Tumbleweeds cleared" reads the sign by the side of the road, a fair indication of the nature of local enterprise. This is the desert east of Palmdale, California, centre of the US high-tech aerospace industry. From the scruffy landscape rise sheer metal buildings housing secret aircraft. To the north are Edwards Air Force Base and Air Force Plant 42, where Northrop hatched the B-2 bomber. A few hundred yards away is Lockheed Skunk Works, birthplace of the F-117 stealth fighter.
But now I'm driving past old ranches with corrals jury-rigged from wire and discarded doors, beneath a sky as wide as the plain, looking for a very different kind of airplane. Then, from a long way off, I see it: a tiny white fleck that comes closer, turning into what looks like a giant white paper airplane, with wing tabs turned down, wheeling over the small airfield set amid greenery and desert. But the strangest thing is its nose: it has no cockpit, no windows. It looks blind; I think of a blind cave salamander, but with wings.
There are no windows in this airplane because there is no pilot. This is Predator, the most recent "unmanned aerial vehicle" to fly for the US military, being tested at the desert airstrip of General Atomics Inc., Predator's builder, in El Mirage, California. Planes without pilots may be the right stuff of the future - a new, digital air force. For a long time, we didn't even have good names for these things. Once, they were dismissively called drones, then remotely piloted vehicles; currently, the term of choice is unmanned aerial vehicle - UAV. Now a new generation of these machines is arriving, relying on advances in electronics and computing, miniaturised sensors and cameras, and digital relay systems.
Ranging in size from a few feet to 150 feet (46m) in wingspan, UAVs are no longer the glorified model airplanes of the past. They have sleek, radar-eluding designs rendered in carbon composite. Some have great, sweeping wings and thin, bird-like fuselages. Others are shaped like manta rays or sea skates. The advent of massive, Cray-level computing power for optimising the design, and of dramatically smaller computers for controlling the flight of these naturally unstable frames, has spurred a revolution in robot craft: software development has pushed close to pilot wetware. And new sensors such as synthetic aperture radar give the spy versions of such planes better "eyes."
The most sophisticated of these machines are self-contained, requiring less and less control from the ground. Their flight plans are written in ROM and are overridden only in exceptional circumstances. Others are flown by ground-based joystick, like a plane in an arcade game, with feedback from on-board radar and video. Like a "real" pilot, the operator of a UAV may have to deal with air-traffic controllers, but he or she will sit in a trailer or a hangar, plotting a mission in advance or responding to video images transmitted from hundreds of miles away.
The Predator is directed by a joystick mechanism, and many of its best pilots, including several women, are those who have flown helicopters. For Pioneer and Hunter, two simple short-range UAVs, there are two "drivers." An "external" operator stands beside the runway, like a model airplane handler, to get the craft airborne and on its way. Then the "internal" pilot - with a game-like joystick (plus pedals, in the Predator) and a video screen tied to a camera on the UAV - takes over for the duration of the mission and returns the craft to the general area of the recovery zone. (Some UAVs are retrieved in nets rather than by conventional landing on runways.) Other versions are programmed to roll out of the hangar, take off, fly their mission, land and return to the hangar without human intervention, though they can be rerouted in mid-course. The interface for these UAVs is far less direct than Pioneer's: it is a series of maps and graphs of way points, modified by mouse and keyboard.
All of these unmanned vehicles are spy planes, though in the future they may also be fighters. Hovering high above unfriendly countries, today's UAVs can relay video, radar, or infrared images via satellite to distant ground stations, watching for anything that moves and taking over the role of manned reconnaissance. The Predator, for instance, can fly more than 2,000 miles (3,200km) round-trip and spend up to two days in the air, where it is virtually invisible to the human eye and hard for radar to spot. The craft is shaped so that despite a 50-foot (15m) wingspan, it shows up as a radar "signature" of only 1 square metre.
A few weeks before I saw them at El Mirage, Predators had shown off in a Scud-spotting exercise as part of Roving Sands - the military's largest field exercise of the year - flying 25 flights in 26 days, high above the simulated battlefield.
