55 Hulls, No Operating Model. Who Runs the Fleet?

Week 27  |  June 2026

In March this year, Australia committed A$176 million to a fleet of 55 Bluebottle autonomous surface vessels. One of the largest sovereign USV procurements anywhere. Two weeks later, the Royal Australian Navy stood up a dedicated unit to operate Ghost Shark, Bluebottle, and Speartooth from shore. Preceding this, in December 2023, the UK's Workboat Code 3 (WBC3) came into force, becoming the first national legislation anywhere to formally define what a Remote Operator actually is, and what they must be able to do.

The navigation problem - the one that consumed most of the last decade's R&D spend, is close to solved. The operations problem is a different matter entirely.

Running a single USV on a 72-hour demonstration mission requires engineering. Running 30-plus hulls on persistent 30-day missions from a shore-based control centre, monitoring propulsion health, managing power consumption, scheduling maintenance windows, and maintaining situational awareness with no crew aboard to notice the smell of an overheating motor or the sound of a fouled propeller, is a different capability requirement. Nobody has credibly demonstrated this at scale. And the gap between demonstration and persistent fleet operations is where the autonomous navy actually lives or dies.

ACUA Ocean's FleetMind platform, announced on 28 March 2026, is the clearest public signal yet that at least one vendor has identified this as the binding constraint.

ACUA Ocean is a Plymouth-based hydrogen-powered USV developer co-founded by Neil Tinmouth (CEO) and Mike Tinmouth (co-founder and COO). Their vessel, Pioneer, is a SWATH-hull autonomous platform certified in August 2025 by Lloyd's Register under UK Workboat Code 3, Annex 2 and is the first vessel to receive that certification.

FleetMind is ACUA’s proprietary Integrated Platform Monitoring System, first deployed on Pioneer in Q2 of last year. ACUA's framing is deliberate and worth taking seriously: existing autonomy software stacks handle command and control, navigation, and collision avoidance. FleetMind targets the engineering stack underneath: propulsion, power management, structural health - the systems a crew would monitor by being physically present. Their pitch is that vessel autonomy and platform monitoring are two separate problems, and the industry has spent years solving the first while largely ignoring the second.

FleetMind’s technical architecture aggregates data from vessel systems across NMEA, PLCs, sensors, and networks into data lakes, enabling what ACUA describes as stateful tracking to power predictive maintenance and automated platform management. The roadmap includes multi-vessel integration for a single-view fleet management dashboard via a Remote Operations Centre. That last capability, multiple hulls on one screen, is the one that matters for any serious fleet operation. It is still listed as a development item.

ACUA reports FleetMind has completed over 7,000 hours of on-water operations since deployment, collecting over 25 billion data points. These are vendor-reported figures with no independent verification, so should be treated as directional, not audited.

One detail worth noting: FleetMind was built in consultation with Lloyd's Register's Human Factors team. That is not a software certification. It is advisory input on the human-systems interface design, how information is presented to a shore-based operator, what gets flagged and when and how cognitive load gets managed across a watch. The fact that a classification society's human factors specialists were at the table at all signals something about how ACUA is framing the problem. This is not a dashboard. It is a regulated operational discipline.

The stakes scale with the vessel. As autonomous maritime systems move beyond small littoral platforms and one-way effectors into workboats and larger ocean-going vessels, the consequences of failure change dramatically. A small USV failing on a coastal mission is an inconvenience. A 12-metre-plus vessel going dark in the Atlantic is a different order of problem: mission failure, costly recovery, and the kind of incident that sets back programme timelines and procurement confidence. Robustness, reliability, redundancy, and graceful failure are not software features. They are the engineering foundation on which operations depend. The complexity of these vessels means failures are inevitable. The question is whether the system fails safely, maintains operational availability, and minimises disruption when it does.

WBC3 came into force on 13 December 2023. Annex 2 is the world's first national legislation specifically drafted for remotely operated unmanned vessels (ROUVs). It defines the Remote Operator, the Remote Operation Centre, and the minimum-safe-manning obligations that apply when no one is aboard.

In practice, WBC3 requires remote operators to hold the same certificates of competency as crew on equivalent crewed vessels, plus supplementary training. Only one individual may hold command of an ROUV at any time. Owner-operators must conduct a minimum-safe-manning analysis that accounts for watchkeeping across the full duration of a voyage, including shift structures and ROC-to-vessel connectivity monitoring. The UK Marine Guidance Note MGN 703 spells out the training and competency requirements for remote operators in more detail.

The UK's Maritime and Coastguard Agency (MCA), the author of WBC3, has acknowledged its national Remote Operator certification standard is still being developed. So, for the UK, the regulatory frame exists, the compliance obligations are live, and the specific competency standard is still being written. That is a procurement problem for any commercial operator standing up a shore-based fleet management function right now.

