World Naval Developments – August 2018
By Norman Friedman*
At the end of August the U.S. Navy chose Boeing’s MQ-25 Stingray as its first carrier-based UAV. At least initially, MQ-25 is to be a tanker, freeing F/A-18E/F fighters for long-range combat missions.
In the past, the E/F fleet has been worn down by its tanker role, the navy having given up dedicated carrier tankers at the end of the Cold War. At that time the judgement was that manned attack aircraft would generally be used for close air support at moderate ranges. Long range attack, particularly against heavily-defended targets, would be reserved for Tomahawks. This judgement made the F/A-18 an acceptable replacement for longer-legged carrier strike aircraft such as the A-6 and the F-14 ‘Bombcat.’ To some extent the choice to develop the F/A-18E/F was a step back. Carriers are being pushed further from their likely targets, and manned aircraft seem to be needed to strike deeper and better-defended targets. The F/A-18E/F is above all a somewhat stealthy attack airplane, and it is more and more obviously wasted as a tanker. With the growth of anti-access measures and pop-up targets, strike range seems more and more important.
Too, the number of missiles any naval force can deliver is quite limited. We still cannot replenish vertical missile cells at sea, so that after a surface ship has launched her Tomahawks she has to withdraw before firing again. Carriers are still the only ships which can easily refill their magazines at sea, hence are still the only way to maintain a sustained aerial attack. Part of the judgement of the early post-Cold War era was that weight of attack was no longer important; it would always be trumped by precision. Experience with precision weapons has shown that mass still often matters. That was particularly clear, for example, in the U.S. strikes against Syrian air bases. It takes an airplane to carry a heavy load – and to return to strike targets again and again, particularly when attempts to repair them can be monitored.
It seems obvious that the MQ-25 was never conceived simply as a tanker. Photographs show a stealthy shape which provides only limited internal volume for cargo fuel. Presumably most or all of the cargo fuel will reside in external tanks. The stealthy configuration, without any vertical tail, presumably makes carrier landings more, rather than less, challenging. It may be that the main adaptation to a non-combat role has been to omit low-observable materials, which can be added later. On the other hand, the tanker mission seems to have been made for a robot airplane. Pilots are expensive to train and to retain; surely they are better used for combat missions on which their initiative and adaptability to surprises is far more important. Perhaps most importantly, the tanker mission entails long endurance. A UAV doesn’t get tired or bored; its endurance is limited only by the reliability of its parts. Too, the UAV flies only when it is needed, which considerably reduces the maintenance load it imposes. Using a UAV as a tanker suggests other relatively routine uses, such as the ASW search mission previously carried out by S-3s, which was dropped at the end of the Cold War. Now that the navy faces a large Chinese submarine force, it may be relevant again – and UAVs may be a good way to carry it out.
It would seem that a UAV conceived solely as a tanker would be a high-volume high-observable airplane, something like the old S-3. On the other hand, it is possible that low observability would make new kinds of tactics possible. Low-observable tankers could operate much closer to a defended area, for example. An attacker which could be sure of tanking just outside enemy territory could arrive with a heavier bomb load, or could penetrate deeper, or could maneuver more violently.
The navy already operates an armed UAV, although it is not called that: the Tomahawk land-attack cruise missile. Current Tomahawks can be retargeted in flight, and they can pass back target and other images to their operators. The one way in which a Tomahawk is not like a carrier-operated UAV is that it is never recovered. Put another way, manned attack aircraft are a recoverable (reusable) way of delivering ordnance on target. Aside from that, their pilots can respond to surprise situations, particularly when communications links are down. The great UAV question has been and is whether the robot can be provided with sufficient intelligence to continue to function effectively when its links (and, probably, its navigation) fail.
That is a considerable barrier to relying on armed UAVs. A key tenet is that responsibility for any armed act must be assignable. A human must be in the loop somewhere. That might mean assigning a target to a missile, which would identify the target as it approaches before hitting. Where does artificial intelligence fit? What happens when data links fail over a heavily-defended area? To what extent are aircraft given pre-assigned targets, and to what extent is pilot initiative involved?
