Market Insight

Uncrewed maritime systems: Operation and regulation

Uncrewed maritime systems are moving fast from defence pilot projects into commercial use, and offshore operators are increasingly in the frame. Following a seminar hosted by HFW, Thales and Austal in Perth, this article sets out the efficiency and safety case for autonomy, the regulatory grey areas it creates, and where legal risk is shifting for operators, designers and technology providers as these systems become part of everyday operations.

Australia, the United Kingdom and the United States have identified the development of payloads and enabling systems for uncrewed undersea vehicles as their first AUKUS Pillar II signature project. That priority says a lot: uncrewed maritime systems are not just a nice-to-have, they are increasingly necessary in a contested and rapidly evolving environment. The same is true of commercial applications: these systems are not confined to defence programs but are already in commercial use, and offshore operators are increasingly in that frame. Understanding what these systems can do, and the legal complexities they create, matters now, not in five years’ time. This is another example of regulation struggling to keep pace with technological development.

Efficiency and safety gains, at a cost

Autonomous, AI-assisted and remote operations technologies offer real commercial benefits to the maritime sector: fewer crew, lower fuel costs and fewer accidents. But realising those benefits requires navigating new operational and legal challenges.

The land transport sector shows what is coming. Autonomous heavy trucks are already being deployed by major retailers in the United States, cutting operating time and fuel use, and reducing the risk of human error. It is inevitable that commercial shipping will head in the same direction with several pilot projects already underway. For passenger and ferry services in particular, where operating costs are a constant pressure, figuring out how far autonomy can be layered into existing operations is a live issue.

One of the most immediate applications for uncrewed systems is in the name: crew reduction. This is already being implemented through AI-assisted monitoring at sea, which takes live camera feeds and layers them with artificial intelligence for continuous monitoring. These systems can recognise a person on deck, a failure in a machinery room, an object or animal in the water, or another vessel nearby, and trigger the relevant adjustment or alert.

Fuel efficiency is another clear commercial driver. Autonomous systems managing large, complex manoeuvres such as docking can deliver measurable savings. Ferry operators around the Isle of Wight, one of the United Kingdom’s most congested waterways, are already using AI systems that predict and plot optimal routes, cutting fuel usage from route adjustments and manoeuvring.

There is a safety case too. An estimated 80% of maritime incidents are caused by human error. Take the human out of the equation, and safety at sea should improve. That is why appetite for these systems is building, but not without resistance.

Solving one problem can create another. For example, reducing crew or automating route planning raises the obvious question of what happens to the seafarers and support staff whose roles disappear. In addition, the more autonomous a system becomes, the more regulatory scrutiny it would ordinarily attract. However, the case of uncrewed maritime systems is one where practice has greatly outpaced regulation, with regulators playing catch-up in seeking to ensure safety at sea. It is important that this post-facto regulation should not stifle development, but issues are frequently discovered only after deployment, which raises enormous risk for regulators seeking to address this emerging problem. Industry and regulators therefore need to work together to develop a regulatory framework which strikes a balance that allows these technologies to continue to develop while ensuring they can be operated safely.

Why this matters for offshore operators

The issues created by the use of uncrewed maritime systems apply directly to offshore operations servicing the oil and gas sector. The sector is already deploying systems of varying degrees of autonomy in remote operations, often in some of the most congested and contested waters in the world. From a regulatory perspective, there is an obvious parallel with offshore assets, as an offshore installation is often subject to oil and gas regulation once its legs are down or connected to production infrastructure, but reverts to vessel regulation once it is disconnected and starts moving. Uncrewed systems present a similar challenge, shifting between regulatory regimes depending on their function and operating environment.

