This article was originally written for the Wavell Room as part of their Maritime series in October 2018. Link.
Warfare is an inherently human endeavour, yet the rate of change in the design, application and capability of unmanned systems has the potential to place humans at a far greater remove from the conduct of warfare than ever before. This article attempts to summarise the key challenges faced by the introduction of unmanned systems, specifically by the Royal Navy, and ultimately aims to present a vision of how we might best incorporate these systems into our concept of operations
A question that is currently gaining popularity is ‘how do you man an unmanned Navy?’ I would argue that this is the wrong question; to be able to consider that, ask first ‘how do you unman a manned Navy?’
The Limitations of Autonomy
To paraphrase Werner von Braun, the human mind is the best computer we can place in a complex vehicle. Given the fundamentally human nature of warfare, in the extreme case automation risks making warfare inhuman and we are already seeing rising public concern over the deployment of unmanned systems, in particular those we now arm. Awareness of these factors should remain central to the development and introduction of any unmanned system if it is to be used ethically within our current laws of armed conflict. It does, however, raise the question of how much autonomy we can, or need to grant.
Consider the construct of a ship’s crew and you see how much autonomy we delegate amongst humans. In past times an Able Seaman would plot radar contacts, the Leading Hand would assess them and evaluate the track, the Petty Officer would classify the track and the Warfare Officer would act within their delegations to deal with it. If that resulted in a missile engagement, the weapon itself carries out a limited form of the same process but within tightly constrained limits. If we are to exploit the potential of new technology, and manage wider public expectation of the employment of unmanned and autonomous systems, we must ask ourselves what rank and authority we are granting our autonomous systems.
At the lowest level, a system will merely collect information and may (or may not) stream it back to the end user; raw data, video, etc. In this situation, the data processing will then take place at the human end (or perhaps an interim unmanned step). More complex systems may be able to carry out a degree of processing, something we are comfortable with in modern systems (current systems are more than capable of track classification); as a result a smaller quantity of more refined data can be transmitted, so long as you can move the autonomy into the platform. Some data may still need human processing, such as synthetic aperture radar data or complex acoustic data. At the highest levels, a system could potentially detect, classify, track and engage a target – thus removing the human from the kill chain completely. Thus depending on the task, environment and presence (or otherwise) of weapons, the autonomy ‘slider’ needs to be carefully calibrated.
Autonomous navigation and sense-and-avoid systems exist but fully autonomous manoeuvring, up to and including evasion is still likely to require human input. The degree of autonomy will increase or reduce the amount of direct control necessary. As an aside, almost all systems rely on GPS to a greater or lesser extent; this is both an enabler and a vulnerability that must be addressed.
Artificial intelligence may enable greater autonomy to be granted to unmanned systems, enabling onboard data processing, autonomous manoeuvring, resistance to jamming/deception, etc. However, this will demand greater hardware and power requirements and may not necessarily represent a zero-sum gain.
The Communications Problem
Forward processing in any unmanned system is clearly attractive, as greater processing inevitably reduces the amount of information that must be communicated back to the end user. Experience shows us that communications is one of the most critical and challenging aspects of warfare, often overlooked until the point of failure (at which point we all know about vocal officers demanding comms!). The communications spectrum is more and more contested, and the military does not have free-reign to transmit whatever and whenever it likes.
Satellite-based networks using SHF frequencies have the advantage of being line-of-sight for security and can support relatively high data rates. It is, however, by far the costliest option and only the United States really has the resources to operate predominantly on satellite connectivity. The UK, with its limited national SATCOM capability, may not be able to rely on satellite as a future pathway; however, the UK does have a growing space industry, with cubesats and similar development opening up the potential for improved secure communications.
As you move down the spectrum, through UHF, VHF and into HF bands, the reducing frequency and increasing wavelength reduces the data rate; these signals are also move vulnerable to atmospheric attenuation. Broadcast networks, which are omnidirectional, already heavily loaded and contend with civilian signals , are challenging to establish and maintain. Encryption, burst transmission and spread-spectrum techniques can help, but broadcast radio networks are susceptible to jamming (both broadband and targeted) and potentially to exploitation, especially command links, as well as potentially compromise security.
