation and doctrine, which I think is the right call. The unit that operates the systems must own the concepts, because doctrine written by people who have never operated them fails on contact. Dane says the unit has a mandate and massive potential, the open question is whether the resourcing and institutional backing follow. On the operating side, Dane offered a measure of the experience Ocius brings to it: in two years of actual Border Force work Ocius has accumulated a substantial number of ‘boat days’ on the water, equal to 17 years of experience at sea, and enough nautical miles travelled, as he put it, equal to go from here to the moon.
Third, and most important, the model is a service. Dane agrees with the argument I made to him. The technology is moving fast, and a Bluebottle is a robot with a number of moving parts that a navy short of people should not have to maintain itself. The New Zealand arrangement shows the alternative. The navy’s autonomy unit rotates officers through, so Ocius is opening a facility in Auckland to do the maintenance, turnaround, re-missioning, and reprogramming. The navy then drives the boat and runs the mission.
I have seen this pattern before. When the Marine Corps bought data from the InSitu UAV in Afghanistan rather than the platform, the company was free to keep changing the aircraft. When the government later insisted on buying the next version outright, the program failed because it could not move fast enough. The lesson is that an acquisition structure can ruin a capability that works as a service.
This connects to a point about mobilization. A service arrangement puts the support workforce in the private sector. Those people do not count as military personnel, but they are available in a crisis. A government that insists on owning everything forecloses that.
There is one piece of homework for the industry. I told the Maritime Border Command leadership I wanted a spreadsheet every year showing how much money the government saved by not sending a crewed asset when a Bluebottle did the job: crew, pension liabilities, fuel, hull life. I am not arguing the boats are free. I am arguing for real data, because without it the debate stays ideological. Dane’s version of the same point is that a ‘capability brick’ or pack of Bluebottles on station in a strategic area lets a commander put his ships, fuel, and people elsewhere. And if the adversary knows the pack is there, he goes somewhere else. That is deterrence as a by-product of a cheap, persistent presence and is also hard to measure but should be included in that spreasdsheet
The same logic scales up from Border Force to a problem I sketched in the first of three vignettes I wrote about MASU: grey-zone coercion against a regional partner in the First Island Chain. Committing frigates and submarines to sustained forward presence carries escalatory risk and depletes the scarce high-end assets needed if the situation worsens. What is needed is persistent, low-signature, attritable presence that populates the battlespace with sensing at acceptable political and operational cost. In that vignette undersea vehicles map the environment and characterize adversary signatures, while Bluebottle-class USVs work the periphery of contested zones carrying passive sensors or electronic support measures. The vignettes are thinking tools, not predictions. But what Dane is doing north of Darwin is the peacetime, law-enforcement version of the same task, and it is already generating the experience a grey-zone mission would require.
What This Means for the Capital Ship
The capital ship question is where Dane and I are least likely to use the same words. I reject the term “hybrid fleet.” It suggests bolting uncrewed systems onto a conventional fleet, and that is not what I mean. I mean rethinking the whole interaction, and I think of force planning in terms of zones and building blocks, with different mixes of systems solving different tasks. Dane said the same thing in his own terms: the goal is not to be bolted on.
He is also more modest than I am about what this means for the ships, and I think that modesty is useful. Nobody at Ocius is claiming to replace the frigate. The Bluebottle takes a task off a ship, such as persistent listening or watching a stretch of ocean, and does it better than a crewed vessel would. Some uniformed officers still want to send a ship, and they do not yet appreciate what they can free up elsewhere. Dane called them out gently. The majority, he said, get it.
This is the kill web logic Ed Timperlake and I have applied for more than a decade. The decisive question is no longer what a given ship can do in isolation, but how effectively it operates as a node in a distributed network of sensors, shooters, and decision-makers spread across domains and allied forces. The mesh fleet is the maritime expression of that logic: an integrated, reconfigurable force of crewed and uncrewed vessels, aircraft, and undersea systems, composed and recomposed for specific missions. Seen that way, the Bluebottle is not a replacement for the frigate. It is a deterrent and a node the frigate’s captain learns to call on.
Several elements of his operation bear directly on how the capital ship changes.
The host and the payload separate. Ocius does not make sensors. It makes a persistent host and runs it globally, while ThayerMahan and Thales supply the arrays. For the first time, a company can accumulate operational experience on a payload-agnostic platform across many theaters. That is a practical example of the payload-utility logic Ed Timperlake and I apply to the mesh fleet, where target acquisition and target engagement can be bought separately. In that logic the capital ship becomes less a source of payload and more a manager and orchestrator of it.
Control has to cross the ship’s boundary. My view is that a combat system that controls only the weapons on its own hull is a structural limit. Cost-exchange ratios are now a central consideration. A weapons officer needs access to effectors beyond his hull, whether launched from aircraft or from uncrewed surface vessels, and the ship’s design should reflect that. Dane’s operations center shows the coalition version of this. During the San Diego exercise, the Americans will supervise the boats during their day. When the exercise closes each evening, Ocius in Australia will take over, keep the boats safe overnight, and hand control back the next morning when the arrays are switched on. Persistent control passes across time zones and national lines. MASU’s shore-based control centre and Ocius’s operations center are two versions of the same idea, and the handoff between a government centre and a contractor’s will have to be designed, not improvised. A crew on a capital ship now has to think about which of its assets are in its hands, and when.
Logistics commonality is a design choice. The diesel hybrid keeps the same hull, winch, trailer, and logistics. Adding a capability without adding a logistics tail is exactly what a navy needs when it cannot afford a new support chain for every variant. It also connects to my point about bifurcating shipbuilding. Capital ships need drydocks, people, and physical space, so they are limited in how fast they can grow. Uncrewed systems have a limited ready-force requirement but a potentially unlimited crisis requirement, so the thing to fund is the supply chain behind them, not the inventory. Dane’s insistence on non-Chinese, military-grade components is the same concern seen from the factory floor. A surge line is only as good as its inputs.
A Caution, and a Claim
I want to be careful not to claim more than the evidence supports. In my books I avoid futurism, because much of what is written about autonomous maritime systems is unreliable. The progress has been incremental, but it is real, and the work now is to find language that captains of capital ships can use to understand how it matters to them.
My own claim is this. A capital ship that does not learn to work with autonomous systems will struggle to survive the ones an adversary fields. A capital ship that does learn can defend itself with them and extend its reach through them.
Dane’s claim is narrower, and I think it is the right one to start from. The Bluebottle is not a prototype and not a weapon of the future. It is here, it does certain things well, and where it does them poorly, something else will do them. The demand will grow as navies work out what they can subtract from their crewed fleets.
The next tests are the San Diego exercise, the Baltic winter trials, the November opening of the operations center, the Fiji decision, and whether MASU is given the resourcing and backing to turn its mandate into doctrine. I will follow each of them, and the question to ask at every stage is the same one: is the navy learning to use the system, or only buying it?