Internationally, the voluntary IMO MASS Code, adopted at MSC 111 in May 2026, is live from 1 July 2026, applying initially to cargo ships. A mandatory code is targeted for adoption by July 2030, entry into force January 2032 at the earliest. For the near-term autonomous programs already underway, the relevant regulatory framework is national, not international. WBC3 is years ahead of the global schedule. It is where the live experiment is running.

Lloyd's Register's involvement in FleetMind makes more sense when you look at their own research on what shore-side fleet management will actually require.

Their 2024 study of bridge alarm management found that alarms had increased 197% per hour for vessels in open sea, compared with data from nearly two decades earlier. Their 2026 follow-up analysed over 40 million alarm-related events from 11 operational vessels across more than 2,000 days. Fewer than half met the benchmark of under 30 alarms per hour. On vessels with unattended machinery spaces, alarms disrupted 63% of rest periods. Cruise ships peaked at 4,691 alarms in a ten-minute window.

These numbers describe a crewed vessel, where a watchkeeper can physically inspect what triggered the alert, use context and judgment, and decide whether the fourth alarm in ten minutes is the same compressor cycling or something that needs attention. Remove the crew. Now, concentrate monitoring of multiple hulls into a single shore-based operator. The alarm management problem does not disappear. It compounds.

A six-month pilot cited in the LR 2026 research cut total alarms by almost 50% without deploying any new technology, through alarm rationalisation and interface redesign alone. That is why it matters that a classification society's human factors specialists were involved in designing how FleetMind surfaces information to operators.

The Australian government's recent A$176 million investment in 55 Bluebottle USVs is the sharpest demand-side signal in the public record. The Bluebottle is a solar, wind, and wave-powered persistent surveillance platform. Individual vessel endurance runs to multiple months at sea. The fleet, once fully delivered under the five-year Program of Record, will be among the largest sovereign USV fleets anywhere.

Someone must monitor those 55 hulls from shore. Persistently. Across multiple time zones of Australian maritime approaches. With maintenance windows, power management decisions, and situational awareness requirements that do not stop at the end of a business day. The shore-side operating model for that fleet does not yet exist in any published form. It is the open question the Ocius contract puts squarely on the table.

The RAN’s Maritime Autonomous Systems Unit represents a real military unit whose explicit function is heterogeneous autonomous fleet operations from shore. How that architecture is structured, what it requires of its operators, and what gaps exist in the tooling - those are operational details the RAN is working through now. They are not published details.

Over in Norway, Zeabuz offers a different data point. Their dual-use autonomy platform, launched in October 2025 in partnership with Damen, is designed explicitly for a single operator to control multiple vessels simultaneously, with a separate supervisory system monitoring vessel health because there is no captain aboard. The MF Estelle passenger ferry in Stockholm operates at autonomy level 4 under remote supervision. Different jurisdiction, different use case, same converging conclusion: the operator-to-hull ratio and the shore-side monitoring architecture are the scaling constraint.

At OTI, our standing argument is that procurement doctrine, certification pipelines, and institutional risk tolerance are the real barriers to maritime autonomy deployment, not technology readiness. The shore-side fleet management problem is that argument applied to the operational layer.

The vessel autonomy stack is close to solved. COLREGS-compliant navigation, collision avoidance, autonomous berthing, these are engineering problems with engineering solutions, and multiple companies have demonstrated them credibly. What has not been solved, what barely has a product addressing it, is the persistent fleet operations layer: the software that monitors platform health across 30-plus hulls, the trained Remote Operators certified under an emerging competency standard, the Remote Operations Centre designed around human factors principles, the shift structures and watchkeeping protocols for a crew that never boards the vessel.

FleetMind is a pre-Series A, grant-funded product from a company that has raised less than £2 million in equity. ACUA Ocean's total disclosed funding is £5.5 million, of which £4 million came from DEFRA and the Department for Transport R&D grants. The most serious public attempt to build the shore-side engineering stack is being bootstrapped. That is an indictment of where the market has chosen to direct capital. Five years of venture money went into hull autonomy. Almost none went into the operating infrastructure that makes fleets viable.

That gap will close. It has to. But the timing matters. Australia has just committed to 55 hulls and the MASU is operational. That is but one example. Procurement offices are making decisions about shore-side architecture now, without a mature vendor landscape, without an established competency standard for Remote Operators, and without a proven operating model for persistent multi-hull missions. The institutional gap, this time, sits at the operational layer.

The autonomous vessel’s other unsolved problem - fuel. Four technologies are competing to change that. One of them is nuclear.

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.