How Technology Failed in Iraq

The largest counterattack of the Iraq War unfolded in the early-morning hours of April 3, 2003, near a key Euphrates River bridge about 30 kilometers southwest of Baghdad, code-named Objective Peach. The battle was a fairly conventional fight between tanks and other armored vehicles – almost a throwback to an earlier era of war fighting, especially when viewed against the bloody chaos of the subsequent insurgency. Its scale made it the single biggest test to date of the Pentagon’s initial attempts to transform the military into a smaller, smarter, sensor-dependent, networked force.

In theory, the size of the Iraqi attack should have been clear well in advance. U.S. troops were supported by unprecedented technology deployment. During the war, hundreds of aircraft- and satellite-mounted motion sensors, heat detectors, and image and communications eavesdroppers hovered above Iraq. The four armed services coordinated their actions as never before. U.S. commanders in Qatar and Kuwait enjoyed 42 times the bandwidth available to their counterparts in the first Gulf War. High-bandwidth links were set up for intelligence units in the field. A new vehicle-tracking system marked the location of key U.S. fighting units and even allowed text e-mails to reach front-line tanks. This digital firepower convinced many in the Pentagon that the war could be fought with a far smaller force than the one it expected to encounter.

Yet at Objective Peach, Lt. Col. Ernest “Rock” Marcone, a battalion commander with the 69th Armor of the Third Infantry Division, was almost devoid of information about Iraqi strength or position. “I would argue that I was the intelligence-gathering device for my higher headquarters,” Marcone says. His unit was at the very tip of the U.S. Army’s final lunge north toward Baghdad; the marines advanced on a parallel front. Objective Peach offered a direct approach to the Saddam International Airport (since rechristened Baghdad International Airport). “Next to the fall of Baghdad,” says Marcone, “that bridge was the most important piece of terrain in the theater, and no one can tell me what’s defending it. Not how many troops, what units, what tanks, anything. There is zero information getting to me. Someone may have known above me, but the information didn’t get to me on the ground.” Marcone’s men were ambushed repeatedly on the approach to the bridge. But the scale of the intelligence deficit was clear after Marcone took the bridge on April 2.

As night fell, the situation grew threatening. Marcone arrayed his battalion in a defensive position on the far side of the bridge and awaited the arrival of bogged-down reinforcements. One communications intercept did reach him: a single Iraqi brigade was moving south from the airport. But Marcone says no sensors, no network, conveyed the far more dangerous reality, which confronted him at 3:00 a.m. April 3. He faced not one brigade but three: between 25 and 30 tanks, plus 70 to 80 armored personnel carriers, artillery, and between 5,000 and 10,000 Iraqi soldiers coming from three directions. This mass of firepower and soldiers attacked a U.S. force of 1,000 soldiers supported by just 30 tanks and 14 Bradley fighting vehicles. The Iraqi deployment was just the kind of conventional, massed force that’s easiest to detect. Yet “We got nothing until they slammed into us,” Marcone recalls.

Objective Peach was not atypical of dozens of smaller encounters in the war. Portions of a forthcoming, largely classified report on the entire Iraq campaign, under preparation by the Santa Monica, CA, think tank Rand and shared in summary with Technology Review, confirm that in this war, one key node fell off the U.S. intelligence network: the front-line troops. “What we uncovered in general in Iraq is, there appeared to be something I refer to as a ‘digital divide,’” says Walter Perry, a senior researcher at Rand’s Arlington, VA, office and a former army signals officer in Vietnam. “At the division level or above, the view of the battle space was adequate to their needs. They were getting good feeds from the sensors,” Perry says. But among front-line army commanders like Marcone – as well as his counterparts in the U.S. Marines – “Everybody said the same thing. It was a universal comment: ‘We had terrible situational awareness,’” he adds. The same verdict was delivered after the first Gulf War’s ground battle, but experts had hoped the more robust technology used in the 2003 conflict would solve the problem.

