On 16 December 2008, Northrop Grumman rolled out the prototype of the U.S. Navy's Unmanned Combat Air System Demonstrator (UCAS-D, also known as the X-47B). As a proposed carrier aircraft, its two critical tests are deck operations (including takeoffs and controlled hooked landings) and air-to-air refueling. The X-47B is about the size of an F/A-18E, with broadly similar performance and capabilities.
Unlike the current fighter, it is intended to be stealthy, so in theory all its weapons must be carried internally in its two bays. Visually, the X-47B differs radically from existing fighters, with an all-wing shape and without any vertical tail surface to reflect radar signals. Part of the demonstration is to show that such a shape can be controlled precisely enough to land safely on board a rolling, pitching carrier.
Assuming the X-47B proves itself, it has the potential to transform carrier aviation. Manned aircraft are limited by the endurance of the human pilots. We have already seen in Afghanistan that badly fatigued pilots can make misjudgments fatal to friendly troops on the ground. Too, a pilot in the air has only limited knowledge of what is happening on the ground spread out around him. His task becomes even more difficult in the kind of counterinsurgency wars the United States now finds itself fighting because there is no obvious visual cue to distinguish enemy from civilian. Matters are even more complex when the enemy deliberately hides among civilians.
Those dispatching aircraft on missions have the benefit of melding intelligence data to create more precise pictures of what is happening on the ground, while an individual pilot has only the ambiguous evidence of his eyes. Those assigning targets, usually by their GPS coordinates, are the best potential decision-makers. When targets are designated only by their coordinates, they are the decision-makers. The pilot's role in the attack seems less essential than in a past of imprecise targeting and much more prominent visual cues. It is by no means clear how often the pilot's own on-the-spot judgment affects how well the target is struck. The pilot's role, for example in realizing that he has to sneak a weapon past obstacles, is likely to shrink as reconnaissance coverage of areas improves.
One lesson of current wars is that all available imagery must be melded to produce the kind of information forces need. Surely the right place to do that melding is in some sort of command center, where large-screen displays and reference information are at hand. Any sort of remote control imposes latencies, but they are probably least important in the sort of ground attack UCAS-D is conceived to do (air-to-air combat may be another matter).
Greater Range
UCAS-D or an equivalent offers a carrier much greater strike range, assuming that the aircraft can be refueled in flight, since it could remain on station within striking distance for many hours at a time, its endurance set by the reliability of its components. The threat of such a virtual air base ought to have significant effects on enemy behavior. With manned aircraft, particularly as air arms shrink (as costs rise), the threat of aircraft is intermittent. If, for example, the air threat limits an enemy's mobility, that enemy can move as soon as a major air strike has been carried out. In Afghanistan the solution in 2001 was always to have some aircraft within a few minutes' time of any potential enemy operation. A more sophisticated enemy would have coordinated operations to saturate this coverage. Coverage in turn was limited by the endurance of the aircraft overhead, usually far from their bases. Although there was little attention to the problem, fatigued pilots tended to make mistakes, the best known being an attack on Canadian forces conducting a live-fire exercise in April 2002.
Less notorious were cases in which aircraft reflexively bombed families celebrating happy occasions by firing their automatic weapons into the air. In the aftermath, the typical cry was that they were in friendly parts of the country—but a pilot flying overhead does not see prominent borders between friendly and unfriendly, and the displays he has tends to be simplified so as not to overload him. A remote controller would work directly from a map-like display into which available intelligence had been integrated. A sensor on the unmanned aircraft would still be able to register ground fire, but the display would make it much easier to understand what that fire actually meant. Such understanding might be particularly important in the sort of counterinsurgency campaign we are now waging.
Lower Cost
Assuming UCAS-D functions as hoped, it can have a massive impact on the cost of naval aviation. Ultimately, as Admiral Joseph Metcalf III used to say (when pressing for a Revolution at Sea in the 1980s), it is a matter of ordnance on target, of how far away the target can be, and how efficiently it can be bombarded. Obviously carriers offer a lot more, including the ability to neutralize an enemy air force, but UCAS-D is a strike bomber. In its case, or in the case of an F/A-18, how much does it cost, in the end, to drop a given amount of explosive exactly where it should go?
The manned airplane carries a great deal of overhead. Landing on a carrier is so difficult that pilots must practice it continuously. The bulk of carrier air flights are made to maintain proficiency, both of the pilot and the deck crew. All such flights exact much the same wear and tear as are experienced in combat, apart from battle damage and some damage due to violent maneuvering. Pilots cost a great deal to train, as does combat search and rescue to save them if they are shot down.
Now imagine the economics of an unmanned air vehicle. It flies only when it is needed, which is a small percentage of the number of flights made by a manned fighter. If it can be refueled, it will remain in the air much longer, so that landings and takeoffs, which are extremely stressful, would be much rarer in relation to flight hours. At least in theory, the cost of maintenance per flight hour is significantly less for an unmanned vehicle. There is no need to reserve part of the force for training.
Because it is about the size of a current fighter, UCAS-D probably costs about as much (it is often said that airplanes are bought by the pound). An unmanned aircraft ought to be somewhat less expensive, because it does not need life support systems, and both it and the manned aircraft need much the same sorts of mission computer and fly-by-wire controls. It is an open question whether an unmanned aircraft needs anything like the sensors of a manned one, since the sensors sending it to a target are the constellation of remote intelligence, surveillance, and reconnaissance systems.
Eliminating pipeline training requirements would cut production, or would release more airframes for combat if numbers were not cut. Cutting the number of sorties to (not by) 10 or 15 percent (by eliminating proficiency flying) would have enormous impact. That would include the indirect effect of reducing the need for tankers to top up carriers every few days (and the tankers will need escorts when the United States again faces a blue-water enemy). Stealthy composite structures are difficult to rebuild, so anything that reduces wear on airframes would surely be well worthwhile as a means of extending airframe life.
Naval aviation has been an enormous success because it combines the vital naval attribute, mobility, with long-range striking power. With its carriers, the U.S. Navy greatly outperforms navies limited to missile ships because its long-range attackers return to the carriers to rearm and strike again. Carrier strike aircraft are sometimes characterized as reusable stages of missiles. Nothing in all of this requires that the aircraft be piloted, only that they be able to fly off a carrier and return to it. If they can do so more economically than in the past, and at greater ranges, they may be heralding a future in which carriers are more rather than less useful, and in which their cost (per mission, per attack) is reduced.
A kind of carrier transformation has been ongoing for some time due to the rise of precision attack and the less-heralded rise of constellations of sensors that provide the attacker with precise target locations. Navy briefers have pointed out the shift from past carriers capable of striking one target a day to current and future ones capable of striking nearly 200 each day. It would seem that UCAS-D or its equivalent, if it works as planned, should extend this transformation further in going from targets struck to enemy targets held at risk by long-endurance precision attackers.
The reason for building a demonstrator is that the required performance has not yet been attained. No unmanned vehicle has yet refueled in the air. Carriers routinely guide aircraft on board in bad weather, so the problem of landing has likely been solved. The most difficult problem may be maneuvering the unmanned vehicle around the flight deck. Human pilots react to a remarkable variety of cues, some of them quite subtle. It would, however, seem strange if the promise of unmanned combat flight failed not because the airframe could not conduct some exotic maneuver or fly in a required formation, but because it could not follow cues from wire to servicing ramp.