Betting the planet on IT

The national missile defense (NMD) system, a $36 billion proj-ect to build

an ultrahigh-tech intercontinental ballistic missile shooting gallery,

faces a crucial technical test this week. The July 7 exercise, which involves

intercepting a Minuteman II missile using a complex space-based warning

system, will go a long way toward determining the fate of the politically

charged project. But it will also say a lot about whether today's information

and telecommunications technology is ready and reliable enough to "bet the

country on," as one critic says in our cover story this week.

Certainly the challenge is technically breathtaking. Ground- and space-based

sensors and tracking systems would have to work seamlessly to detect the

equivalent of a fast-moving nuclear golf ball high over Seattle — from hundreds

of miles away. Without any electronic countermeasures to confuse the system,

that would be difficult enough. But NMD would have to discriminate a real

threat from among a throng of decoys and other disguised systems.

Is our information technology — and the management systems that bind

it together — really trustworthy under such circumstances? Probably not.

As in the realm of network security, a more fiendish countermeasure is always

in the making. So the system is never likely to be fail-safe.

We have also learned from our experience with modernizing the nation's

air traffic control system that, at least at present, even the top guns

in the technology field have trouble developing and managing billion-dollar

software projects. NMD, like air traffic control systems, requires integrating

multiple systems — with each link, like each system, representing another

potential point of failure. Even a positive outcome Friday leaves the possibility

for plenty of failures in the future.

NMD represents a powerful stake in the arms control debate. And though

it may not be a hard shield, it could provide a layer of insurance against

an actual attack. But NMD should not be considered an effective, reliable

application of information technology. That's a view that will only backfire.