The Pentagon points to the Iraq War’s many networking successes. During the blinding sandstorm that lasted from March 25 to 28, 2003, a U.S. radar plane detected an Iraqi Republican Guard unit maneuvering near U.S. troops. Bombers moved in to attack using satellite-guided bombs that were unaffected by poor visibility. And the vehicle-tracking system (known as Blue Force Tracker) successfully ensured that commanders knew the locations of friendly units. Overall, command headquarters in Qatar and Kuwait sported “truly a very impressive digital connectivity” that “had many of the characteristics of future network warfare that we want,” Brig. Gen. Robert Cone, then director of the Pentagon’s Joint Center for Operational Analysis and Lessons Learned, said in a Pentagon briefing last year.

Yet connectivity in Qatar was matched by a data dearth in the Iraqi desert. It was a problem all the ground forces suffered. Some units outran the range of high-bandwidth communications relays. Downloads took hours. Software locked up. And the enemy was sometimes difficult to see in the first place. As the marines’ own “lessons learned” report puts it, “The [First Marine] Division found the enemy by running into them, much as forces have done since the beginning of warfare.” Describing the army’s battle at Objective Peach, John Gordon, another senior researcher at Rand and also a retired army officer, put it this way: “That’s the way it was done in 1944.”

Military intellectuals call them “revolutions in military affairs.” Every few decades, a new technology or a new “doctrine,” to use the military jargon, changes the nature of war. Single technologies, like gunpowder or nuclear weapons, spur some of these revolutions. New doctrines, like Napoleonic staff organization or Nazi blitz tactics, drive others. And some are the result of many simultaneous advances, like the airplanes, chemical weapons, and machine guns of World War I – which achieved new rates of slaughter.

The newest revolution is known to Pentagon planners as “force transformation.” The idea is that robotic planes and ground vehicles, empowered by an ever expanding range of sensing, targeting, imaging, and communications capabilities (new technologies), would support teams of networked soldiers (a new doctrine). According to its most expansive definition, force transformation is intended to solve the problem of “asymmetric warfare” in the 21st century, where U.S. forces are not directly confronted by conventional militaries but rather must quell insurgencies, destroy terrorist cells, or mitigate regional instability. Among other things, more nimble, networked forces could employ tactics like “swarming” – precise, coordinated strikes from many directions at once.

The technologies driving force transformation are incredibly complicated. It will take at least 31 million lines of computer code to run something called Future Combat Systems, the centerpiece of the Pentagon’s transformation effort. An army-run program expected to cost more than $100 billion, it consists of a suite of new manned and unmanned machines, all loaded with the latest sensors, roaming the air and ground. Software will process sensor data, identify friend and foe, set targets, issue alerts, coordinate actions, and guide decisions. New kinds of wireless communications devices – controlled by yet more software and relaying communications via satellites – will allow seamless links between units. Currently, 23 partner companies, many with their own platoons of subcontractors, are building the systems; Boeing of Chicago and Science Applications International of San Diego are charged with tying them all together and crafting a “system of systems” by 2014.

In this grand vision, information isn’t merely power. It’s armor, too. Tanks weighing 64 metric tons could be largely phased out, giving way to lightly armored vehicles – at first, the new 17-metric-ton Stryker troop carrier – that can avoid heavy enemy fire if need be. These lighter vehicles could ride to war inside cargo planes; today, transporting large numbers of the heaviest tanks requires weeks of transport via land and sea. “The basic notion behind military transformation is that information technologies allow you to substitute information for mass. If you buy into that, the whole force structure changes,” says Stuart Johnson, a research professor at the Center for Technology and National Security Policy at National Defense University in Washington, DC. “But the vision of all this is totally dependent on information technologies and the network. If that part of the equation breaks down, what you have are small, less capable battle platforms that are more vulnerable.”

