Autonomous Technology: Sea control through decision advantage
Autonomous surface vessels can enable decision advantage in sea control missions, but there are challenges to adoption, writes Andy Watts.
Some technologies are so revolutionary that requirements must be matched to technology, not the other way around as in traditional acquisition practice. Examples include breakthroughs in artificial intelligence, machine learning, and edge computing, and now autonomous air, surface, and sub-surface vehicles.
Modern defence practitioners accept this inversion (although acquisition culture may not) and look for ways in which these breakthroughs can be applied to deliver military advantage and better value for money in defence acquisition.
This is not without risk, in that the bureaucratic safety delivered by carefully defined requirements approved across three to four successive layers of authority is not available. But if defence forces are to manage change at the speed of relevance, root and branch change is needed.
This article offers a discussion on the applications of autonomous technology in the naval sphere, including the challenges that must be overcome but that are seldom acknowledged by the more ardent advocates of autonomy.
“The commander’s detailed situational awareness, though supported by strategic intelligence systems, is limited by the range of own ship sensors and those of the other ships and aircraft with which the ship is cooperating, shared via data links. “
Sea Control
Autonomy is a by any measure a revolutionary technology. The capability/requirements inversion referred to above applies; naval practitioners must seek applications for autonomy or be left behind. But the potential of autonomy for New Zealand cannot be realised unless the operational context in which RNZN capability protects our vital interests is considered.
In an article published by the New Zealand Initiative, Major General John Howard (Retd) makes a well-articulated case for sea control capability. It’s a case that is all the more valuable for having being made by a very senior soldier – proof, if any were needed, that the bad old days of single service parochialism are behind the New Zealand defence establishment.
John’s thesis, if I interpret it correctly, is that the free movement of goods and information across and beneath the world’s oceans is vital to our national interests and that we must therefore make a proportionate contribution to defending it. That is what sea control is – defending our ability to use the oceans for our own legitimate purposes.
Sea control also allows us to project hard and soft power by sea – transporting and landing land and humanitarian forces without interference from hostile powers, and it allows us to protect the borders, resources, and ecological balance of our marine environment. In essence, this is what New Zealand naval combat power must do – enable us to exercise sea control in conjunction with partners and in our own right.
Some time ago, I attempted to support an MoD policy paper with solid data that refined the geographical areas in which shipping carrying goods vital to our national life needed to be defended – which obviously extended beyond the sea lanes leading directly to New Zealand.
I couldn’t do it. International supply chains are now so numerous and diffuse that it would require months of access to commercially sensitive shipping data to even begin to identify where cargos destined for New Zealand were being carried in ships. The global supply chains for iPhones begin in 40 different countries, the raw materials for electric car batteries are mined in nine different countries, and their manufacture is similarly diffuse.
We are thus vulnerable to disruption to shipping on the routes to and from New Zealand, but also to disruption to shipping that doesn’t come within thousands of miles of our borders. Defence against this requires sea control capability, both in the approaches to our country (which we will almost certainly have to defend unaided except by partner nation strategic intelligence systems), and in the wider region.
Combat Capability
As stated above, sea control requires naval combat capability. So what should our sea control capability be able to do, and what should our naval combat capability thus consist of? The most important point to note is that sea control is delivered by a capability system, not by a particular type of ship or aircraft. Within the following paragraphs, I offer a grossly over-simplified description of that system.
At the apex are the commanders who determine how resources should be allocated and missions assigned across a theatre, and the strategic intelligence systems that collect, process, analyse and disseminate information enabling multi domain situational awareness across our area of strategic interest. The strategic intelligence system consists of sensors, computing power, and people capable of applying the nuanced judgement needed to make sense of the vast amount of information that sensors and computers can generate. It informs Commanders’ decision making and supports the situational awareness of forces throughout the theatre of operations.
The next layer consists of the tactical HQ (including Maritime Operations Centres (MOC)) and deployed forces that perform the missions assigned by Commanders. These forces are broken down into sub-forces that perform the specific tasks needed to achieve the missions.
For instance, surveillance aircraft, crewed and uncrewed, might be brought together in a Task Group that refines the intelligence provided by strategic systems with more detailed surveillance. Surface forces might be formed in a Task Group providing direct protection to shipping and land forces in transit. Surface forces, crewed and uncrewed, would also contribute to theatre situational awareness as they patrol or pass through a particular area, in addition to using their own sensors to attain local decision advantage over adversaries (see below).
This is the form of sea control task that New Zealand naval combatant forces are most likely to be called upon to perform, and it’s in the context of this task that I hope to be able to explain the relationship between the crewed and autonomous systems that I believe we should be considering.
Decision Advantage
All military operations depend for success on the attainment of what is known as decision advantage, which is a state of cognition in which our own commanders can make better decisions more quickly than their opponents. This depends on both the quality of the strategic intelligence and analysis available to the commander, and the commander’s own ability to make sense of the environment in which they operate.
The commander’s detailed situational awareness, though supported by strategic intelligence systems, is limited by the range of own ship sensors and those of the other ships and aircraft with which the ship is cooperating, shared via data links.
Autonomous surface vessels can greatly enhance local situational awareness and thus decision advantage by extending sensor coverage. They can do this in close cooperation with a crewed combatant, allowing the localisation of potential threats through the triangulation of bearings detected by both the combatant and the autonomous vessel, or through independent operations at some distance from the autonomous vessel’s control combatant, expanding the area under direct and detailed surveillance.