A couple of days after my visit, four Predators were on their way to the former Yugoslavia to conduct nearly round-the-clock observation of forces on the ground. Built by General Atomics, a company whose primary business lies in the area of nuclear energy but which is increasingly focusing on UAVs as a technology of the future, the Predator used to beam its information to a chase aircraft or ground station but now relays signals via satellites.
Since the autumn of 1994, the US Army has been testing the Predator and training its pilots at Fort Huachuca, Arizona, near the US-Mexican border - not far from where airplanes themselves were used in war one of the first times, for the excursion against Pancho Villa in 1911. The plane's initial wartime role was reconnaissance, and few at that time anticipated it would be useful for more. But it was soon clear that military aircraft would take on more aggressive service, and many today are just as certain that one day UAVs will move beyond recon to act as fighters.
But will unmanned aircraft ever be able to manoeuvre with the agility of a fighter? The evidence is convincing.
Already, encounters between fighters and the missiles that seek to kill them match machine against man, and Air Force officials, according to Aviation Week, have become concerned about an air-to-air missile gap. There's a need, some argue, for cheap, smart missiles that can use radar and infrared seekers to respond rapidly to evasive manoeuvres. On the current trend of development, some generals fear, a US$200,000 (£125,000) missile will soon make a $20 million (£12.5 million) fighter aircraft obsolete.
Similarly, UAVs can develop faster and drop in cost more readily than fighters. Already, the names of these craft - Hunter, Raptor, Talon and Predator - hint at an aspiration to be real combat planes, and some think it's quite likely that by the middle of the next century they will take over the role of the fighter pilot. The joystick in the cockpit may be replaced by one on the desktop, and Top Gun may be replaced by Captain Nolo - traditional Air Force lingo for "no live operator." Even remote driver Captain Nolo is on course to vanish: more sophisticated sensors and onboard computing power could make future remotely piloted vehicles entirely autonomous in dog- fights.
Out in Palmdale, as I watch the Predator land, a pickup truck emerges from a hangar and heads toward me. There is no law against looking, but I still think it wise to hop into my car and head off. Government agencies have always made me uneasy. Glancing nervously in my rear-view mirror, I take the road back west, past Blackbird Air Park, a strip of grass near the Lockheed Skunk Works where examples of its proudest achievements are parked: the A-12 and SR-71 Blackbird aircraft.
Along with the U-2, these planes made Lockheed - now Lockheed Martin, which recently announced the acquisition of Loral Corp., in this age of military-contractor consolidation - the company the CIA and the Pentagon turned to most often for their eyes in the skies. But beside the long, sleek black shapes of those planes sits another, often over- looked craft, a curious drone called the D-21. This pilotless plane was dropped from beneath the wing of a B-52 to overfly China in the early 1970s, looking for evidence of nuclear testing and troop movements. But it didn't work well - one crew member died trying to launch the thing - and, in a sense, the D-21 was an idea ahead of its time.
Up ahead loom the huge gray and blue buildings of the Skunk Works. I have come to the high holy of American aviation to watch the unveiling of DarkStar, the Pentagon's newest UAV, soon to begin flight testing at Edwards Air Force Base. These days, aerospace companies must sell themselves hard, and inside the hangar called Building 602, the Skunk Works has prepared a bit of theatre for the brass, politicians and press invited for the unveiling. A huge curtain has been drawn across the hangar, splitting it in two. The lights go down and there is a great rumbling sound from above. I look up and see that the yellow roof crane that spans the whole hangar is sliding slowly in the dark, pulling back the blue curtain, a cluster of orange lights hovering somewhere high above. As music from the film The Rocketeer plays, a fog machine spreads a soft and ghostly fog around the craft: a white object that looks like nothing so much as a flying saucer with a large porthole. It takes a few seconds to see that narrow wings grow from the sides of the disc. It is DarkStar, the new so-called Tier III Minus UAV.