The Iraq War represented something of a midpoint – and an early proving ground – in the move toward this networked force. The U.S. offensive did include the old heavy armor, and it didn’t sport all the techno-goodies envisioned by the promoters of force transformation. But it did presume that satellite- and aircraft-mounted sensors would support the fighting units on the ground. The war’s backbone was a land invasion from Kuwait. Ultimately, some 10,000 vehicles and 300,000 coalition troops rumbled across the sandy berm at the Kuwaiti border, 500 kilometers from Baghdad. Desert highways crawled with columns of Abrams tanks, Bradley fighting vehicles, armored personnel carriers, tank haulers, Humvees, and of course, fuel tankers to slake the fleet’s nine-million-liter daily demand for fuel.

Several communications links were designed to connect these vehicles with each other and with commanders. First, and most successfully, at least 2,500 vehicles were tracked via Blue Force Tracker: each vehicle broadcast its Global Positioning System coordinates and an ID code. This thin but critical stream of data was in essence a military version of OnStar. Commanders in Qatar saw its content displayed on a large plasma screen. Marcone, like some other commanders in the field, also had access to it, thanks to a last-minute installation in his tank before the invasion.

Once the invasion began, breakdowns quickly became the norm. For the movement of lots of data – such as satellite or spy-plane images – between high-level commanders and units in the field, the military employed a microwave-based communications system originally envisioned for war in Europe. This system relied on antenna relays carried by certain units in the advancing convoy. Critically, these relays – sometimes called “Ma Bell for the army” – needed to be stationary to function. Units had to be within a line of sight to pass information to one another. But in practice, the convoys were moving too fast, and too far, for the system to work. Perversely, in three cases, U.S. vehicles were actually attacked while they stopped to receive intelligence data on enemy positions. “A lot of the guys said, ‘Enough of this shit,’ and turned it off,” says Perry, flicking his wrist as if clicking off a radio. “ ‘We can’t afford to wait for this.’”

One Third Infantry Division brigade intelligence officer reported to Rand that when his unit moved, its communications links would fail, except for the GPS tracking system. The unit would travel for a few hours, stop, hoist up the antenna, log back onto the intelligence network, and attempt to download whatever information it could. But bandwidth and software problems caused its computer system to lock up for ten to 12 hours at a time, rendering it useless.

Meanwhile, commanders in Qatar and Kuwait had their own problems. Their connectivity was good – too good. They received so much data from some of their airborne sensors that they couldn’t process it all; at some points, they had to stop accepting feeds. When they tried to send information to the front, of course, they found the line-of-sight microwave-relay system virtually disabled. At the command levels above Marcone’s – the brigade and even the division levels – such problems were ubiquitous. “The network we had built to pass imagery, et cetera, didn’t support us. It just didn’t work,” says Col. Peter Bayer, then the division’s operations officer, who was south of Marcone’s battalion on the night of April 2 and 3. “The link for V Corps [the army command] to the division, the majority of time, didn’t work, to pass a digital image of something.”

Sometimes, intelligence was passed along verbally, over FM radio. But at other times vehicles outran even their radio connections. This left just one means of communication: e-mail. (In addition to tracking vehicles, Blue Force Tracker, somewhat quaintly, enabled text-only e-mail.) At times, the e-mail system was used for issuing basic orders to units that were otherwise out of contact. “It was intended as a supplement, but it wound up as the primary method of control,” says Owen Cote, associate director of the Security Studies Program at MIT. “The units did outrun their main lines of communications and networking with each other and with higher command. But there was this very thin pipe of information via satellite communications that allowed the high command to see where units were.”

The network wasn’t much better for the marines pushing forward on a separate front. Indeed, the marines’ lessons-learned report says that First Marine Division commanders were unable to download crucial new aerial reconnaissance photographs as they approached cities and towns. High-level commanders had them, but the system for moving them into the field broke down. This created “a critical vulnerability during combat operations,” the report says. “There were issues with bandwidth, exploitation, and processes that caused this state of affairs, but the bottom line was no [access to fresh spy photographs] during the entire war.”