For some years we have been moving into a world in which reconnaissance systems, including those in space, find crucial targets, which are then assigned to aircraft. Our record with moving targets, such as Scud launchers in Iraq almost two decades ago, may not be very good. In a very permissive environment, the problem did not really arise. A UAV could orbit, its sensor watching (say) a road along which a target car would likely come. When the UAV’s remote crew saw the car, they could order the UAV to strike. There was no question of who was responsible. It is much more difficult to assign responsibility when a target is nothing more than a set of coordinates for a GPS-guided bomb, and those dropping the bomb never even see it. That is now the rule for fixed targets. How is an airplane dropping a GPS-guided bomb on a preassigned target different from a UAV doing the same thing? If the UAV can be made reliable enough, is it as good as a manned attack airplane?
Probably the most important role of the MQ-25 program is as a learning experience: the carrier navy will learn how to operate a UAV at sea. Ironically, perhaps, from a UAV point of view the most difficult part of the carrier operation is what a pilot might consider the simplest: handling the UAV on deck. A human pilot has little trouble following hand instructions to maneuver his airplane once he has landed. A UAV is a very different proposition. It might be easier to set it up so that a ground handler can simply jump on board after the UAV is arrested. The deck handling issue, incidentally, is very typical of what we are seeing in many applications of artificial intelligence. Computer systems can learn to solve structured problems, such as finding cancers on scans (or winning highly-structured games such as chess). Applying experience flexibly seems to be far more difficult, perhaps even impossible. The same UAV which can follow way-points to a target and drop a guided bomb at preset coordinates may find it far more difficult to taxi out of arresting gear and take an appropriate position based on where other aircraft are.
We tend to focus on airplanes rather than on the systems within which they operate. How many readers are familiar, for example, with the way in which a carrier intelligence center works to direct the carrier’s strike aircraft, or the way in which individual pilots are assisted in planning their flights in and out of a target area? If you think of the carrier strike aircraft as part of a larger attack system, the intelligence center is something like the Aegis system directing defensive missiles. This system looms a lot larger when the airplanes are unmanned. We already invest heavily in UAV control systems, including software and links to UAVs. The UAVs themselves are often inexpensive, but the central system is not. Although they are not expendable, UAVs may be most analogous to missiles wielded by a complex weapon system like Aegis. The biggest investment in the MQ-25 program, then, is the new carrier –based UAV control system. It has to combine the sort of mission planning a carrier already does with in-flight control. The further away the UAVs fly, the more the system must work through external connections, almost always via satellites. They impose time lags and the communications loads on them may also impose time lateness. That is certainly already true of UAVs operated over land, but MQ-25 offers high performance which may make a high degree of current connectivity more urgent.
The other difference between carrier UAVs and those which have been so successful ashore is that they operate from a moving environment. It has to be flexible enough to keep track of where its base is. Carrier pilots do that all the time, and GPS helps considerably – unless, of course, it is jammed or otherwise disrupted. We keep hearing about Chinese interest in what amounts to space warfare, which would surely affect GPS. How well would a carrier UAV function if it lost its source of precise navigational information? How well can it adapt to fuzzier information, the kind humans know how to use? The great wartime successes of UAVs have been in very permissive places, but the carrier force is valued for its ability to fight where it definitely is not wanted.
A big UAV like MQ-25 is probably at least as expensive as a carrier airplane. However, it may impose a much lower lifetime cost. It flies only when it is needed; there is no need for proficiency flying. That should considerably reduce its maintenance load. If it never need fly to train pilots, there is no need for a big training pipeline. The production run can be much shorter. Without the need for proficiency flying, a carrier would not burn as much jet fuel, which might reduce the need for tankers in forward areas. Buying a few tankers would not have much impact. If, however, MQ-25 (or its equivalents) took on more roles, then there really would be economic impact. Carriers might become far less expensive to operate – we might even operate more of them, at a time when their strike power is more and more vital.
* Norman Friedman is author of The Naval Institute Guide to World Naval Weapon Systems. His column is published with kind permission of the US Naval Institute.