The Tier designation has its origins in some obscure Pentagon paper and delineates the pecking order of unmanned aerial vehicles - by range, time in the air, size and other features. In this recondite categorisation, the Predator, a so-called Tier II UAV, follows the Tier I "Gnat 750," which the CIA tried out over Bosnia from Albanian bases. Tier II Plus is being constructed by Teledyne Ryan in San Diego and should be able to fly at 65,000 feet (about 20km) for 24 hours or more; this craft, yet to fly, is a semi-rival of Lockheed's DarkStar, with a less stealthy shape. DarkStar, which will cruise at 288 miles per hour (463km/h) using a single jet engine buried inside what looks like the porthole, can reach 45,000 feet (or about 14km) where it surveys a huge area of some 16,000 square miles - about 41,000km2 - using synthetic aperture radar or electro-optical cameras made by Sony. But its UFO-like shape makes it harder for radar to spot, track and shoot down than the Tier II Plus - and, at least in theory, safer.
After the smoke around DarkStar has dissipated, and the oohs and aahs have faded to silence, I talk to Major General Ken Israel, head of the Defence Airborne Reconnaissance Office. DARO - along with ARPA, the Advanced Research Projects Agency, which gave us the original stealth fighter - developed DarkStar, and Israel is its champion. During the 1960s, Israel flew on an EB-66, a bomber converted to monitor Soviet radio traffic and telemetry. He paints a glorious future for UAVs, a revo-lution in aviation.
Israel's leading arguments for UAVs are humanistic: "In the next century, we will definitely rely more on pilotless aircraft to place people out of harm's way." But they are also couched in terms of the new Pentagon fashion - 'infowar.' "We need to know what's on the battlefield before we get on the battlefield," says Israel. UAVs can give the generals not just desktop infowar but desktop infowar in real time.
Look, too, he says - using the economic terms of today's budget-conscious Pentagon - at "cost of ownership." Including development, production, maintenance, and other costs, the SR-71 Blackbird costs $38,000 (£23,750) an hour to fly, a U-2 $6,000 (£3,750) an hour, but a UAV only $2,000 (£1,250). And with its ability to linger over an area, Israel argues, a UAV can "view the battlefield with impunity." It can wait until the prey moves, the Scud scurries.
The logic for UAVs has been obvious to some for years. The former head of the Skunk Works, the late Ben Rich, touted pilotless planes in a book last year, and Pentagon visionaries regularly include them in their conception of future high-tech battlefields. But even as early as the '60s, the legendary founder of the Skunk Works, Clarence "Kelly" Johnson - who took a handful of engineers, set them up in a few old engine crates in 1944 and created the first successful US jet fighter - predicted that the future of military aviation would belong to unmanned aerial vehicles.
They can outperform manned craft, pointed out the designer of some of the best manned aircraft in history, because they can tolerate more Gs. And they cost less, because they don't have to carry systems to protect and nurture the pilot at risk. The ill-fated D-21 I saw at Blackbird Park, however, was the closest Johnson came to realising his vision.
Modern UAV history really began in 1982, when Israeli pilots flying over the Bekáa Valley in Lebanon used unmanned planes designed by a California-born engineer named Al Ellis to knock out Syrian surface-to-air missiles. Those remotely piloted vehicles, called drones, waited high in the air while a volley of smaller decoy planes entered the valley, carrying electronics that created the radar impression of a larger fighter plane and fooling the SAM radars into locking on. Once locked on, the radar could be spotted by Israeli bombs, incapacitating the SAM sites.
For the most part, the US military ignored these developments in the Middle East. But after the Reagan administration stationed US Marines in Lebanon, the Israelis demonstrated the usefulness of another kind of remotely piloted vehicle called Mastiff. From the camera of one such craft, they reportedly shot a videotape of then-Marine Commandant General P. X. Kelley visiting his troops. In case the point was too subtle, they fixed his head in the camera's cross hairs. The result impressed Kelley's boss, Secretary of the Navy John Lehman, who ordered several of the Mastiff RPV systems for his own forces.
Meanwhile, the US set about developing its own remotely piloted vehicles, the old-fashioned Pentagon way. While the Mastiff and other Israeli systems were basically overgrown model airplanes carrying a variety of electronics, the American Aquila would be a multimillion-dollar superplane lacking only a pilot. Lockheed's Texas branch devoted years and many unsuccessful test flights to the project, which was finally killed. The Mastiff had been developed in five years for half a million dollars. The Aquila - Latin for eagle - took a decade and nearly $1 billion (£625 million), only to fail.