Fortunately for U.S. forces, they faced little resistance during the Iraq War. The Iraqis launched no air attacks or Scud missiles. Iraqi soldiers shed uniforms and boots and walked away barefoot, studiously avoiding eye contact with the Americans. When they did fight, they used inferior weapons and vehicles. To be sure, U.S. units racing forward would run into stiff “meeting engagements” – jargon for a surprise collision with enemy forces. But such meetings would end quickly. “They [the U.S. forces] would succeed in these meeting engagements,” Cote says. “But we were far from the vision of total knowledge. You can easily see how we would have paid a big price if it were a more robust opponent.”

The problems are acknowledged at high levels. However, Art Cebrowski, retired vice admiral and director of the Pentagon’s Office of Force Transformation, cites “existence proofs” that networking was generally successful in Iraq. In previous conflicts, combat pilots were briefed on targets before takeoff; hours would elapse between target identification and an actual attack. In the Iraq War, more than half of aerial sorties began without targets in mind, Cebrowski says. Instead, targets were identified on the fly and communicated to the airborne pilots. “Combat was moving too fast; opportunities were too fleeting. You had to be in the networked environment” for it to work, says Cebrowski.

Clearly, networking during the ground war was not as successful. “There were certainly cases where people didn’t have the information they needed. This was a very large operation, so you would expect to see the good, the bad, and the ugly in it,” Cebrowski acknowledges. But it would be a mistake to use these problems as an argument against phasing out heavy armor, he says. Big tanks require not only considerable time and energy to move into battle but also larger supply convoys that are themselves susceptible to attack. According to Cebrowski, by keeping heavily armored tanks your main line of defense, “you simply move your vulnerability to another place on the supply chain.”

Some defenders of force transformation argue that the troops’ problems were doctrinal, not technological. According to this line of reasoning, the networking of the Iraq War was incomplete – because it was fatally grafted onto old-fashioned command and control systems. Sensor information went up the chain of command. Commanders interpreted it and made decisions. Then they passed commands, and tried to pass relevant data, down the chain. The result: time delays and the magnification of individual communications failures.

Better, some say, that information and decision-making should flow horizontally. In fact, that’s how the 2001 war in Afghanistan was fought. Special-operations forces organized into “A teams” numbering no more than two dozen soldiers roamed the chilly mountains near the Pakistan border on horseback, rooting out Taliban forces and seeking al-Qaeda leaders. The teams and individuals were all linked to one another. No one person was in tactical command.

But despite the lack of generals making key decisions, each of these teams of networked soldiers had a key node, an animal once confined to corporate IT departments: the alpha geek, who managed the flow of information between his team and the others. The U.S. special forces also maintained a tactical Web page, collating all the information the teams collected. And this page was managed by a webmaster in the field: the metageek of all alpha geeks.

How did the page perform? Postmortems and reports on special-forces operations in Afghanistan are more secret than those from the Iraq War. A report on one major special-forces operation, Operation Anaconda – an attempt to encircle and root out al-Qaeda in March 2002 – is due soon from National Defense University. Still, anecdotes are trickling out of the special-forces community. And they provide a startlingly different view of warfare than Marcone’s tank-level vantage. One account, not previously reported, comes from John Arquilla, an expert in unconventional warfare at the Naval Postgraduate School in Monterey, CA.

The scene was a cold night in the late fall of 2001. In New York City, the World Trade Center ruins were still smoldering. In Afghanistan, a U.S. Air Force pilot en route from Uzbekistan noticed flashing lights in the mountains below, near the Pakistan border. Suspecting that the flashes might be reflections from hooded headlights of trucks bumping along, he radioed his observation to the webmaster. The webmaster relayed the message across a secure network accessible to special forces in the region. One team replied that it was near the position and would investigate. The team identified a convoy of trucks carrying Taliban fighters and got on the radio to ask if any bombers were in range. One U.S. Navy plane was not far off. Within minutes, the plane bombed the front and rear of the convoy, sealing off the possibility of escape. Not long after, a gunship arrived and destroyed the crippled Taliban column.

