Taliban troops and tanks had massed on a ridge within view of a US forward air controller, who scanned the position with a laser range finder and relayed the coordinates to the theater command center. The target was passed to a B-52, flying unseen above the clouds. Guiding on a signal from space, the B-52 struck with deadly precision. Time elapsed: 19 minutes.

Several things made the quick reaction possible. The target was clear and visible. The FAC was close enough to make positive identification. The ridge was in a location where the FAC was authorized to release a strike on certain kinds of targets, such as tanks.

For various reasons-some technical, some procedural, some political-most targeting of airpower takes much longer than 19 minutes.

In the Kosovo operation three years ago, the representatives of more than a dozen nations got to vote on what to bomb. The approval of fixed targets took, incredibly, an average of 14 days.

The constraints are seldom that severe, although The New Yorker and the Washington Post have reported delays in the approval of targets in Afghanistan that may have let some Taliban and al Qaeda leaders escape.

Avoidance of civilian casualties is important. Mistakes can kill the wrong people. They can also undercut the objectives for which the war is being fought.

It is not easy to balance such considerations against military requirements, especially in a theater like Afghanistan, where the Pentagon says 90 percent of the targets have been of the mobile or "emerging" variety.

The crux of the matter is how soon we can absorb and act on data from intelligence, surveillance, and reconnaissance sources. When the data is ambiguous, it stretches out the targeting decision.

Part of the solution is more capable sensors. In general, though, our ability to gather data is greater than our ability to translate it into usable information.

The Air Force believes the targeting loop can be shortened substantially. This is a high priority for Gen. John P. Jumper, who became Air Force Chief of Staff last September. He has been pushing the idea for his last three assignments.

Jumper says that the "kill chain"-the time from when the target is spotted by a sensor until a shooter locks on to it-can be reduced to 10 minutes or less.

To get there, we must modify intelligence practices held over from the Cold War, when the collection of data was paramount. "First we collect, then we analyze, then we report. Does that sound time critical to anybody here?" Jumper asks.

He prescribes the "horizontal integration" of aircraft and spacecraft, which would exchange data with each other, directly and immediately. A small preview was seen in Afghanistan, where Predator reconnaissance drones provided a streaming live feed from their video cameras to AC-130 gunships.

The place where it all comes together is the Combined Air Operations Center, which Jumper says the Air Force "will consider as a weapon system and crew it like we crew an airplane."

The automated CAOC of the future will combine information from satellites and manned and unmanned aircraft with everything that's in any database. In the past, such information came in separately from various sources, platforms, and priesthoods, which diminished its value in making decisions.

The CAOC will obtain and display, without being asked, useful kinds of information about targets and the battle area. This will be done almost instantly, machine to machine, and it will go a long way toward resolving ambiguities of target location and target identification.

Political constraints are not within the military's purview to decide. However, the stronger the confirmation of the nature of a potential target, the greater the confidence of the politicians in the proposed action is likely to be.

Reducing the targeting cycle is not a new idea, nor is the fusion of intelligence data. The difference this time is the degree of data integration that technology now makes possible and the amount of horsepower behind the effort. In the past, attention centered on platforms and programs, and integration was a by-product.

Among other evidence of the commitment, the Air Force is establishing a new three-star position on the Air Staff, a Deputy Chief of Staff for Warfighting Integration. Look for this to be a major theme in the months ahead.

The value of rapid targeting is obvious at the tactical level of war. If the process is slow, time-urgent opportunities may be lost or mobile targets may be long gone by the time an air strike gets there.

The significance at the operational and strategic levels may be less apparent.

For the past decade or so, the United States has pursued, with considerable success, the concept of parallel warfare. The attack is launched on all sides at once, and is simultaneous rather than sequential.

There is no time for the enemy to react, adjust, adapt, mount a counteroffensive, or escape. The faster the attack, the more effective it will be and the sooner the conflict will likely be concluded and over.

Tightening the sensor-to-shooter loop is long overdue, and the effort to do it has an excellent chance to succeed.