Birth of a Bomb

Sixty-one days after the United States won its first war with Iraq in February 1991, Air Force Chief of Staff Gen. Merrill McPeak scrawled a note that would change the course of the next three wars fought by the U.S. military. "We need to lay down a requirement for an all-weather precision-guided munition," McPeak wrote on May 1, 1991. That brief order began the quest for a new kind of bomb and helped launch the age of precision warfare.

McPeak's new bomb-an updated version of some of the oldest bombs in the military's stockpile-now rides on the wings or in the belly of nearly every bomber or fighter flown by the Air Force and Navy. What's new is the 5-foot steel strap-on tail kit containing a central computer and tail fins bolted to each bomb. The kit adds brains and steering to "dumb" 1,000- and 2,000-pound bombs, which otherwise rely on gravity and luck to hit their targets. The kits save the Defense Department hundreds of millions of dollars by updating an outdated arsenal of aging munitions.

When bomb and guidance kit are married, they become a Joint Direct Attack Munition (JDAM), one of the world's most accurate and cost-effective weapons. When a 2,000-pound JDAM hits, its 14-inch-wide steel warhead explodes into an 8,500-degree fireball of white-hot metal that moves faster than sound and leaves a crater at least 20 feet deep and often more than 50 feet wide. JDAM kits can be attached to small bombs when the military seeks to limit damage, and for use against underground or reinforced targets, the kits can be strapped onto "bunker buster" bombs with thicker, harder shells. Because their location constantly is updated by satellite, JDAMs can be launched through clouds and thick smoke from up to 45,000 feet above and 15 miles away from their targets.

More than two-thirds of the 30,000 bombs dropped on Iraq in March were laser- or satellite-guided. Of the guided weapons, nearly a third were JDAMs. They had an accuracy rate of over 90 percent and were used against some of the war's most sensitive targets.

In surprise strikes that began the war, JDAMs dropped from F-117 fighters destroyed palaces, houses and underground bunkers in a southern Baghdad complex, known as Dora Farms, where Iraqi leader Saddam Hussein was thought to be meeting with senior military advisers. And on April 7, near the war's end, an Air Force stealth bomber unleashed JDAM bunker busters on houses in the upscale Baghdad enclave of Mansour, where intelligence reports suggested Hussein was holed up with his sons. Those strikes leveled three buildings, left a 30-foot-deep crater, uprooted orange trees and broke windows and doors on houses more than 300 yards away, but probably missed Hussein.

A Bomb Problem

After the 1991 Persian Gulf War, McPeak skipped the ticker-tape parades, believing victory was a poor teacher. Instead, the former fighter pilot appointed only nine months earlier by President Clinton, pored over after-action reports from the Gulf. He was looking for flaws in the seemingly invincible armor of air power. It didn't take long to find a big one.

The Air Force had a bomb problem. About 90 percent of the bombs dropped during the war were 1,000- and 2,000-pound unguided munitions, which, on average, fell only within 200 feet of their intended targets. Dozens had to be dropped to take out a single target-an improvement from thousands during World War II, but not one that made air commanders much more confident. Commanders avoided using unguided bombs in urban areas for fear of hitting civilian targets.

But McPeak knew bombing precision was attainable. He was a forward air controller over Vietnam's Ho Chi Minh Trail in the spring of 1969, where he called in some of the first strikes using laser-guided bombs. The munitions consisted of a laser sensor attached to the tip of a dumb bomb. Pilots, or the weapons officers seated behind them, guided the bombs from aircraft by aiming handheld lasers at the targets. "The accuracy was amazing," McPeak says. "I'd call for them to cut off a street, they'd drop a bomb and leave a big crater right there." Twenty-two years later, those bombs were equally effective during the Persian Gulf War, accounting for only 7 percent of ordnance dropped, but delivering 75 percent of all damage wrought by the air war.

