Pilotless Planes: Not Cleared For Takeoff?
At noon on Nov. 4, an 11-foot-long experimental plane topped with a mushroom-shaped communications dome took off from Hondo, Tex. It proceeded to trace lazy racetrack-oval patterns in the sky, as it would when performing its intended mission: reconnaissance. The plane did have trouble maintaining its assigned 2,200-foot altitude, but don't blame the pilot. There was no pilot--not in the air, at least.
The aircraft, called Outrider, is controlled from the ground, using a laptop computer. That's so flying low-level spy missions over dangerous territory won't put pilots at risk. After the 15-minute flight, Alliant Techsystems Inc. executives were elated: No major snafus clouded the test--the first flight with a new and more reliable engine. And despite the glitch in the tail controls that hampered level flight, the plane "made a picture-perfect landing," recalls Outrider chief Don E. Cattell.
Alliant, based in Hopkins, Minn., has big plans for Outrider. But they're in jeopardy because of the plane's checkered past. And it comes on the heels of botched efforts to develop unmanned aerial vehicles, or UAVs, in the past (table, page 110). Outrider has been so problem-plagued that on the eve of the test flight, the Pentagon gave Alliant 30 days to show progress or the plane would be history. Lawmakers, angry that the military has so little to show for the $3 billion plowed into a batch of UAV designs since 1979, had already slashed Outrider's 1998 budget to $45 million, from $83.3 million. "The program appears to be on a path to failure," a recent congressional report concludes.
PERFECT PERCH. That's bad news not only for the unmanned aircraft business but also for the Pentagon's plans to use civilian technology to grease future progress. Outrider was supposed to be a model of procurement reform, showing how off-the-shelf commercial technology would speed up development, save money, and produce more advanced weapons systems throughout America's armed forces. Many similar fast-track programs are under way for such projects as precision targeting systems and remote monitors that would warn of biological weapons. Outrider, however, has become more of an object lesson in what pratfalls to avoid.
Pentagon officials still insist UAVs will one day give U.S. troops a decisive edge in warfare. If the Outrider works as Alliant expects, it could "loiter" above a battlefield for several hours, enabling Army and naval commanders to use its electro-optical cameras and infrared sensors to pinpoint enemy troops and follow their movements minute by minute. "It's like sitting on a flagpole in the middle of the battlefield," says Alliant Chairman Richard Schwartz.
Eventually, UAVs may fly more than recon missions. Some next-generation ground-attack planes may also be pilotless. UAVs tickle the fancy of war planners because unmanned means cheap. Outriders will cost as little as $300,000, and the tab for attack UAVs might be just $10 million--compared to $30 million for the proposed Joint Strike Fighter. And combat training for UAV jocks would be like learning a video game--far cheaper than logging thousands of hours of flight time. At some point, predicts Defense Secretary William S. Cohen, "you'll find much greater utilization of UAVs than ever before."
VIETNAM VETS. Despite current troubles, pilotless aircraft have a long record in surveillance. The U.S. flew 1,000 low-tech UAVs during the Vietnam War. But the information they gathered was on film, which was ferried to Washington for analysis. That usually took several days, making much of the intelligence outdated by the time it got back to Vietnam. Still, Israel Aircraft Industries (IAI) refined the U.S. technology and has been selling combat-tested UAVs for 20-odd years. IAI rakes in $160 million a year from its pilotless planes, which are relatively simple, with narrow missions.
The Defense Dept. sees new technology as a way to overcome those shortcomings, but gold-plated wish lists often create engineering nightmares. A recent General Accounting Office report casts the Pentagon and defense contractors as virtual Keystone Kops. In the 1980s, the Army squandered more than $1 billion on Aquila, a short-range UAV that Lockheed Martin Corp. tried to develop. It featured a secure communications link to protect against Soviet jamming. Unfortunately, the security was so tight it degraded the quality of Aquila's own video transmissions.
The only real high-tech success has been Predator, a 27-foot-long recon plane from General Atomics Aeronautical Systems Inc. Used in Bosnia, it soars up to 25,000 feet to provide a bird's-eye view of the landscape and funnel vast amounts of data to the military brass.
Outrider was conceived as a mini-Predator that would provide battlefield data to local commanders. By harnessing commercial technology, it was supposed to be a snap. So in May, 1996, Alliant got a contract covering just two years and $53 million. Then the program began skidding. The Pentagon decreed the plane must meet both Army and Navy needs, believing joint projects not only save money but improve interservice coordination. That makes sense for paper clips and communications systems but not always for weapons. For example, the Army sought a UAV with only a 50-kilometer range. But the Navy stipulated a 200-kilometer range so ship-launched UAVs could see over the horizon at sea.
Complicating matters further, the Navy demanded an engine that burns "heavy" fuel--diesel--so highly combustible gasoline wouldn't have to be stored aboard ships. Since no suitable diesel engine existed, Alliant converted a McCulloch 50-horsepower gasoline engine. It never worked properly. Finally, the company switched to a gasoline engine made in Britain by UAV Engines Ltd. But by then, the project was several months behind. "We were forcing too hard to look for one solution to satisfy all requirements," concedes Paul G. Kaminski, who was then the Pentagon's acquisition czar.
DANGEROUS BUGS. Other problems may bode ill for drafting commercial technology for military service. While 90% of Outrider's components are off the shelf, some had to be modified. One example: The electrical system in today's planes has a reset button to recover from interruptions. But there's nobody to push the button in a UAV. So a switching circuit had to be installed.
Similar woes bedevil just about every UAV program. Allegheny Teledyne Inc.'s Global Hawk, a $10 million craft essentially designed to be a U-2 spy plane on autopilot, is behind schedule and over budget. Rigorous software testing has been the main snag. That's because software bugs caused Lockheed's DarkStar, a stealthy reconnaissance jet, to crash on its second flight. After takeoff, it tried to gain altitude too quickly, so its software repeatedly forced the nose down to gain speed--until it hit the ground. Major General Kenneth R. Israel, head of the Defense Airborne Reconnaissance Office, says a third flight is expected sometime early next year.
Even as Defense struggles with this generation of UAVs, another is on the drawing boards. In October, Lockheed Martin won a contract to study "uninhabited combat vehicles" (UCAVs) that could be launched from a ship's deck or a submerged submarine. UCAVs would take on risky roles, such as knocking out surface-to-air missile sites before the U.S. establishes air superiority.
While much of the technology that UCAVs need exists, the biggest challenge is coordinating their operation alongside manned aircraft. That's "really a key," says Armand Chaput, head of Lockheed Martin's UCAV team. How long it will take to resolve this is anybody's guess, but UAV backers point out that the first spy satellite program in the 1960s, called Corona, also had its share of failures. Only 12 of the first 30 missions were productive. But eventually the technology became indispensable.
Three decades later, even the use of commercial technology doesn't guarantee fast-track results. Still, the Pentagon has little choice. It can't afford $335 billion for new manned aircraft alone over the next two decades. UAVs can help plug the budget gap. It just may take longer than the brass would like.