Underwater communications based on sound can be long-ranged with sufficient power but also suffer from environmental degradation and far greater noise levels. They are also limited to low data rates, and cannot breach the air/water barrier easily although experimental techniques are in development. In order to be effective, a surface communications node is likely to be necessary for some time to come.
Possibilities that could enhance the ability of manned and unmanned systems to cooperate, such as laser communications (see LiFi) or quantum cryptography and communications, are in development. All will need to be evaluated as they emerge, but are unlikely to fundamentally change the need for the physical platforms or the sensors loaded onto them.
Generally unmanned systems bring advantages in size and weight by removing the human element, reducing performance requirements and generally focussing on a single capability rather than several. They also do not suffer from crew fatigue or seasickness which is one of the most significant constraints on operations. However, they are still subject to the same limitations of manned systems in terms of power density and efficiency as well as methods of launch and recovery, and are not as flexible or adaptable as manned systems.
Fixed wing aircraft are more efficient than rotary winged ones; they are simpler, lighter and achieve better power-to-weight ratios. However, they require runways (or long flight decks) to operate, unless they are of a sufficiently small size that they can catapult launch (such as Scan Eagle) or exist as compromise designs (such as tilt-rotors like the Boeing Eagle Eye). Rotary wing platforms are less efficient but can carry greater loads than small fixed wing systems, and the ability to launch and recover from confined areas is a significant advantage. As unmanned aerial vehicles and systems (UAV/UAS) evolve, we are seeing ever greater endurance times as light helicopter derivatives evolve.
In the maritime environment, small surface vessels are inexpensive and can endure moderately high sea states; furthermore, a wealth of propulsion options exist from wind-powered craft down to diesels and pure electric. Larger vessels offer greater endurance for conventionally powered platforms, but above a certain size the vessel will no longer be recoverable to a larger one and must operate from a port, or else be designed to be refueled and repaired/maintained at sea. Submersible vessels can operate more freely and more stealthily than surface vessels, but have their own power and speed issues (generally restricted to battery power) and are often much more difficult to recover in even moderate sea states, which you must to repair/refuel. Shifting to complex propulsion systems (such as air-independent propulsion and nuclear) introduces a heavy maintenance burden that autonomous systems cannot manage. Complexity = cost, and if cost is a key driver for the introduction of unmanned surface vessels (USVs) then you will self-restrict your choices to smaller platforms.
In terms of payloads, I will briefly cover sensors and electronic payloads; weapons are too numerous to list. Optical sensors are continually evolving and now include multi-spectral and hyper-spectral capabilities in lightweight mountings. There are also advanced concepts such as ViDAR and laser radars, including laser systems that can reach underwater for survey and mine detection. ViDAR in particular is a good example of how processing can be shifted forwards to the platform. Radar is diversifying from simple mechanical scanners to multi-panel AESA systems, with lightweight systems such as Osprey leading the charge. These systems are also capable of synthetic-aperture and inverse synthetic-aperture operation, providing moving target indication and mapping functions. Other sensors include electronic intercept systems, sidescan sonars, lightweight active variable depth sonars (as fitted to helicopters) and passive detection arrays, some of which can be combined into a single array (such as the GeoSpectrum TRAPS system). Lastly, do not overlook environmental sensors that provide update meteorological and oceanographic information that optimises all of the above to improve detection.
Other electronic payloads include communications relays and nodes, essential for establishing line-of-sight networks without needing satellite relays. Furthermore, electronic warfare systems, including jammers, can provide additional capability in this increasingly important and connected environment.
Jam Tomorrow, or Marmite Today?
The Royal Navy, and the UK Armed Forces in general, are cash-poor; every new capability must be balanced against a tight budget and competing priorities. However, unmanned systems make a powerful case for significantly enhancing the potential of manned platforms, by extending specific weapons and sensor capabilities at lower cost than a manned platform. We must avoid over-specifying our requirement or looking too far ahead, as the ‘jam tomorrow’ promise brings less utility than perhaps the ‘Marmite today’ alternative.