Recently, the Persian Gulf War gave impetus to UAV proponents, especially in light of the difficulty of locating Scuds on the ground. While most American TV viewers saw Desert Storm as a model of information efficiency and intelligence gathering, General Norman Schwarzkopf and other commanders complained about their lack of real-time information. The cool images we saw of smart bombs riding lasers down air vents were actually gun-sight films, carried back to base and developed. Real-time information about enemy targets was much harder for the generals to get. By the time satellite photos and other images reached the field from Washington, the tanks had often moved, the Scuds had been shifted.
This flow of information reflected the old order of Pentagon thinking. Today, both are being 're-engineered,' and UAVs are a tool in the process. The old Pentagon was a sort of mainframe, top-down system, but now field generals want the real-time intelligence a UAV can provide - the equivalent of a desktop PC, pushing decisions down to the front.
To be sure, the new Pentagon thinking is still the classic case of fighting the last war, but it also offers a look at the information war of the future. War-making today is wrapped in the same clichés as private enterprise - colonels at the Pentagon now talk about re-engineering, downsizing and empowerment, just like middle managers in business. UAV proponents tout their craft as the PCs of intelligence - inexpensive, adaptable, customisable - laptops for the general on the go.
But what bothers traditionalists about this vision is the lack of a cockpit, of windows - of a guy with guts in there. The image of the piloted plane is still central to the US Air Force. The fighter ace, the lone-dog jet jock, the right stuff - all this is as basic a part of American fighting folklore as the cowboy on horseback. It is not easily challenged, a fact which makes Ken Israel and his cohorts identify with George Patton and other proponents of the original tank design, who struggled against the dominant horse-cavalry mentality of the early 1900s.
The UAV champions in fact identify the most with Billy Mitchell (played by Gary Cooper in The Court Martial of Billy Mitchell), who became a martyr to the futuristic idea of dropping bombs on people in distant cities and was driven out of the Air Corps. Mitchell embarrassed the US Navy by sinking ships with bombers, proving the vulnerability of the battleship. In the process, he laid the foundation for the fire-bombing of Europe and Japan, the atomic bomb, and General Curtis "Bombs Away" LeMay's Strategic Air Command.
"We are like Billy Mitchell," say Israel and other UAV proponents, invoking the one-time rebel, now a saint in the Air Force pantheon. But pilots will not accept this evolution so easily. Israel's vision of a future war in which bat-like robot planes swoop and dive in the sky too acutely challenges the status of the traditional hero. There is no reason at all, Israel says, that UAVs could not take over the job of the manned interceptor - that Captain Nolo could not supplant Chuck Yeager. It is only a matter of time before fighter planes without fighter jocks inside them joust with each other in some robot battle in the sky.
Lieutenant Colonel Jim Greenwood, like Israel, is a DarkStar enthusiast and has no trouble envisioning the air jockeys of the future. The first controllers of DarkStar, he explains, will be experienced pilots. "But in the future," he adds, "you might take people straight off the street, give them pilot training and instrument rating, then have them stop flying real planes and go to UAV school."
The prospect of videogame stars taking over from Top Gun doesn't seem to faze the colonel. Although he concedes that "it's a pilot's air force" and some pilots will resist UAVs to their last breath, when it comes to computers replacing the "human element" at the controls, says Greenwood, that began happening long ago. Computers fly airplanes much more often than pilots like to admit. Many aircraft, such as the F-117 stealth fighter, are unstable without controlling computers that translate the motions of joystick and pedals. And commercial airliners full of hundreds of souls are more readily trusted to computer systems than to human pilots for bad-weather landings at the world's major airports.
Pilots have grudgingly accepted the co-operation of computers in such contexts, but it is far harder for them to face the prospect of being grounded in favour of a videogame geek.
Yet considerable incentive for replacing the fighter with unmanned vehicles, despite all our affection for the chivalry and heroism of the dogfight, comes in part from the Gulf War, which made clear how sensitive the US has become to any loss of human life during conflicts. The Gulf War syndrome - in other words, total intolerance of casualties - has replaced the Vietnam syndrome. And the US is especially unwilling to accept the national humiliation occasioned by shot-down pilots who become prisoners, even hostages, to be displayed for the cameras.
