In June Secretary of Defense Robert M. Gates fired Secretary of the Air Force Michael Wynne and the service's chief of staff, General T. Michael Mosley. The immediate cause was gross negligence concerning U.S. nuclear weapons, but it has been suggested that the deeper reason was Gates' sense that the Air Force had badly bungled the use of unmanned aircraft in Iraq and Afghanistan. Somewhat earlier Gates had chided the Air Force for its unwillingness to drop "old ways of doing business," which he felt denied U.S. forces on the ground the support they needed.
As in the past, the Air Force view has been that control of air assets should be centralized. Central control may seem efficient, but in practice it has often denied deployed ground units the close-in reconnaissance support (perhaps equivalent to close-air support) they need. U.S. Navy P-3s have found themselves making up much of the difference, the result being serious metal fatigue problems that could prematurely ground this fleet. It might be time to re-think the value of centralized control of air assets in situations in which ground action is dispersed and there is no concentrated enemy, as in both Iraq and Afghanistan and quite possibly in future analogous situations. That may have been Secretary Gates' point.
The unmanned reconnaissance aircraft can be part of a further ongoing military revolution, the transition toward picture-centric (usually called network-centric) warfare. In the past, pilots generally found movable targets and then struck them (fixed targets were a different proposition). Now forward observers and unmanned reconnaissance aircraft often identify the targets. Strike aircraft deliver their bombs to coordinates; sometimes the targets themselves cannot be distinguished clearly from the air. Indeed, friendly-fire casualties are often the result of ambiguous target descriptions.
That prompts a question: If we are moving into an era of reconnaissance-led air attacks (because enemy forces are thinly dispersed), and if the pilot's eye no longer matters nearly as much in a strike, how important is the strike pilot's skill? An unmanned aircraft can surely deliver a weapon to the same coordinates. The strike pilot may be much better at evading enemy air defenses on the way in, but an unmanned aircraft can maneuver more violently, and it may even detect enemy missiles more efficiently.
Moreover, a controller directing a strike may be much better situated than a pilot to take account of no-fire zones and even of moving friendly forces (using blue-force trackers). When targets are spread thin, there are many more opportunities for error, and in a war for hearts and minds, errors can be very expensive. This idea is already widely accepted, with efforts to limit ourselves to precision strike weapons and those that reduce collateral damage.
The Role of Humans
It also seems that the future, even in large-scale war, will belong to dispersed forces, which make poor targets for the mass attacks of the past—precisely because previous wars have demonstrated that such forces are often destroyed before they can attack. In a war of dispersed forces, much more effort will go into locating targets in the first place, and those targets will probably be difficult for a fast-moving pilot to identify from the air. Instead, we will continue to rely on a massive intelligence and reconnaissance effort requiring interpretation by experts on the ground. Their product will be target coordinates, just as it is right now in Iraq and Afghanistan. Given systems like GPS, forward observers and troops requesting support will also generally submit their requests in terms of coordinates.
None of this is to advocate a move toward war by machine. Human decision-making is, and will continue to be, key. Humans provide initiative and creativity. The question is where in a weapon system the humans ought to be. Surely pilots are too valuable to be reduced to the role of chauffeurs driving trucks to deliver weapons to preset addresses.
Secretary Gates' frustration with the Air Force seems traceable at least in part to that service's insistence that each unmanned aircraft be flown by a rated pilot (the other services operating UAVs seem not to feel that way). That is certainly not the case with cruise missiles (such as Tactical Tomahawk), which one might see as one-way unmanned aircraft. In their case targeters develop attack plans, often by several missiles working together. Many such missiles can be placed in waiting patterns, their controllers diverting them to targets as required. How different is that from having one controller handle a swarm of unmanned attack aircraft?
Perhaps the main difference right now is that the unmanned aircraft can be recovered to attack again. Under a current contract with Northrop Grumman, the Navy is to test a prototype unmanned carrier attack aircraft, which should be able to refuel in flight. Assuming that is successful, the aircraft would gain enormous endurance, unlimited by a pilot's human weakness.
Pervasive Surveillance
Now go back to the new kind of picture- or network-centric warfare. Imagine a situation of pervasive and almost continuous reconnaissance. Surveillance can never be perfect, but it would be difficult for an enemy to know when he was and was not under observation. No enemy would ever be sure that any particular surveillance system indicated that an attack was imminent. Virtually all attacks would be horrifying surprises, and the strain of always being subject to attack would have considerable psychological effects.
This is not a matter of speculation. Naval warfare using ocean surveillance has many of the characteristics of the new picture-centric warfare, including the strain of continuous vulnerability. You can see its effects in, for example, U-boat operations once the Allies were reading most German codes (after about May 1943). Until that time, a U-boat commander could be reasonably certain that he was safe, once in the open ocean, until he gained contact with a convoy. After mid-1943 he never knew, when surfaced, if an airplane would come out of nowhere to attack him. The effect on U-boat morale was shattering (the situation improved once the Germans could snorkel).
The natural complement to pervasive surveillance is a continuous threat of attack. This means a swarm of armed attack vehicles always available to strike anything the surveillance system sees. Human pilots cannot provide this sort of presence, but unmanned vehicles—particularly if they can refuel in flight—can. We already see the germ of this idea in the form of Predators armed with small missiles like Hellfire. They can orbit for very long periods in areas where enemy personnel operate, continuously scanning the terrain until an identifiable target appears. Predators have been successful because the current enemy seems to have little ability to see them at ranges at which they can be effective, hence the hostiles tend not to evade them. On the ground, their operators have enough data to identify the targets.
In a less permissive environment, unmanned vehicles passing continuously overhead would offer about the same degree of surveillance, but the surveillance aircraft would not be the shooters. In either case, what matters is that the surveillance system as a whole provides a human decision-maker with enough information to conduct a vital attack, within what may be restrictive rules of engagement. It may literally be impossible for a pilot to make such decisions in flight.
A sophisticated unmanned aircraft is about as expensive as its manned equivalent, and it requires about as much maintenance effort. However, it is not subject to nearly as much wear and tear, because it is never flown to maintain pilot proficiency. Nor need it be bought to populate a vital pilot training pipeline. Almost certainly, then, the same budget can buy a lot more unmanned surveillance and, probably, strike aircraft. Alternatively, if we need only so many such aircraft, money becomes available for other purposes.
The U.S. Navy may be particularly well placed to take advantage of such economies if unmanned aircraft prove suitable for carrier operations. A carrier could maintain a pervasive attack capability over most places of interest to the United States, without the need to also maintain a nearby base. Land-based aircraft would require such massive tanker support (for sustained presence) as to be unaffordable. Without the need for continuous proficiency flying by many of her aircraft, the carrier could operate much more economically than at present. She can replenish her weapons with relative ease, and hence can sustain the threat to enemy forces (individual unmanned aircraft would return to the carrier to rearm, and to a tanker area to refuel).
This is the potential toward which Mr. Gates' "new ways of doing business" would seem to point. Students of service cultures will add that, whereas the Navy is devoted to gaining and using control of the sea—including efforts exercised through the air—the Air Force seems much more concerned with maintaining manned air operations. That has meant limited interest in what the aircraft actually do to support operations on the surface. The Navy's broader mandate has made it much more adaptable to past revolutionary changes in military technology. "Battleship admirals" existed more in myth than in reality.