But laser-guided bombs must "see" their targets to hit them. During the Gulf War, one in three laser-guided bomb attacks was canceled because the lasers couldn't point out targets through storm clouds, sandstorms, and smoke and haze from oil fires. Military planners expected Operation Desert Storm's opening air war would last only 15 days, but bad weather caused so many bombing runs to be scrapped that the campaign lasted 43 days. Even in good weather, the use of laser-guided weapons was limited by their cost. At $100,000 each, they were too expensive for the Air Force to stockpile. In addition, only a handful of Navy and Air Force fighters, and no Air Force bombers, could drop them. In 1991, McPeak was frustrated that more than 20 years after they starred in Vietnam, guided bombs hadn't come down in price or become more versatile. He sought a cheap guided bomb that could be dropped under any conditions from any aircraft. And so, on May 1, 1991, he sent his memo. "The idea was: Don't drop any more unguided munitions, period," McPeak recalls.

The Technology Debate

Air Force Maj. Gen. Joseph Ralston would oversee the quest for the new bomb. In the spring of 1991, Ralston was the Air Force's director of tactical programs, managing the development and purchase of many Air Force weapons. Ralston believed the best guidance system for bombs was the global positioning system (GPS), a constellation of satellites orbiting 12,660 miles above the Earth. Launched between the mid-1980s and early 1990s, the satellites created an Information Age version of the North Star for military, and later civilian, users. GPS receivers determine their exact longitude, latitude and altitude by picking up signals from up to four satellites every second.

Before Operation Desert Storm, the military purchased 10,000 commercial GPS receivers. The handheld, high-tech compasses won high marks for tracking tank and troop movements. They even allowed the Army to quickly deliver meals to troops on the front lines. Ralston envisioned another use for GPS receivers: steering bombs. "I had no doubt it would work," he recalls.

But there were plenty of doubters with more stars on their shoulders than Ralston's. They were wary of a bomb that would use a commercial GPS receiver that anyone could buy for $200 at Radio Shack. They wanted a military-unique guidance system that relied on built-in radar. Ralston believed a bomb built with military-specified parts would be too expensive and take too long to build. Air-to-air missiles that used such radar cost hundreds of thousands of dollars each. "I was the one preaching, 'You have to keep it cheap,'" he says. Ralston called the Air Force's munitions laboratory at Eglin Air Force Base, Fla., and ordered up a prototype of a satellite-guided bomb.

As it turns out, Eglin engineer Louis Cerrato had been working on and off for six years to find a cheap way to make dumb bombs smart. In the mid-1980s, Cerrato had become intrigued with a concept that had been around for a decade, known as inertial guidance. Inertially guided munitions contain a central computer processor loaded with target coordinates. The processor uses data from onboard navigation hardware, such as gyroscopes and accelerometers, to direct tiny motors that maneuver tail fins to steer the munitions. The engineer pondered how to build large quantities of inertially guided weapons on the cheap.

Cerrato knew laser seekers were strapped onto dumb bombs, so he decided to create a strap-on inertial guidance system. The guidance kits could be attached to hundreds of thousands of dumb bombs the military had been stockpiling since the 1950s. But his idea languished. The Pentagon wanted to build new and bigger weapons, not retrofit old bombs. What's more, bombs guided by the inertial navigation system tended to drift off target. The research dollars dried up. "It was an orphan. Nobody wanted it," says Cerrato, who saved his notes anyhow.

Ralston's order sent Cerrato back into action. Finally, he would get to build the kits he had envisioned years before. But there would be a key difference: a commercial GPS receiver. By updating the bomb's location every second, the receiver would correct the tendency to drift. Cerato was so certain the concept would work, "I didn't even think a test was necessary," he recalls. But Ralston believed a live demonstration was the only way to convince the skeptics.

In the winter of 1992, an F-16 roared over a bombing range at Eglin carrying a 1,000-pound warhead packed with concrete instead of explosives and a GPS guidance system. Its target: a wooden outhouse built just for the occasion. Ralston wanted a dramatic target that could be videotaped to make his point. The location of the outhouse had been programmed into the test bomb. Ten thousand, 15,000, 20,000 feet, the fighter climbed into position. The pilot hit the pickle, a button on top of the fighter's steering joystick. The bomb fell away, plunging at 1,200 feet per second. The bomb updated its position via satellite second by second. A minute after the bomb fell free, the outhouse broke into splinters and disappeared in the dust of a 10-foot-deep crater. The debate was settled. The new bomb would be guided from space.

