COLUMN ONE : Turning Facts Into Attacks : Intelligence-gathering by the U.S. in the Gulf War had severe lapses despite cascading rivers of data. New technology will bring order to information and put knowledge into action.

On a remote battlefield in the Middle East, an Army major punches a few keys on a computer console. A split second later, his screen is bright with new satellite photos--taken only a few minutes before--of a column of enemy tanks rumbling toward American lines.

He hits the keys again, and the computer identifies the enemy tank unit, lists its armor and cannon range, summarizes freshly intercepted enemy radio messages and even offers up a capsule biography of its commander. All over the area, other U.S. units are doing the same.

That scene still is fantasy, but, according to military intelligence experts, a system very similar to that may well become a reality in just a few years.

Spurred by intelligence lapses in the Persian Gulf War that could have proved serious, intelligence agencies are moving to capitalize on a new generation of exotic technology that would enable combat forces to monitor enemy units almost as an attack is imminent.

Pentagon strategists say that, in future wars, U.S. military commanders may be able to tap into sophisticated information systems that correlate data from photo and radar satellites and other sources--and even point out hidden targets.

The Gulf War demonstrated the many strengths--and several shortcomings--of tactical intelligence-gathering and established how crucial these skills will be if the Pentagon is to cope with threats that are likely to crop up in future conflicts.

“The war has really focused attention on these issues,” said Rep. Norm Dicks (D-Wash.), a member of the House defense appropriations subcommittee.

Dicks predicts that the Gulf conflict will give a push to some existing development programs and point up the need for others. It may also provide ideas on how to improve the way the intelligence-gathering bureaucracy serves battlefield needs.

Until now, most effort has gone into detection and gathering of intelligence. Military analysts say that the next generation of equipment will focus on how to process and distribute what is gleaned.

Such new technology should enable troops to fight their battles from even greater distances--and keep U.S. casualties even lower.

Technological progress already has been breathtaking. Over the last few decades, scientists have accomplished miracles in extending the reach and capability of intelligence-gathering and processing equipment.

Today, Keyhole-11 satellites in space can read three-inch-high letters on the ground. During the Gulf War, eavesdropping satellites recorded conversations whispered in underground Iraqi bunkers. And they picked up virtually every other important communication.

The next generation will be miraculous for different reasons. Its aim will be to bring order to the cascading rivers of intelligence data, filter out the huge proportion that is irrelevant and deliver pertinent information in a coherent form to the battlefield where it is needed.

These new programs are aimed at creating what the Pentagon calls quick-response battlefield systems. And military men say the prospects are tantalizing.

The varieties include:

--Information systems such as the battlefield computer station, which correlates data from photo and radar satellites, intercepted enemy radio conversations and human sources to provide instantaneous information to field commanders about the forces they are facing.

--Advanced computers designed to ease intelligence bottlenecks by scanning thousands of satellite photos to help pick out hidden targets, such as enemy tanks, planes and artillery that would not otherwise be easily spotted.

--Reconnaissance aircraft that can beam integrated radar and photo images of enemy weaponry to dozens of sites on the battlefield simultaneously and virtually instantly--without the delays inherent in relying on film, which must first be developed.

--Unmanned aircraft that can hover over enemy positions undetected and provide TV photos clear enough to enable friendly ground units to adjust their artillery or tank fire and destroy the enemy within seconds.

Paul B. Stares, a defense specialist at the Brookings Institution, says the push ahead is a necessity. For years, the technology for collecting data has outpaced the technology for processing and distributing it, Stares says. “The goal now is to strike a better balance.”

The U.S. experience in the Persian Gulf War demonstrated graphically what might happen if intelligence-gathering hits a snag. Although the quantity and quality of intelligence reports were unprecedented, they were also conspicuously uneven.

Battlefield commanders complained that too often the intelligence data was passed through too many middlemen and took too long to reach the field.

For example, Air Force officers got detailed photos taken by cameras from the Keyhole-11s, but the satellites only passed overhead every 1 1/2 hours or so--and often did not cover the same area a second time.

U.S. pilots received relatively fresh photos from spy planes fitted with special television cameras. But most U.S. spy planes were forced to use Vietnam-era film cameras, which involved several hours’ delay for developing the film--allowing Iraqi tanks to escape.

As a result, U.S. military experts say, the effectiveness of the intelligence-gathering operation for U.S. troops in the Kuwaiti theater was decidedly mixed.

“Some of the intelligence pictures we got were less than an hour old when we got them--and sometimes we didn’t get the pictures at all,” said Air Force Col. Steve Plummer, vice commander of the 4th Tactical Wing.

Battlefield commanders complained also about the delays that the military services experienced as a result of requirements that they channel their photos through the CIA and Defense Intelligence Agency, which are outside the military’s control.

During the first week to 10 days of the war, fighter pilots reported having trouble finding their targets. Sometimes weather or Iraqi camouflage was to blame, but often the intelligence was simply outdated and the target had moved.

“Our results were not what they should be with modern weapons,” said Lt. Gen. Charles A. Horner, commander of the Air Force in the Gulf War. “On a scale of 1 to 10, it was a 5. We (eventually) corrected it to a 9.”

