Science Corp.'s ‘Magic Bullet’ for Bomb Detection Misfires With FAA : Terrorism: Government agency, disappointed over bulky machine’s accuracy in detecting explosives in baggage, is scrapping plan to buy up to 150 of the $1.1-million machines.


Two years ago, a bomb detection device developed by Science Applications International Corp. of San Diego was hailed as the best chance government and airline officials had of making airplanes safe from the kind of terrorist explosives that caused the crash of Pan Am Flight 103 over Lockerbie, Scotland in December, 1988.

Today, the Federal Aviation Administration concedes that the high hopes it once held for SAIC’s thermal neutron analysis (TNA) machine were overly optimistic, and that the agency’s plan to buy as many as 150 of the $1.1-million machines has been scrapped. Four of six machines installed at U.S. airports have been pulled from on-line use.

And, for the foreseeable future, the FAA says its bomb detection and anti-terrorism efforts will revolve around experienced and highly trained human inspectors, not the massive, 10-ton machines made in Science Application’s Santa Clara plant that are capable of scanning up to 600 bags an hour.

The need for bomb detection technology to ensure airline safety was highlighted last week after two Libyan nationals were indicted for the Flight 103 bombing, which killed 259 on board and 11 on the ground. The Libyans allegedly had a bag with plastic explosives placed aboard a flight from Malta, with the bags being transferred to the ill-starred Flight 103 in Frankfurt.


At the time of the crash, the FAA had five of the SAIC machines on order. But the agency asked the San Diego firm to move up delivery by six months, so strong was the technology’s promise as a stand-alone “magic bullet” capable of sniffing out plastic explosives. In addition to the FAA’s imprimatur, the Science Applications machines also received support from international airline trade groups.

But that promise has since faded, and blame for shortcomings of the machines is partly that of the FAA and partly that of Science Applications, a 13,000-employee contract research firm specializing in high technology, defense contracts and system integration work.

The FAA concedes that the six machines that Science Applications delivered met the specifications given to the company, but now says those specifications were too narrow, adding that it erred in its estimates of the types and quantities of explosives that were capable of causing a catastrophic aircraft explosion. For security reasons, Adm. Clyde Robbins, director of the FAA’s Office of Intelligence and Security, declined to be more specific about the nature of the FAA’s errors.

But Robbins also described the six SAIC machines that were delivered as unwieldy and unreliable. “They are large, expensive and not as capable as we had hoped. They have a false alarm rate higher than we would like. . . . The size of them is a restricting factor, because they are heavy and they would be difficult to put in most terminals,” he said.


The machines are 12 feet long, 7 feet wide and 6 feet, 4 inches high and weigh 10 tons each.

As a result of the machines’ shortcomings, the FAA is using the ones still in operation only as part of a system that involves intense hands-on inspection. “The bottom line is, we still don’t have a piece of equipment that can do the job. Until we get one, we will rely on human element to find explosives,” Robbins said.

A Science Applications executive said Monday that the machines were never intended to be “panaceas” to solve the terrorist bombing threat.

“No system can be used as the only weapon” against bombings, said Cathal Flynn, SAIC’s senior analyst and the company’s liaison with the FAA. “As people have looked at this problem in a broader sense, they realize they need to have a system that includes a lot of things, no single one of which solves it all.”

Airlines and government inspectors, Flynn said, will always need “profiling of suspected terrorists, screening passengers through questioning and X-raying of bags . . . to increase the probability of finding an explosive.”

Insisting he has no intention of “bad-mouthing” Science Applications, which he said is still the only company to have developed a plastic explosive-detecting device, Robbins nevertheless said the FAA is helping fund the development of alternative bomb-detecting technologies, and that several companies may soon be challenging SAIC.

Science Applications’ technology is based on a radioactive source that directs neutrons into baggage moving along a belt. The neutron bombardment causes the bags to emit gamma rays, which in turn are measured by a computer for characteristics resembling the “signature” emissions of explosives.

The competing companies also developing bomb detection equipment, including Imatron of South San Francisco and Thermedics of Woburn, Mass., each use different technological approaches, Robbins said. One company is targeting measurement of vapors that are emitted by explosives.


Robbins said he envisions an overall system using several technologies. “There is no silver bullet, no one that does everything for everyone. Each of these pieces of equipment does a particular thing well, and we’ll have to put them together.”

SAIC couldn’t agree more, company spokesman Chuck Nichols said Monday.

“Everybody wishes they had (a bomb detector) that costs $25,000, is no bigger than a breadbox and which weighs less than 5 pounds,” Nichols said. “Unfortunately, it doesn’t exist, and there is no indication that anything like it will exist in the near future other than TNA and trained people who can find explosives.”

Despite the disappointing response of the FAA to its machines, SAIC has not laid off any of its 80 Santa Clara-based employees working on the TNA project, Nichols said, adding that the company is continuing development of its TNA technology.

SAIC is also going forward with its contract with the developers of the tunnel connecting the European mainland with Great Britain to “augment security devices for vehicles going through the tunnel,” Nichols said. The tunnel is scheduled to open in 1993.