We should be exploiting available systems now, in order to understand the potential and gain operating experience. Where we can, we should drive ruthless (but not ruinous) commonality and accept that this may constrain some of our desired choices; speed and developmental agility has its advantages over mass. We should exploit appropriate and relevant platforms, avoiding the gold-plated iPhone when a cheap Android alternative will do the job – we can and should leverage UK scientific expertise in QinetiQ and our universities and tech companies, developing new modular payloads and spiral upgrades. An 80% solution today offers more utility, and more opportunities to build experience, than a 95% solution tomorrow. And we must not forget that launching and recovering these platforms also drives a high demand for commonality in the necessary equipment and support arrangements.
We also need a more collegiate approach to system development. Most of the systems below will have equal value to the RAF and Army, especially in ISR, distributed lethality and cargo delivery. The RAF has plenty of experience with the Reaper system although such MALE systems are unlikely to become ship compatible; equally the Army has developing experience with Watchkeeper which is a close relation to the Searcher concept I envision below. The RAF also owns the Space environment and, together with the emergence of lower cost satellite and HAPS systems, has much we can benefit from.
A detailed study of potential options for RN unmanned capabilities is beyond the scope of this article, but in considering some of the currently mature and developing capabilities, the following shortlist is not unreasonable (forgive the names; it helps to conceptualise the ideas).
Unmanned Littoral Operations:
- Searcher. An ISR capability based on a small, fixed wing (for endurance) UAV family able to deploy from a wide range of platforms. This should provide medium endurance, low-medium altitude ISR that is particularly useful for littoral forces including amphibious operations. ScanEagle/RQ-21 Integrator or similar. Enables longer ranged detection of surface contacts and over-land surveillance with a relatively low signature; indeed, signature reduction should be desirable where feasible.
Sender. A communications node – a fixed wing (for endurance) UAV that could be launched from larger platforms or from land. ScanEagle/RQ-21-class UAVs could provide this, or larger long-endurance platforms (such as the eye-wateringly expensive Global Hawk, or the Airbus Zephyr already in MOD service). Enables networked communication at lower powers without relying on satellite connectivity, over local to regional distances. It is possible that an airship might have utility, but its low speed and conspicuous nature rather count against it.
Sidearm. A short ranged weapon carrier – a UAV that can enhance defensive surface warfare firepower, act as ASW weapon or sensor carrier, or provide short ranged ISR (such as overwatch for a boarding operation). Likely to be a rotary winged platform to maximise payload over endurance and operation from multiple platforms. Risks overlapping capability with the Searcher concept, but could be designed to operate in partnership as well as with manned assets such as Wildcat. The FireScout or Scheibel Camcopter are good examples.
Supplier. A short-ranged logistics enabler that can supply small items quickly for defect repair, either within a task group or possibly over the beach. Amazon delivery drone or similar; there is plenty of civilian potential at low cost. This would also be capable of conducting aerial inspection of ships structures inaccessible whilst at sea. Potentially could be combined with the Sidearm concept above.
Scout. Tactical reconnaissance for amphibious forces. A variety of small quadcopter and nanocopter derivatives, as already demonstrated by the USMC. Ideally partnered with a concept such as Searcher, again as already demonstrated by the USMC.
Sentry. A USV, possibly a semi-submersible design, that could provide persistent forward deployed signals intelligence collection opportunities. Appropriately camouflaged and designed with low-observable masts or superstructure, such a craft could operate in the same manner as an SSN or SSK over long periods ahead of littoral operations.
Subverter. A concept that envisions a family of vehicles that operate as persistent tactical decoys. Most applicable in amphibious operations, a carefully tailored force could deploy a range of static and mobile decoys, both active and passive (including potentially jamming and providing comms deception), that could enable deception operations to disguise true intent and targets (think Operation FORTITUDE but deployed forwards).