The way Horner and his crews improved their performance was to dispatch a team of “Killer Scouts”--specially trained F-16 pilots who flew ahead of the bombers, found the targets and sent back information on their positions to the bombers.

The snafus did not seriously hinder the campaign, but they were grave enough to be a major focus of Pentagon post-mortems of the conflict--as well as a part of the sweeping study of U.S. intelligence operations now under way in Congress.

What was needed to solve these problems? “Technology--absolutely,” said Sen. John Glenn (D-Ohio), a former Marine officer now a member of the Senate Select Committee on Intelligence. U.S. forces, Glenn said, often simply “didn’t have it.”

A senior U.S. military source said the Pentagon would like to be less dependent on satellites that are controlled by the national intelligence agencies--or at least be able to supplement them with information from alternative sources.

One answer may be the JSTARs--highly sophisticated, radar-evading reconnaissance aircraft that rely on radar and infrared sensors to generate computer-enhanced images of the battlefield. “You need complementary systems,” the senior official said.

Lt. Gen. William C. Odom, a former director of the National Security Agency, has called on the Defense Department to set up a separate agency to handle image intelligence--much in the way that the NSA itself handles radio intelligence.

“The whole imagery intelligence area is broken,” Odom said, “and the Defense Department needs to put it back together again--in a hurry.”

But, although there still is not a consensus on how the United States should organize its intelligence-gathering, there’s no dispute at all about the value of the next generation of equipment.

Researchers are experimenting with new “fusion centers”--such as the Army’s high priority All-Source Analysis System--that sift raw intelligence data, splice useful strands together and distribute it.

John Pike, a space policy analyst at the Federation of American Scientists in Washington, says the military had nine separate--and incompatible--information systems in the Gulf. The next step is to link them with unifying software, so analysts can cross-reference the data from each.

Planners say that, once such a system is in place, commanders will more often be able to stage complex operations--such as the now-famous attack in Baghdad in which an American pilot sent a laser-guided missile down the elevator shaft of a Iraqi factory.

The pilot knew about the shaft because of intelligence data from several sources, Pike recounts. With the All-Source Analysis System, “that kind of planning would be that much quicker.”

The information from these systems most probably wouldn’t be useful to commanders of small units in the field. It simply is too general to answer their questions about what lies over the next ridge.

But the Pentagon is planning other intelligence aids for those units. One is a laptop computer unit that would provide field commanders with periodic updates on the disposition of enemy forces, Pike said.

Because of the experience in the Gulf, the military is increasing its reliance on unmanned reconnaissance aircraft. Such drones were a flop in the Vietnam War, and more recent Army efforts to use them effectively ended in a costly failure.

But, in the Gulf, the Navy successfully used a Vietnam-era drone called the Pioneer to help direct gunfire from its battleships. Video images flashed from Pioneers 20 miles away enabled gunnery officers to see what they had hit and to readjust their sights accordingly.

Analysts speculate that the Pentagon may soon develop programmable drones that have greater range, radar and heat-sensitive sensors and the ability to evade enemy sensors.

The two JSTAR reconnaissance craft used in the Gulf War have been widely hailed as successful.

The aircraft, part of a $6-billion program that was nearly scrapped before the war, provided 250-kilometer-square views of the Kuwait theater that showed almost instantaneous images of Iraqi tanks, artillery and armored personnel carriers.

The JSTAR aircraft beamed their images to a half-dozen ground stations, where the photos were used to direct artillery and missile attacks and air strikes by F-15As on enemy Scud missile launchers.

Although vehicles appear only as points of light on JSTAR images, Army officials say the system’s operators were able to pick out Scud launchers with accuracy.

Some analysts are skeptical of those claims. But it is clear that the next goal for the JSTAR program will be to increase its ability to discriminate among moving objects.

The next generation of tactical intelligence equipment is also likely to include the first computers to help sift through the mountains of images collected by satellites and reconnaissance aircraft.

Researchers have long discussed so-called automatic target recognition to lighten the load of the thousands of intelligence analysts who spend their days poring over often fuzzy images of terrain.

Pike, of the Scientists’ Federation, said he is skeptical that researchers can devise a system any time soon that has a low enough error rate to be able to handle the job. Nevertheless, he conceded, even a small improvement “could speed up the process several-fold.” In any case, experts say the new generation of equipment is not far away. Next time the United States goes to war, a computer station may be as much a part of an Army unit’s field gear as its binoculars and maps.

INSTANT BATTLEFIELD INTELLIGENCE

The Persian Gulf War proved the value of high-tech intelligence-gathering but also showed that our ability to process information lags behind our ability to collect it. Some hoped-for advances:

Battlefield computer stations will assemble data from photo and radar satellites, intercepted enemy radio conversations and human sources to provide instantaneous information to field commanders.

Computers will scan thousands of satellite photos to select hidden targets, such as enemy tanks, planes and artillery not easily spotted otherwise.

Reconnaissance aircraft will beam digital radar and photo images instantly rather than waiting to develop film of enemy weaponry.

Unmanned aircraft will hover undetected over enemy positions to provide TV photos enabling quick adjustment in artillery or tank fire.