Mapping Route to Obsolescence

What began as an ambitious Defense Department effort in 1982 to leapfrog existing technology in crafting military maps ended a decade later in an obsolete computer system.

The computerized mapping system, which was built by the Defense Mapping Agency at a cost of at least $2.6 billion, had been designed to meet the military threat from the Soviet Union. Building it took so long that the Cold War ended before the system was completed.

At its worst, the mapping computer was riddled with an estimated 82,000 software defects, critics say, and it never achieved the efficiencies in map-making that had been expected.

Just a few months ago, in an action based in part on the problems with the mapping system, Defense Department officials said they had abolished the Defense Mapping Agency and handed over its operations to a new organization.

The problems at the mapping agency have drawn none of the publicity that has accompanied, for example, the IRS’ faltering efforts to automate the processing of tax returns.

Yet the map agency’s problems illustrate many of the pitfalls that have plagued a host of the federal government’s large information-technology programs for the past 20 years.

Overly ambitious goals were set. Accountability was not clear. The timetable was far too long. Much of the work was conducted in secrecy. Some inside critics were swept aside. The agency lacked the talent for the complexity of the task. Federal regulations bogged down the contracting process. Some top officials jumped ship to work for the outside contractors building the system.

The lament of the military about the inadequacy of maps is even older than the country. In the heat of the Revolutionary War, Gen. George Washington once complained, “The want of accurate maps has been a grave disadvantage to me.”

Latter-day generals are in a similar fix. When the Cold War ended, military leaders examined the new era of military threats and the need to respond quickly in unfamiliar places. They realized that the mapping computer--known as the digital production system--was poorly suited to the post-Cold War need to quickly produce and distribute maps of various regions.

In hindsight, few defense officials boast about the project.

“It took an awful long time,” acknowledged Emmett Paige Jr., the Defense Department’s chief information officer. “It used technology that was current then but, by today’s standards, is obsolete.”

*

Military commanders, meanwhile, face a 15-year backlog for maps they have ordered, according to a Defense Science Board report issued last year.

After the report was issued, the agency decided the need for the maps was “obsolete.” The Defense Department merged the mapping agency into the National Imagery and Mapping Agency, which includes other intelligence offices.

The digital production system was intended to replace a large part of the laborious and painstaking process of making maps by hand. The new computer used sophisticated imagery software to construct maps from the agency’s database of photographic images taken by spy satellites.

Mapping agency officials conceded they are not sure even now if productivity has improved with the system. This is a typical problem in government, where performance measures often do not exist. Paige said it became clear to him two years ago, however, that the mapping system had grown too costly to operate and that the Pentagon needed a modern, flexible system.

The agency produces some garden-variety road maps. But it also makes topographic maps, aeronautical charts, oceanic charts and maps that show gravity variations around the globe. It has maps of railroads, dams, power lines and much else.

About half the data used by the agency to make maps comes from spy satellites, according to the science board report, and the rest from foreign maps, other foreign sources and oceanic surveys.

Many of the problems in the new program were predicted at the outset by Basharat Jamil, a mathematician at the mapping agency who said he was pushed out of his job after he began to express his concerns.

Jamil, a member of the technical evaluation team for the project in 1982, recalled warning the agency that it was moving too fast in awarding contracts and lacked sufficient in-house skills to manage a large number of outside companies.

Indeed, the agency had decided that it alone would be the systems integrator, taking responsibility that all parts of the automated mapping system would work together. It assigned major contracts to Hughes Electronics Corp., Intergraph Corp., Lockheed Martin Corp., E-Systems Inc. and General Dynamics Corp..

“We needed a certain expertise to talk to contractors, and we didn’t have even that,” Jamil said.

According to three other mapping agency employees, who requested anonymity, the system breaks down frequently and has never fulfilled its original goals. It requires substantially more human involvement to operate than had been anticipated.

“The original expectations were way out of line,” said a cartographer at the agency. “They had no idea of what they were doing. No system can mimic the accuracy of the human eye. A fully automated system is a pipe dream, or it will be for another 25 or 50 years.”

An engineer at the agency delivered a terser verdict: “The whole thing is a big fiasco and doesn’t work.”

The Defense Department inspector general investigated the mapping computer in 1994 and substantiated a number of allegations that Jamil and others had made. It was the inspector general’s report that revealed the system had 82,000 software defects and, in some cases, was not performing as planned.

The report also found that the mapping agency had maintained secrecy on the program far too long, even classifying the identity of the contractors. It said the Pentagon had failed to conduct external reviews and that the program was exempt from many of the Pentagon’s acquisition safeguards.

Agency officials dismissed much of the criticism, saying the system is a backbone of its operation that produces maps daily, though not at the speed originally expected.

“It is a success story,” said Roberta Lenczowski, associate deputy director for operations. “What [the digital production system] did was drive the state of the art for the remarkable commercial capability that exists today.”

*

Some of the mapping agency’s top executives have moved to the very companies that are increasingly performing the mapping agency’s work under contract, using better technology and commercially available satellite photographs. They include the former mapping agency deputy director and former civilian director.

Jamil was stripped of his responsibilities 11 years ago and subsequently filed for workers’ compensation, citing job stress. He has been conducting an unsuccessful legal battle to get his job back.

In a series of letters to Defense Secretary William J. Perry, Jamil has alleged that rampant fraud and waste went unchecked in the mapping project and should be investigated.

Jamil also has contended that the official $2.6-billion estimate for the cost of the system is substantially understated and that total spending on the program probably exceeded $10 billion, a claim backed up by Patrick Weed, the former president of the union representing agency employees.

Defense officials dismissed the assertion that the system cost more than $2.6 billion. Paige, the assistant defense secretary, said Pentagon officials had investigated Jamil’s allegations and that the matter was now closed.

Meanwhile, the new National Imagery and Mapping Agency has already begun to abandon parts of the digital production system, and plans over the next year to find commercially available replacements for much of the equipment.