The episode, as recounted by Arquilla, shows what’s possible. “That’s networking. That’s military transformation right there,” Arquilla says. “Some of the problems in Iraq grew out of an attempt to take this cascade of information provided by advanced information technology and try and jam it through the existing stovepipes of the hierarchical structure, whereas in Afghanistan we had a more fluid approach. This is war by minutes, and networking technology allows us to wage war by minutes with a great probability of success.” In this case, service members on the battlefield collected data, shared that data, made decisions, and ordered strikes.

Perhaps Pentagon optimists are right. Perhaps the success of Blue Force Tracker, of the special-forces assault on the Taliban column, and of air force operations in Iraq accurately foretell the full digital transformation of war. But to many observers, the disruption of communications between the main ground combat units in Iraq was not a very promising sign at all. “If there is this ‘revolution in military affairs,’ and if this revolution is based on technologies that allow you to network sensors and process information more quickly and spread it out quickly in digestible form, we are still just scratching the surface of it,” says Cote of MIT. “If you look at the performance of a lot of the components of the first efforts in that direction, it’s a pretty patchy performance.” And then there’s the question of terror and insurgency. Even if the Pentagon transforms war fighting, the meaning of the word “war” is itself undergoing a transformation. More Americans died in the September 11 attacks than have subsequently died in Afghanistan and Iraq. And the Iraq insurgency challenges the meaning of the Iraq military victory. Future wars will be fought in urban zones by low-tech fanatics who do not follow the old rules. They are unlikely to array themselves as convenient targets for the U.S. to detect and destroy. Indeed, a leading cause of death among U.S. soldiers in Iraq today is improvised bombs targeting passing vehicles such as Humvees.

Arquilla says some networking technology can be – and is being – brought to bear against the Iraq insurgency. While actual strategies are secret, some general tactics are known. Suspicious vehicles can be tracked, and their connections to other people and locations determined. Small drone aircraft can deliver video feeds from urban buildings as well as from desert battlefields. Sensors can help find a sniper by measuring the acoustical signature of a bullet. And jamming devices can sometimes block radio-controlled detonation of roadside bombs. But old-fashioned tips from humans are likely to trump technology. “Our networks don’t really have the sensitivity to keep up with unconventional enemies. All the network does is move information around, but the information itself is the key to victory,” says Loren Thompson, chief operating officer of the Lexington Institute, a think tank in Arlington, VA. “It’s a little hard to derive meaningful lessons from networked war fighting when you are dealing with such modest threats.”

The welter of postmortems from the Iraq and Afghanistan wars tell many stories. But one thing is clear: Marcone never knew what was coming at Objective Peach. Advanced sensors and communications – elements of future networked warfare designed for difficult, unconventional battles – failed to tell him about a very conventional massed attack. “It is my belief that the Iraqi Republican Guard did nothing special to conceal their intentions or their movements. They attacked en masse using tactics that are more recognizable with the Soviet army of World War II,” Marcone says.

And so at a critical juncture in space (a key Euphrates bridge) and time (the morning of the day U.S. forces captured the Baghdad airport), Marcone only learned what he was facing when the shooting began. In the early-morning hours of April 3, it was old-fashioned training, better firepower, superior equipment, air support, and enemy incompetence that led to a lopsided victory for the U.S. troops. “When the sun came up that morning, the sight of the cost in human life the Iraqis paid for that assault, and burning vehicles, was something I will never forget,” Marcone says. “It was a gruesome sight. You look down the road that led to Baghdad, for a mile, mile and a half, you couldn’t walk without stepping on a body part.”

Yet just eight U.S. soldiers were wounded, none seriously, during the bridge fighting. Whereas U.S. tanks could withstand a direct hit from Iraqi shells, Iraqi vehicles would “go up like a Roman candle” when struck by U.S. shells, Marcone says. Sitting in an office at Rand, Gordon puts things bluntly: “If the army had had Strykers at the front of the column, lots of guys would have been killed.” At Objective Peach, what protected Marcone’s men wasn’t information armor, but armor itself.