Pushing Boundaries

In 1993, not long after the bomb debate ended in splinters and dust, McPeak gathered a dozen Air Force procurement officials in his small conference room in the Pentagon's outermost ring. The general wanted to talk cost. He looked down the conference table toward the newly appointed leader of the GPS bomb program, Terry Little. An outspoken and sometimes arrogant civilian manager, Little had spent the last 15 years at Eglin turning around troubled classified weapons programs.

"What's the estimate?" McPeak asked.

"$65,000 apiece," Little replied, looking down at a sheet of figures provided by Defense contractors.

"And what's your goal?" McPeak asked, clearly unhappy with the figure. Little paused for a moment. Then, guessing he might be able to cut the price by a third, he blurted, "$40,000."

"By God, if it's one cent over, I don't want it," McPeak thundered, pounding his fist on the wooden table.

McPeak wasn't bluffing. The Cold War was over; the Reagan-era defense buildup was history. Between 1990 and 1997, the Defense budget would be slashed by a third. Not long after McPeak and the procurement managers met, then-Vice President Al Gore hammered an ashtray to bits on TV's David Letterman show to illustrate the Clinton administration's disdain for costly government specifications. Congress was preparing to pass the 1994 Federal Acquisition Streamlining Act, designed to speed and ease government purchasing of everything from fax machines to weapons.

Little saw his opening. In the classified procurement world, he often stretched buying rules. Contractors and auditors paid less attention to secret programs, and the need to deliver them fast overrode concerns about the rules. Little viewed McPeak's low-price mandate as license to push the boundaries again. This time, the one-time union organizer-who cultivated a maverick image by puffing Lucky Strikes and nipping from a bottle of gin in his desk drawer-would need all the operating room he could get. He had just promised to build a brand new bomb for $40,000 apiece when the cheapest guided bombs to date ran six figures each. Little convinced the Pentagon to waive more than 80 buying rules for the bomb kit contract.

The most important change exempted companies from having to reveal how much they spent designing and building the bomb kits. Normally, Defense requires weapons manufacturers to keep and share meticulous records on the cost of every component of a weapons system. The Pentagon uses cost figures to determine the price it's willing to pay, usually the total manufacturing cost plus a 10 percent to 15 percent profit. Little didn't care what companies spent building bomb kits; his only concern was getting them at the price he'd promised McPeak.

So Little proposed a deal to seven companies that said they could build the bomb kits for $65,000 each. If the companies would agree to build as many bombs as the military wanted for no more than $40,000 apiece with a 20-year warranty, they would not have to reveal their costs, could choose their own suppliers, and the winner would be the sole JDAM supplier in perpetuity. Contractors leapt at the deal. "We were willing to take a risk up front to get a long-term commitment to a price," says Kim Michael, the bomb kit program manager for Boeing, which now owns contract winner McDonnell Douglas.

By early 1994, Lockheed Martin and McDonnell Douglas had beaten five other companies to become finalists. The Air Force and Navy planned to buy 84,000 kits. The two companies had 18 months to come up with the best, most cost-effective design. Such competitions usually are tightly controlled. A cadre of contracting officers obsessed with fairness oversees them. Companies get little feedback from the government until the contract is awarded. But Little believed this approach was too rigid and that the lack of communication led to flawed designs and overpriced products.

Little used another waiver to assign JDAM program officials who had spent their entire careers being tight-mouthed around companies to become contractors' confidants. They were to work side by side with the firms at their plants and at Eglin. They were to show contractors their design weaknesses and likely places to cut costs. They would win or lose together. A few months into the competition, Little set up two teams-each with a dozen engineers, contract specialists and program analysts. He assigned a team to each contractor with orders to do everything possible to help its company win the deal.

