Missile Impossible?

(3 of 4)

And a crafty foe wouldn't limit itself to the Pentagon's single, simple decoy. The enemy could slip its warhead inside a decoy balloon and deploy it along with a dozen identical balloons, forcing the Pentagon into a futile effort to destroy all of them. The warhead might be cloaked in a shroud of liquid nitrogen, chilling it so that the interceptor's heat-seeking sensors couldn't find it. Chemical or biological weapons might be deployed in dozens of bomblets far too numerous to destroy.

Distinguishing between warheads and decoys requires a wealth of information the Pentagon wouldn't have in a real attack and wouldn't be likely to get. But Pentagon officials insist their relatively crude discrimination technologies will keep improving. They say measuring subtle differences in projectiles' mass, motion, reflection and rotation will enable the Pentagon to pluck the real warhead from among the decoys. But the officials decline to detail the technical wizardry behind their assertion, saying that divulging their techniques would only aid potential foes.

Back in the sky over the Pacific on Friday, physics and chemistry will take over. The interceptor's three sensors--two detecting heat and one detecting visible light--all share the telescope that juts out its front end. The visible-light sensor will get the interceptor into the right neighborhood, but only the infrared sensors can guide it into its target, gently steering it with minithrusters powered by 30 lbs. of liquid rocket fuel. For the heat-detecting sensors to "see" anything, they must be chilled to -330[degrees]F using nitrogen and krypton, funneled to the sensors through a 0.0035-in. diameter pipe.

Each sensor's 65,000 pixels will feed signals into the interceptor's brain, where lightning-fast calculations involving heat, light, mass and motion are cranked into databases searching for the ballistic fingerprints of enemy warheads. As the interceptor rushes toward its possible targets (the warhead, the balloon and the launch container), it will keep them all within view for as long as possible before discarding the ones its computers say have the least likelihood of being the warhead.

The interceptor's thrusters will fire in precisely choreographed microsecond bursts to guide it into a collision with its chosen target. There are no explosives aboard the interceptor. The sheer kinetic force of the crash at a combined speed of nearly 17,000 m.p.h. makes explosives superfluous. The process is often described as "hitting a bullet with a bullet," but that imagery is misleading. There is no head-on collision. In fact, when the collision occurs, interceptor and target are both on the way down, pulled by gravity; the interceptor hits the side, not the nose, of the target. If all goes according to plan, both the interceptor and the warhead will disappear into a fiery cloud of what Pentagon officials like to call "space dust." If things go amiss, the only thing turning into space dust will be the $100 million cost of the test.

Like the term "military intelligence," the phrase "Pentagon testing" is something of an oxymoron. Military officers and contractors have long bent the rules and faked results to keep programs on track and money flowing. Watching the preparations for this week's test, Coyle remains skeptical of the rush to field the missile shield. He says the push to build a system with "immature" hardware and inadequate testing is a hallmark of troubled Pentagon programs.

The U.S. military, in missile tests conducted for several programs, has succeeded in "hitting a bullet with a bullet" in space only four times in 14 tries, or about 30% of the time. "While the four successful intercepts provide support for the hit-to-kill concept," the General Accounting Office noted two weeks ago, "the 10 failed attempts raise questions about reliability." No one wants to build a $30 billion missile shield that would shoot down less than a third of incoming warheads. But that statistical shortcoming has only increased the pressure on this test--and its consequences for the federal budget.

The Pentagon is demanding that each interceptor have about a 90% chance of killing its target. To achieve that, the military plans to have the final operational system fire four interceptors at each suspected warhead, pushing the chances of a kill to above 95%. But that will drive the cost higher too. In fact, those odds are unprecedented for a system of such complexity, especially one that must be on perpetual alert. Even the B-2 bomber, perhaps the Pentagon's most pampered weapon, has proved to be capable of performing its mission only 43% of the time.