Cases in point go back as far as Francis Gary Powers and the infamous U-2 spy-plane shoot-down of 1960. But another example conveniently popped up the day after Darkstar's unveiling, when an F-16 carrying pilot Scott O'Grady was shot down over Bosnia. UAVs reduce the need to send aircraft over such areas - and the chance of pilots becoming the enemy's pawns.
In fact, Predators helped find O'Grady, although they weren't fully deployed until a month after his rescue. Had they been used earlier, UAVs might also have warned of the menace of SAMs on the ground.
If the right-stuff tradition is deeply embedded in American military thinking, another ethos is equally strong: the belief in better guns and faster planes, longer-range bombers and deadlier lasers. There is an abiding faith in high-tech weapons to win wars and save lives in the process. One member of the DarkStar team, Boeing's Randy Harrison, puts it this way: "UAVs are part of the great American tradition of sub-stituting technology for human beings."
This substitution, though, has not always occurred as a linear progression. The pilots and ground crews from El Mirage and Fort Huachuca, for instance, found themselves facing new problems when they packed up and headed to the Balkans. On 12th June, one of the Predators I had seen at El Mirage overshot a runway and was damaged. And in Bosnia, the quality of the images brought back by the first Predator flights proved disappointing. In early August, two of the four Predators flying from Gjander airfield in Albania were lost over Bosnia. One simply vanished, presumably shot down. The other, after developing its own problems, was intentionally flown into a mountainside and destroyed to protect its sensitive technology from falling into the wrong hands.
Critics cited these losses as evidence that unmanned aerial vehicles would never replace recon planes. But Bosnia had already shown that UAVs were too important to be abandoned. DarkStar's test flights had been delayed by software problems, yet the same week as the Predator crashes, the Army was taking delivery of its fifth Hunter craft, and the Israelis unveiled a new generation of UAVs. The first official squadron of Air Force UAVs is now taking shape in the Nevada desert, at Indian Springs, a disused World War II facility on the edge of the huge, secretive Nellis Air Force Base test range, and has been designated the 11th Reconnaissance Squadron.
Meanwhile, General Joseph Ralston, commander of the Air Force Air Combat Command, has now gone on record as believing that UAVs are a high priority. But ironically, it is the US Army, not the Air Force, that proposes to move UAVs beyond reconnaissance into weaponry. The Army recently proposed a force of Teledyne's Tier II UAVs to fire "boost-phase intercept missiles" - antimissile missiles for shooting down Scuds or other intermediate-range ballistic missiles on their way up, when they are slow, easy to spot and vulnerable to attack. In this vision, UAVs could also linger above the battlefield and knock out enemy tanks the moment they crept out of cover, with the new generation of smart munitions borrowed from the Star Wars programme, known as "Brilliant Pebbles."
So, back in the desert, the engineers working on tomorrow's UAVs thumb through their newspapers tracking the performance of the robot planes in the Balkans. With the deployment of US for- ces to Bosnia, Predator and Hunter have a chance to assume an even more important role - and an opportunity to demonstrate the effectiveness of UAVs in keeping the peace. But ultimately, for the UAV champions, like the prophets of tanks and air power in the 1920s and '30s, only the actuality of war can prove the effectiveness of the weapons they believe in.
Phil Patton (pattonp@pipeline.com) is a regular contributor to Wired.
The Benefits Of UAVs
Fighter pilots are American heroes - but they also have their drawbacks.
Presented by an Air Force colonel at a meeting regarding UAVs:
They don't need crew rest.
They don't hog all the seats at the O Club during happy hour.
You don't have to call them by some dumb nickname.
It doesn't bother them that they don't know their parents.
They don't have egos greater than all outdoors.
They don't embarrass you at social functions by molesting the opposite sex and playing aircraft carrier on a beer-soaked table.
They don't throw up on your wife's new dress.
They don't need US$100 sunglasses to hide their bloodshot eyes.
They don't need no stinkin' leather jacket.- from Skunk Works Digest mailing list