"Terry Little told me, 'Your [performance] rating is going to be based on what the McDonnell Douglas program manager thinks of you,' " says retired Air Force Lt. Col. Joseph Schearer, who headed the Eglin team working with the St. Louis-based contractor. Schearer knew the only way he'd win McDonnell Douglas' respect was by helping the company bring home the contract.

Paring Costs

Schearer was concerned he'd drawn the weaker partner in McDonnell Douglas, and his fears were confirmed during his first meeting with the firm late that summer. Long a top but pricey player in the aerospace market, McDonnell Douglas had devised a "Cadillac" design. Shearer figured the elaborate strap-on device would cost nearly $10,000 more than the one being crafted by a Lockheed Martin team behind a closed door in the same building at Eglin. He told his new colleagues they'd lose the work if they didn't redesign the bomb and lower its $28,000 price tag by contract award time in the fall of 1995. That gave them a year.

McDonnell Douglas had little choice but to listen to Schearer. Like most weapons builders, the defense giant had drastically reduced its workforce-from 121,400 to 65,800 employees-as the Pentagon's budget fell between 1988 and 1994. In 1994, the company had lost its biggest missile program, the $1.5 billion Tomahawk cruise missile, to Hughes, mainly due to cost. McDonnell Douglas, which specialized in building planes, had few missile programs left. Without the JDAM, the company would lose its foothold in the munitions business. So the cost cutting began. The team consolidated 11 wires to four, combined multiple circuit boards and replaced an expensive GPS receiver housing with a low-cost casing that protected the entire kit from electromagnetic interference.

McDonnell Douglas cut deals with suppliers to lower the cost of key parts-for example, agreeing to let the manufacturer of a GPS circuit decide which antenna to use in return for a better price. The cost cutting was relentless. "We've got to balance risk with cost," a McDonnell Douglas manager once warned Schearer. But Schearer fired back, "Keep taking out cost. I'll tell you when to stop."

The cutting ended in August 1995. Costs had been trimmed to $6,500 for the inertial navigation system, $3,000 for electric motors and fins, $2,000 for the central computer, and $1,800 for the GPS receiver.

Schearer believed McDonnell Douglas' proposal would be competitive with Lockheed Martin's. He was right. In September 1995, the Air Force announced McDonnell Douglas had won the contract with a price of $14,000 per kit-about half the $28,000 price projected just a year earlier.

Now it was time to prove the bomb could work.

Testing Woes

In October 1996, a 6-foot wooden pole rose from an Eglin test range awaiting obliteration. An Air Force F-16 flew overhead and released a 1,000-pound mix of high explosives and Information Age technology-a test Joint Direct Attack Munition. A few miles away in the test range command and control center, Oscar Soler, the new JDAM program manager, stared at a bank of color television screens waiting for video cameras on the range to pick up the blurry image of the bomb heading toward its target. Ten seconds. Twenty seconds. Forty seconds. Soler grew impatient. "What happened? Where is it?" he asked frantically, cursing at the screen. The 10 airmen monitoring test measurement equipment didn't have to respond. Soler knew the answer: The bomb had gone off target and landed out of camera range. The wooden pole stood in the Florida Panhandle soil unscathed.

The failed test-caused by a faulty wire, not a design problem-was just the latest in a string of JDAM setbacks. Soler seemed to find a new problem at every turn in the 10 months since he had taken over from Little. The fins in the bomb's tail kit were too flimsy. Software controlling the bomb's autopilot system wasn't working. The Pentagon scrutinized the program because its low cost had been such a shock. "I had to go to Washington every month to explain the problems," Soler recalls. A key challenge was convincing JDAM engineers and contract managers that changes or fixes had to be made even if they might drive up the price.

The tail kit problem was a case in point. McDonnell Douglas and Air Force program officials, including Soler, spent about a year trying to stabilize the tail fins. Adding springs didn't work, nor did driving more bolts into the tail. Nothing did. "Finally, it dawned on me that we were so obsessed with acquisition reform that we would not look at design changes that would cost more money," says Soler. Eventually, the laws of physics won out and the fins were redesigned, adding a few hundred dollars to the cost of each kit. "When you try to fix something on the cheap it doesn't work," Soler says.