Independent autonomous vessel operations can also be carried out under the control of other parts of the sea control system, including shore based tactical HQ such as Maritime Operations Centres (MOC). Strategic intelligence systems provide an overarching view of our Extended Continental Shelf (ECS), Exclusive Economic Zone (EEZ), and Territorial Waters (TW). This overarching view is amplified and refined by information collected by P-8A and, in the future, autonomous aerial vehicle patrols.
A long-range autonomous vessel could complement these capabilities by close and continuous monitoring of targets of interest detected by strategic and aerial systems, pending interdiction by a crewed ship should this be necessary. It could also carry out long duration surface patrols, which while nowhere as effective in terms of coverage as a crewed or uncrewed aircraft, could focus on areas of special interest, such as a marine protected area, relieving some of the burden on aerial assets and much of the burden on crewed patrol ships.
The other role that has been widely touted for autonomous vessels is the provision of offboard magazines. Vertical Launch Systems (VLS) fitted to surface combatants (including the Anzac class) offer very rapid and efficient missile launch, but they are all but impossible to replenish at sea. Autonomous vessels equipped with VLS and controlled from a surface combatant increase magazine depth, which is likely to be an issue in high intensity combat.
Finally, the mere presence of autonomous vessels in a sea control force complicates the adversary’s surveillance and targeting problem by increasing the number of force elements that must be detected, localised, classified, and monitored. Our own and our partners’ forces will almost certainly be numerically outmatched in any period of tension or conflict; the part played by autonomous vessels in offsetting this could be significant, whether the adversary is a state actor or a criminal organisation engaged in drug or people smuggling.
I would rank the near term potential applications of autonomous vessel technology in the following order of significance (noting that they are already extensively employed for expeditionary reconnaissance and mine countermeasures):
- Enabling sea control in our TW, EEZ, ECS and those of our partner nations with focussed, long duration surveillance operations complementing other elements of the sea control system, including strategic systems, crewed and uncrewed aircraft, and crewed interdiction vessels.
- Enabling sea control on deployed operations by extending combatant sensor coverage.
- Complicating an adversary’s surveillance and targeting problem.
- Assuming challenges around the protection of control and communications data links are overcome, by increasing combatant magazine depth.
Challenges
And now to the challenges of autonomous. Firstly, in line with almost every other UN member state, New Zealand is a signatory to the UN Convention on Certain Conventional Weapons (CCW), and has endorsed the US-led Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy. As I read it, a fully autonomous warship governed by AI and pre-mission programming is currently inconsistent with nation states’ obligations under the CCW and the Law Of Armed Conflict (LOAC). Autonomous vessels must be under human control such that any decision to deploy lethal force is subject to that control.
Secondly, the effectiveness of an autonomous vessel depends on radio frequency communications and control links that are highly resistant to disruption or compromise. An autonomous vessel patrolling an EEZ could revert to pre-planned mission control logic if communications are disrupted, but links would need to be restored for it to remain fully effective. However, an autonomous combatant consort extending sensor coverage or providing magazine depth would need even more robust communications links.
This could potentially be achieved using the light spectrum and by keeping the autonomous vessel within the range of line-of-sight communications (which would limit its effectiveness), but the challenge involved cannot be assumed away. Dependence on radio frequency communications also creates a vulnerability to cyber-attack which must be allowed for in autonomous vessel capability development.
Thirdly, it could be argued that the current state of maritime autonomous technology does not allow an autonomous vessel to comply with the International Regulations for the Prevention of Collision at Sea (COLREGS) with the necessary degree of assurance unless it is under direct (albeit remote) human control. This applies particularly in areas of high traffic density. However, significant advances are being made in this area, and it is possible that COLREGS compliance requirements could be satisfied in the future.
Fourthly, an autonomous vessel is a vessel. Vessels need maintenance, and they break down – often at the worst possible time. An uncrewed vessel operating in our EEZ could be provided with a ‘limp home’ mode in the event of breakdown and the necessary propulsion redundancy. However, an autonomous consort to a combatant will have to deploy over the same ranges and in the same environmental conditions as the combatant.
What would happen if it breaks down on a surveillance task 200 miles from its parent combatant, particularly if communications are lost? How would the risk of autonomous vessel breakdown in environmentally sensitive areas such as the Great Barrier Reef be managed? How would damage caused by high sea states be repaired?
Finally, the autonomous vessel will carry a sensitive, classified payload. How is it defended against attack, including boarding and seizure by hostile actors at home or abroad?
I would rank these challenges in the following order of significance:
- CCW and LOAC compliance.
- Maintaining and repairing on task autonomous vessels and payload systems.
- Preserving communications data and control links with autonomous vessels in a contested electro-magnetic environment.
- Cyber protection of autonomous control and payload systems.
- Defending autonomous vessels from attack and seizure.
- COLREGS compliance.
In Conclusion
Notwithstanding the challenges identified above, uncrewed and autonomous systems are becoming an increasingly important component of the sea control system. It is probable that at least some of the above challenges will be addressed in the near future, given that advanced partner navies have set in place programs for the wholesale adoption of advanced autonomous systems. The Australian defence industry, including companies such as Austal (to which I consult) and Greenroom Robotics, is heavily committed to developing autonomous solutions to maritime operational problems.
In the here and now, autonomous surface vessels can play a part in attaining decision advantage in sea control missions, particularly in our own EEZ and those of our Pacific partners. But wholesale adoption on the scale advocated by some depends on progress in addressing the challenges identified above.
