Maritime autonomous systems: What it would take to make the Hybrid Navy work
Back in December, before Father Christmas visited and, if you're anything like me, we were all half a stone lighter, I wrote about why the First Sea Lord's Hybrid Navy strategy was both an expedient necessity and a significant opportunity for the Royal Navy.
I have been an advocate for the adoption of maritime uncrewed and robotic systems for over a decade and have spent quite some time thinking about, and working on, what it will take for the Royal Navy to lead and win through the innovative and robust exploitation of maritime autonomous systems (MAS).
That's not a vision that has always been shared.
Fewer ships? That's surely fine
I recall a meeting over 10 years ago when I was sat in a meeting room in Royal Navy Command HQ, briefing a two-star-chaired steering group when one Commodore (I was a Commander at the time) piped up: "But what if all of these robots mean that we need fewer ships?"
My response was that the purpose of the Royal Navy was to deliver military effect in and from the sea to fulfil the Government's policy objectives, and if we could do that more effectively without any (or fewer) ships or sailors that was surely fine?
The point was grudgingly accepted.
I'm genuinely pleased to see that the mindset has now shifted, and the right priority is now being given to these transformational technologies throughout the service.
Crewed ships are still going to be necessary to fulfil the Royal Navy's purpose for the foreseeable future, but if MAS are to be integrated with them to achieve the necessary combat capability, and thus deterrent effect, what will it take to make this strategy successful?
The first thing to note is that the uncrewed platforms themselves are the easy bit (relatively speaking).
Building robot boats, ships and submarines shouldn't be that much harder than building crewed versions. Indeed, so the logic goes, it should be simpler and cheaper because there’s no need to accommodate the humans in the design. That latter part is not an argument that I am wholly comfortable with.
USVs that need to operate and survive for extended periods at sea without human maintenance will need levels of resilience and reliability that generally don't exist in current ships. That's going to be a driver of both cost and complexity.
Alternatively, we simply accept that many of them will be lost and won’t come back. Fine if you're really talking about small, cheap USVs intended to be consumable – but what about the sophisticated air defence assets that have been touted by the First Sea Lord?
These will be expensive and important capital assets that will need to endure.
AI or human input
Nonethless, if the vessels themselves aren't the toughest nut to crack, what are the biggest challenges?
The core of this is what I describe as the "cognitive architecture" of the system of systems and the associated information architecture. That is to say, who or what makes what decisions when and where, and under what circumstances?
Understand this, and implement it, and we can really start to iterate rapidly. It is the substrate on which the structure can be grown. Without such foundations, we're on shaky ground.
Taking uncrewed surface vessels (USVs) as an example, it may be that we want to give the onboard computer the ability to make the decisions of the boat's Cox'n – the person who steers the boat and is responsible for its safe and effective operation.
That is to say, deciding on manoeuvring the USV to reach a destination and, perhaps, the route it takes. Or we may not want that at all, and the boat should simply follow the navigation track to a waypoint without any deviation unless ordered by the remote Cox'n supervising from afar.
Moving in the other direction, we may, in time of conflict decide that we cannot rely on the communications and data links and so need the vessel to autonomously and tactically navigate itself to a launch point and then fire its payload at a pre-allocated target.
At the extreme end, although this is currently contrary to UK Government policy, we may want the vessel to seek, find and engage enemy targets without any input from operators ashore or elsewhere.
Each of these scenarios is indicative, but each comes with different implications for the nature of the sensors needed on the vessel, the reliability and bandwidth of communications between the USV and its commander (wherever located), the sophistication of the algorithms/AI/processing on board, and the doctrine/tactics that would be employed in the operation of such vessels.
Transmission impossible
Such thinking should incorporate the scenarios when transmissions are impossible owing to the need for covert operations, or when the enemy's pretending to be Bob Marley: they're jammin'! (you know that you've been in this game too long when you have a favourite electronic warfare joke). Anyhoo….
When the enemy's activities in the electromagnetic environment make timely and assured communications impossible, how will the mission be executed? Such systems, to be combat-credible, need to be able to operate in the "digital dark".
Associated with this is the need for the USV to know where it is. This is essential for effective navigation.
It is highly likely that access to satellite-based navigation and timing services will be jammed or spoofed during conflict (it's happening quite a lot already). There are alternatives and ways to increase the resilience, but this too drives cost and complexity into the solution.
Deciding upon this cognitive architecture means having a clear understanding of the doctrine. This is really just a fancy way of saying "do we know what we want to do with these things?".
How do we want it to work?
Ukraine is using USVs and drones very effectively to undertake Sea Denial operations. They are denying the Russian navy the ability to operate its ships in the northern Black Sea.
This is different from Sea Control, where maritime forces exert sufficient influence to allow their own forces to use the controlled area for their own purposes, such as operating a Carrier Strike Group. The problem is a different one, and the tactics and capabilities that you need are different.
Sea Denial means that you pose a big enough risk to the enemy's forces in that area of the sea that they cannot use it.
Sea Control means that you have to reduce the risk to your own forces to a tolerable level to enable them to operate there.
That requires a different level of endurance and reliability from the vessel, including the likely requirement for at-sea sustainment (refuelling on the go), and a much higher complexity of tactical coordination as part of an integrated and, likely, international task force.
None of this is impossible, and none of it means that we shouldn't be attacking the challenge of the Hybrid Navy with alacrity. What it does mean is that the approach should be strategic. Ensuring that the solid foundations of a resilient and robust cognitive and information architecture are in place that allows for scalable and tailorable autonomy with a heterogeneous mix of robotic and autonomous systems.
And we must ensure that those robot boats, ships, submarines and aircraft that we attach to that web of command and control are credible for the task in the environment and against the threat to deliver the effects required for as long as is needed.
As so often with big challenges, if you tackle the toughest problem first, you're much less likely to commit to a course of action that leads you down a dead-end – and much more likely to be able to accelerate towards your goal once the conditions are set fair.
Full steam ahead!
Related topics
Join Our Newsletter