By 1998, the Air Force had a final design and Boeing, which merged with McDonnell Douglas the previous August, began building the first 476 kits, which would undergo rigorous testing. The test results went to the Pentagon, which weighed them in deciding when the weapon was ready for full production. The JDAM faced final exams at five training and test ranges from Guam to Barksdale, La. Almost immediately, the testers found problems. They worried about the maturity of the commercial software used to program the bombs. They fretted about the cheap fiberglass containers used to ship and store the kits. In a December 2000 final report, they deemed JDAM "operationally effective, but not suitable."

Seeing Action

Meanwhile, in early 1999, the United States and NATO had gone to war in the Balkans against Serbian strongman Slobodan Milosevic. Increasingly, the air war over the former Yugoslavia was encompassing urban areas. War planners feared collateral damage-the unintended killing of civilians who live or might simply be present near allies' targets. Laser-guided munitions weren't a good option because they couldn't negotiate the daily rain, snow, fog and sleet of springtime in the Balkans. The planners asked for better bombs.

Ralston, by 1999 a four-star general serving as vice chairman of the Joint Chiefs of Staff, remembered the bomb he'd fought for 10 years earlier, the one guided from satellites. There's little doubt that as No. 2 military officer at the Defense Department, he had a large say in which weapons would fall on the Serbs that spring. But military leaders of the B-2 bomber wing at Whiteman Air Force Base, Mo., made the final call. The JDAM had done well in testing when dropped from the bat-winged stealth bombers, so the several hundred bombs at the base would be put to use.

Boeing ramped up deliveries from its JDAM plant to Whiteman, a few hours northwest. At the base, JDAM kits were strapped onto 2,000-pound bombs and loaded 16 at a time onto B-2s flying their first combat missions-30 hours to Kosovo and back. The JDAMs saw action almost immediately. Military planners wanted to take out a petroleum plant the Serbs used for refueling. But the oil tanks were located across the road from a centuries-old church. JDAMs destroyed the fuel tanks without cracking a window in the house of worship. Bad weather had prevented the allies from hitting the Zezeljev Bridge stretching over the Danube River in Yugoslavia's second largest city, Novi Sad. JDAMs easily knocked the heavy concrete bridge into the water. The 652 JDAMs dropped during the 78-day air war with Serbia were 98 percent accurate. The testers' concerns were rendered moot. Shortly after the war, the Defense Department placed an order for 84,000 kits. The era of smart bombs had dawned.

A year and half later, in November 2001 in Afghanistan, a Northern Alliance militia leader radioed allied forces that hundreds of Taliban soldiers were gathering on a ridge about a mile from the city of Kunduz, one of the last strongholds of the government that had harbored al Qaeda. In the wake of the Sept. 11 terrorist attacks, the United States was backing the Northern Alliance in a campaign to root out al Qaeda and its supporters. The militia leader requested that the Taliban troops be bombed within 24 hours to enable his forces to advance. Help arrived in 19 minutes. A U.S. Special Forces operator nearby fixed a laser designator on the Taliban troops. He translated their location into GPS coordinates and relayed the request to U.S. combatant commanders headquartered in Saudi Arabia. Headquarters staff contacted B-52 pilots in the area. The pilots programmed the location of the Taliban force into 12 JDAMs in their aircraft. Within minutes the sky rained iron. The Taliban soon gave up Kunduz.

In Afghanistan, JDAMs showed their mettle on short notice, delivering immediate blows to advancing troops. About 6,650 JDAMs have been dropped on Afghanistan since October 2001, and operations continue there. In early 2002, JDAMs had become so popular that the Defense Department nearly tripled its order, to almost 240,000 kits. Boeing suddenly had to churn out 3,000 kits a month from a factory already too small to build half that number.

