Autonomous Vessels and the Future of Maritime Warfare: Integrating AI and C4ISR Systems
Written by Steve Prest, (Commodore (Rtd) Royal Navy).
On 26 May 1941 the hunt was on for the Bismark. The German battleship had, two days earlier, sunk HMS Hood killing all but three of her 1,424 crew. Partly motivated by the need to strike an immediate revenge blow, and partly through the strategic imperative to prevent this major threat to the Allies’ Atlantic shipping routes, the Royal Navy was throwing everything they could at the chase.
It was Swordfish aircraft - propeller driven biplanes - from HMS Ark Royal and HMS Victorious that found and crippled the Bismark, allowing the Royal Navy’s big guns to close their prize and finish her off. Over 2,000 German sailors died in the action. Later in the war, over in the Pacific, there were numerous further examples of the power of the aircraft carriers and their air wings, which replaced the battleship as the unit of currency of naval power. Aircraft carriers beat battleships.
At this point, you may be thinking that you’re reading the wrong article. What on earth has this got to do with the subject: Autonomous Vessels and the Future of Maritime Warfare: Integrating AI and C4ISR Systems?
Steve Prest
Steve Prest is a veteran with his most recent appointment having been Deputy Director of People Change within the Royal Navy.
The next evolution in naval warfare
Before I left the Royal Navy, I was the Senior Responsible Owner of the Mine Hunting Capability (MHC) programme. This programme is replacing the Navy’s mine countermeasures vessels, the Hunt and Sandown Class mine hunters, with a heterogeneous suite of uncrewed and autonomous capabilities deployed from host support vessels. The traditional mine countermeasures vessels were ‘battleships’ – they closed with their targets and used their onboard sensors to detect and then prosecute at close range. The UUVs and USVs, in time to be supported by UAVs, find and defeat the threat at range, reducing the risks to both the operators and the ships, and increasingly the effectiveness of the mission.
In trials, the ‘aircraft carrier’ approach of the MHC programme was shown to be an order of magnitude more effective, more readily scalable, and with a new agility to tailor the capability to the specific mission. The evidence before us is overwhelming – the adoption of uncrewed and autonomous systems at scale will give capability advantage to those navies that can implement and exploit this approach.
The current conflict in Ukraine offers us some clues as to the direction of travel, but whilst remotely operated and autonomous systems have played vital roles in this conflict so far, I would argue that we have seen volume rather than integration at scale. And it is this integration at scale where the real opportunity in Autonomy and AI resides.
Autonomy and decision-making
Uncrewed systems themselves are evolving rapidly. There are increasingly vehicles of various sizes, shapes, complexity and cost to fill all manner of use cases. The technology is advancing rapidly too – materials for construction, design, energy storage and propulsion technology are making significant strides as they are tailored specifically for the UxV market rather than being repurposed from elsewhere.
In terms of payloads, there is important work being conducted in low space, weight and power (SWAP) sensors, processing and communications equipment, as well as effectors (weapons and other devices) to defend our own Force or strike the enemy. The real trick, of course, is in the “smarts”. What I like to describe as the cognitive architecture: who or what is making which decisions where and when, and what information is needed to inform these decisions?
A useful way to think about this is authority that the Navy gives to different ranks in its operations rooms. In the last century, radar detections were plotted manually to generate tracks. These Radar Plotters (RPs in the jargon) were of Able Seaman rank but their role was replaced by automated radar track extractors.
These automated tracks are now fused together by the picture Compilers (Leading Hands). Some of their work is automated, but much of it still requires human intervention. And, after the Compilers, are the warfare Directors (Petty Officer/Chief Petty Officer), and then the Principal Warfare Officers and so on.
The battlespace of tomorrow
To gain the real benefit from such distributed systems, especially in an environment where the electromagnetic spectrum, and thus communications, is likely to be contested and congested, many more of these functions, higher up the rank scale, will need to undertaken and decided upon by the Artificial Intelligence and algorithms onboard of our autonomous systems and in the Force C2 hubs. It may not be possible, for example, to pass back in real-time all of the radar or sonar data of an uncrewed picket, stationed far ahead of the main Task Group. And so, the onboard systems will need to be trained to identify, classify and prioritise the tracks of interest that are communicated to the main Force C2 hub, filling in the rest when the bandwidth is available.
Back at such C2 hubs, whether afloat or ashore, the volumes of data generated by the heterogeneous fleet of crewed and uncrewed systems will quickly overwhelm a traditional Ops Room staff with legacy C4ISR systems. The sifting, categorising and prioritising of data, and the decision-making upon it, in this machine-speed and data-saturated battlespace requires an altogether different tempo of response. This can only be achieved by rapid and AI enabled data processing, bringing to the attention of commanders only what truly matters, when it matters. Given the range and complexity of modern warfare, this must be synthesised on a pan-domain basis. Any force that fails to master this holistic information battle will be bested by the one that does. It is not an optional extra, but the vital ground of modern conflict.
It may even be, for example, that where an autonomous system detects, tracks and classifies an incoming threat missile or drone, that it is authorised to take action against such targets to defend the Force. Maybe this is in the form of Electronic Warfare, or maybe even the engagement of hard-kill weapons. In any case, this form of autonomous surveillance, with the associated decision-making, self-synchronised across the Task Force, will be an essential part of a layered, real-time C4ISR picture of the future. It will not be optional if we are to defeat the range and volume of threats that confront us in the modern maritime environment.
To realise such a vision requires moving away from a platform centric view of the Force, even beyond a data-centric view of the Force and into a cognitive view of the architecture of the Force. Yes: platforms, data, bandwidth, sensors, weapons, etc are all essential to be understood, architecture and managed; but in a world of autonomous, off-board systems it will be the ability to implement mission command in machines that will really give the tactical edge, and the ability to scale.
