Case Study: Borderline Success at the Department of Homeland Security - ' Piece'

The staged development-and-release methods behind US-VISIT set it apart from typical federal IT project planning, which has tended toward long timetables that promise huge systems to be delivered at a single stroke, says Cooper. He says it is possible to recognize the project's shortcomings in terms of both management and deployment, and still discern in its history a new model for building federal systems.

Maybe, he suggests, it's time to ditch the traditional big-bang approach for a release-and-refine method. Change of some sort seems a necessity, given the miserable record of bloated government IT jobs, which overrun deadlines and budgets with numbing frequency.

"It's probably true in a lot of places where the government makes critical policy decisions that the most important priority should be followed by the next most important, to start on a prioritized list and see where it goes," agrees Jim Stolarski, the US-VISIT project manager for lead integrator Accenture. While DHS has made most of the decisions on the project, he says, "We tend to favor a phased approach, especially in an environment like this one where there is still learning going on."

Private companies, notes Cooper, are already less likely than in the past to engage in "zillion-dollar programs." Instead, he says, "they are defining smaller chunks that fit together into larger systems. It's a model worth exploring." He looks to the Office of Management and Budget, which guides many federal departments, along with the Federal CIO Council, a group of top technology executives at government agencies, for leadership on the way government programs are deployed. Nobody would recommend that the hurry-up methodology behind US-VISIT be copied exactly, but there is a case to be made for learning from it in positive ways, as well as negative ones.

DHS backed into its iterative strategy on US-VISIT. There was just no way, says Cooper, to do everything at once and still hope for a timely release of the most important piece—the entry-screening system. So that became the priority: Tie key immigration databases and watch lists together, and get some biometrics to U.S. Customs checkpoints, pronto; the latest and greatest screening systems could replace available tools at a later date. Then, roll out things like RFID readers for new, chip-bearing passports. And only then address the exit portion of the system, which was seen as less critical to national security, and which involved infrastructure and regulatory challenges as well as technology solutions. The benefit of catching miscreants entering the U.S. was obvious, and the costs of waiting on the exit system seemed bearable.

By rolling the project out in stages, DHS was able to rapidly deploy a working version of the system most critical to national security. The biometric entry-check network isn't perfect. It will have to be upgraded, and its companion pieces must be added over time. But as the saying goes, The perfect is the enemy of the good; US-VISIT got something good done, fast.

Next page: Development on the Run

1 | 2 | 3 | 4