Smart Bomb-Making

Not long after the Northern Alliance seized Kunduz in early 2002, Boeing introduced lean manufacturing-a process to shorten and streamline production-at its newly enlarged JDAM plant. There, in a barbed wire-ringed complex in St. Charles, Mo., about 20 miles west of downtown St. Louis, first McDonnell Douglas, then Boeing, had built cruise missiles for almost two decades. Building No. 505, where JDAMs are assembled, doesn't stand out, even with its new two-story addition. It's about two hundred yards long and 100 yards across. From the road, it resembles an oversized storage shed. But inside, workers assemble the brains of one of the world's most precise bombs in 78 minutes.

Parts from 22 JDAM suppliers arrive continuously throughout the week. Boeing exercises tight control of its supply chain, through which 887,000 parts moved last year. The company sends its own trucks to transport parts so it can determine when and how often they are picked up and delivered. There are no stockrooms or rows of pallets inside the factory. Boeing saves space and money by storing no more than six days' worth of components. As trucks unload, forklifts carry their cargo directly to one of 20 lines of silver metal rollers that extend 20 feet from loading dock to assembly line. Most parts never touch the factory floor.

The only steady noise inside is the drone of the ventilation system, occasionally interrupted by the high-pitched warning beep of forklifts backing up and the crash of boxes onto the metal conveyors. Nothing's cut or welded here. All the shipping containers are reusable-having to trash them would waste space and cost money. At the end of the metal rollers, a munitions mechanic wearing a special white lab coat to protect circuit boards from static electricity places a GPS circuit card inside a steel 12-inch control unit. He grabs a black automatic screwdriver hanging overhead. Twenty-seven turns later, half-inch silver screws connect the card to the control unit. In six minutes, the mechanic creates a bomb's sense of direction. Forty-two munitions mechanics on two shifts working five days a week can build 3,000 JDAM kits each month.

Hiss. Thwap. Hiss. Thwap. An automatic machine whips thick translucent green plastic straps around a dark green fiberglass box. The boxes carry JDAM kits out the door and around the world. Stacked two high and three across, 54 containers holding 108 kits can fill one 18-wheeler. Most days, the factory fills a trailer and a half. In May, a truck pulled out, bound for an Army arsenal; Boeing wouldn't say where. Upon arrival, the kits it carried likely were stored in temperature-controlled bunkers alongside the bombs they would one day steer. The Defense Department generally gives Boeing a week's notice of a shipment's destination, but that can change en route. One destination is known: At least a half dozen 18-wheelers full of JDAMs wound up on the aircraft carrier USS Constellation heading for the Persian Gulf.

Baghdad Bound

In mid-March, bomb builders clad in red turtlenecks and camouflage pants hustled about the fluorescent-lit bunker inside the Constellation. The young sailors had been working 18 hours, seven days a week, for more than a month assembling bombs, including JDAMs, for war with Iraq. The "Bomb Build-Up Chart" on a whiteboard detailed which tails and fins were to go on which bombs and how many were to be built each day. "I can build two or 2,000 bombs," said Gunner Joe Thompson, a former enlisted man turned naval officer who managed the Constellation's 108 ordnance handlers.

Bombs come in pieces-munitions in wooden crates and tail kits encased in fiberglass. It takes two ordnance men about 30 minutes to bolt silver tail kits to the dull gray bombs sporting two yellow rings around their noses. After a quick electronics test, the bombs are stored. On the eve of war, bombs were stacked three-high, 10 rows deep in the ship's arsenal awaiting the elevator trip up 10 stories to the flight deck. On a clear night during the third week of March, the wind whips hard across the Constellation's flight deck. It takes two or three ordnance men to pull two JDAMs in a small cart toward the fighter planes parked diagonally at the flight deck's edge. They attach two bombs under each wing of the F-14 Tomcats and F-18 Hornets. Five minutes later, pilots climb in and roll the planes to the ship's steam-powered catapult, where they're latched on. A safety officer taps a white light on the deck.

Yanked by the catapult, the planes jerk upright and accelerate from zero to 150 miles an hour in two seconds, barreling down the deck and off the carrier. Each one hovers precariously just off the ship's bow, weighed down by 4,000 pounds of bombs. Then flames shooting out the tail from the afterburners grow longer and whiter. Each fighter begins to climb and disappears into the stars over the Persian Gulf to deliver its JDAMs to Iraq.