Jurisdiction, both in terms of nationality and domain, adds additional layers of complexity. These systems are increasingly built, sold and operated across borders, so certification tends to happen system by system, and the applicable law depends heavily on where a system is manufactured, deployed and operated. There is also a live demarcation question as to where maritime law ends and aviation law begins. Take the example of a vessel that launches a drone to assist in the manoeuvring of a platform into position, or the coupling of a platform with an accommodation vessel. If reliance on the drone causes damage to the vessel or rig, the question remains open as to which legal regime should apply. That could have profound consequences, as maritime liability is in general ‘limited’ and based on concepts of tortious ‘liability’, whereas liability in an aviation context is on the whole ‘unlimited’ and ‘strict liability’. A vessel operator could inadvertently void its insurance coverage (or at the least, compromise it) by deploying an aviation asset that damages maritime assets. This is a legal conundrum that courts globally have yet to grapple with, but that day is inevitable and fast approaching. Add to this the fact that there is almost no international or national regulation of subsea environments, at least within the water column, and a quagmire of legal uncertainties emerges across domains in which there has been historically little need for coordination between the relevant regulators, such as aviation and maritime regulators.

Where the real risk sits

Standard contractual risk allocation may not anticipate the systems designer or technology provider at all. Knock-for-knock arrangements between vessel and asset owners, where each party agrees to bear responsibility for their own property damage and personnel injuries, are well understood. But absent specific contractual protection, a designer could be brought into a claim outside of that arrangement. This means that the designer or manufacturer of an autonomous system incorporated into a vessel might be held liable for damage caused by that vessel. This is a critical issue that operators and their contracting counterparties should turn their minds to now – not after an incident. A further layer of complexity arises when the law of tort is considered in the context of the design, manufacture and installation of the autonomous system, and the relationship of each stage to the incident causing damage. Potentially, that could involve four or more separate legal jurisdictions and questions as to where the tortious negligence occurred, which would invoke complex questions of conflict of laws. Yet further complexity would arise given that a vessel owner would prima facie be able to limit its liability under the applicable limitation of liability regime, but that ability may be compromised if an aerial drone were implicated in causing the incident. Furthermore, the designers, manufacturers and installers may not have the same capability to limit liability. That ‘complexity’ could create opportunities for claimants to ‘shop around’ for defendants with the least ability to limit their liability.

Connectivity failure is a related risk. If an uncrewed system loses its control link, the onboard systems will largely determine what happens to that vessel, including how much decision-making capability sits locally. Such events, or even the possibility of them, raise questions about how a vessel or platform can or should operate safely when effectively left unsupervised.

Legal precedent for errors made by autonomous systems remains sparse. That will not last. As deployment increases across commercial contexts such as oil and gas, disputes will inevitably test how existing legal principles apply to decisions made, in whole or in part, by a machine.

Looking ahead

Effective regulation is already struggling to keep pace with deployment. Autonomous systems are actively used in commercial contexts, including oil and gas, and in congested waters such as the Strait of Hormuz. A flexible, outcomes-based approach, rather than a rigid, ground-up prescriptive model, is more likely to be fit for purpose in a domain where technology is moving faster than any legislature can respond.

The most significant current challenge in developing a full regulatory framework is that decision-making can occur without human input: systems can sense and gather information, but the harder questions arise from what they do with it once no human is in the loop. Legal precedent for autonomous system errors remains underdeveloped, and while legal rules could, in principle, be coded into a system’s decision-making, there is no agreed position on what those rules should look like, how they might be followed in novel circumstances, or how they could result in unexpected outcomes.

There is clearly a need for regulators and industry stakeholders to work together now to address the current regulatory vacuum and establish an operational framework that provides clarity rather than waiting for the courts to determine how risk and obligations are to be apportioned in response to a casualty. That is easier said than done. The first step may simply be bringing together regulators from sectors that have historically had little need to engage with one another, let alone with each other’s regulatory frameworks.

In the meantime, offshore operators considering uncrewed systems should:

review contracts with systems designers and technology providers now to confirm whether existing knock-for-knock or liability allocation arrangements actually capture them and what recourse exposure may arise;
map out, before deployment, which regulatory regime applies to a system at each stage of its operation – particularly where it moves between maritime, subsea or aviation contexts;
build contingency and fail-safe planning into procurement and operational protocols for loss of connectivity, rather than treating it as a remote risk; and
engage early with regulators and insurers, given the case-by-case and evolving nature of the regulatory landscape in this space.