Unmanned Anti-submarine Warfare:
Shark. A USV that can operate as a tactical sonar sensor, above/underwater communications relay and underwater environmental data gatherer. Depending on whether this is best suited for open-ocean or littoral operations, the size will vary according to the desired endurance. Such a platform potentially offers greater potential for multi-static active sonar operations. There is potential hull commonality with other littoral USV concepts. The Elbit Seagull is an example, though larger hulls may be necessary for the open ocean as the US Sea Hunter  is demonstrating.
Manta. In all forms of ASW, it is essential to gather accurate, long-term oceanographic data in order to best exploit the sonar conditions for ASW. This platform will necessarily be of longer-endurance than the USV above with consequent limitations on speed. Many examples already exist, including those developed by NATO CMRE and civilian research agencies. These should be deployable by air or by surface, in a similar manner to a torpedo.
I have previously looked at the ASW potential in a separate blog post.
Unmanned Mine Countermeasures and Survey
Unmanned systems in the mine countermeasures field are already well advanced. Existing platforms include wide area surveillance devices (such as those from Hydroid) and shallow water sweeping systems (specifically ARCIMS); however, a shallow water hunting system is arguably necessary to achieve high levels of confidence and concepts already exist. USVs can operate less conspicuously than large manned surface ships, allowing the mothership to sit out of sight over the horizon and network the remote capabilities at range; here, a higher level of autonomy is desirable. Future capabilities include the ability to deploy disposal systems remotely, and more advanced disposal systems that leverage existing land-based counter-IED technologies (which are much more challenging in the underwater environment, if only because of the constant movement of the water).
To bring my thoughts together, I submit the vignette below as an illustration of how we might achieve an enhanced capability in the near future. It is intended to illustrate how we could create capability quickly, using existing or near-future platforms .
The Littoral Task Group Commander sits offshore onboard HMS BULWARK, just over the horizon and radar silent, waiting for the first reports from the raiding party. His combat system plot shows blue force positions and status updates. Evidence of enemy activity up and down the coast is still abundant, flushed into activity by the decoys seeded into three different potential amphibious operating areas and supported by deception strikes from HMS PRINCE OF WALES group further offshore and well to the north. Two of BULWARK’s Searcher UAVs were still airborne, the last launched just an hour previously, scouring the coast for enemy movements and listening for any potential chatter that might provide warning to the raiders. A third Searcher is tracking an enemy destroyer to the south; the LSG’s Type 31’s have it covered, missiles already programmed with coordinates and constantly updated. The escorts have already swept the area for hostile SSK’s well before the LSG moved into position, and maintain their patrol with their Shark USVs in formation; two Manta drones were still trolling slowly north and south of the group. HMS SPEY has recovered her remote minesweeping and minehunting vehicles having cleared the boat lanes to the beach, and is back in the close Force Protection station on BULWARK, two armed Sidearm UAVs ready on her flight deck and her own weapons trained into the night. She had been here for a week already, slipping into the coast in the dark and deploying her remotes without exposing herself. A Supplier buzzes past the Bridge, ferrying an urgent spare to the beach-head; the message sent half an hour previously, relayed through the Zephyr battlefield comms node, had been processed quickly.
Action! The raiders have destroyed their objective; they are en-route back to the ships. They have hostiles in pursuit; the southerly Searcher has them on camera. Coordinates are swiftly transferred; the nearest Type 31 lights up with rapid salvoes inshore from its main gun. No sense hiding now; the task group goes active, radars flooding the area. Two craft turn towards the group; hostile FPBs inbound. SPEY actions her Sidearms; a Wildcat lifts from the nearest frigate, and the group homes in on the threat, neutralising it with a salvo of Martlet missiles. Now the Marines are inbound; the two Merlin troop transports dust off and the Wildcat veers in to cover the landing craft now powering off the coast. Two F35 are en-route to cover the force, their diversionary mission complete. Missiles flash out from the southerly Type 31; the destroyer is alerted and closing, but with only radar bearings has no solution of its own. The enemy is off-guard and trying to cover a wide front; the LSG recovers its forces in swift order and sails clear, mission accomplished.