'Railhead' Project Derailed by Technical Failures, Contractor Mismanagement

Bob O'Harrow at the Washington Post reported this week that a $500 million project to upgrade the National Counterterrorism Center's (NCTC's) terrorist watch list has been crippled by technical failures and contractor mismanagement.

The Terrorist Identities Datamart Environment (TIDE) contains centralized data on more than 400,000 known or suspected terrorists, and delivers daily updates to the nation's 16 intelligence agencies (one former NCTC director described it as "the mother of all databases").  A February 2008 report by the Office of the Director of National Intelligence (ODNI) declared that TIDE is an "innovative solution...to increase information sharing and collaboration in support of the counterterrorism mission."

But according to documents obtained by the Post and a memo prepared by staff on the House Science and Technology Committee's Investigations and Oversight Subcommittee, TIDE has been much more of a problem than a solution.  For instance, it was discovered that an earlier version of the database was only storing 60 percent of the data from CIA cables.  There have also been critical problems with the database's search engine.

In 2006, the NCTC initiated a five-year, $500 million program called "Railhead" to upgrade TIDE and other information technology systems, and to improve the sharing of counterterrorism intelligence data throughout the government.  However, internal documents show that the Railhead project has suffered from additional technical problems, contractor mismanagement, and weak government oversight.  For example:

• The new database will not provide access to the websites and data sources of the CIA, DIA, FBI, NSA, and other agencies;
• Software tests have uncovered critical errors in the new database's storage and search functions;
• Five task orders worth an estimated $92.9 million are "significantly off-plan";
• Nearly a dozen government positions on the Railhead project have remained vacant for over a year; and
• Insiders have raised questions about the government's apparent use of nearly $200 million to retrofit a Boeing office building in Herndon, Virginia, so that it could house some of the 800 contractor employees working on the Railhead project.

In recent weeks, the majority of the contractor employees working on Railhead have been laid off, leaving the project on the "brink of collapse" according to a press release by the Investigations and Oversight Subcommittee.  Rep. Brad Miller (D-NC), Chairman of the Subcommittee, has written a letter to the ODNI Inspector General calling for a full investigation into the program.

It's worth noting that a portion of the Railhead program has been managed by a lead systems integrator (LSI), Boeing's Space and Intelligence Systems Mission Systems division.  An LSI is a contractor or team of contractors "hired by the federal government to execute a large, complex, defense-related acquisition program."  POGO has been a longtime critic of the LSI model, which has resulted in cost overruns and schedule delays on the Army's Future Combat Systems and the Coast Guard's Integrated Deepwater System program.  Program management is a perfect example of an inherently governmental function that should not be left to the private sector.

We've also argued that the use of LSIs increases the potential for conflicts of interest, since the LSI might favor its own subsidiaries or subcontractors when reviewing contract proposals.  Speaking of conflicts of interest, the Investigations and Oversight Subcommittee staff have raised questions about a cozy relationship between the government's Railhead program manager, Dirk Rankin, and the program manager from one of the Railhead contractors, SRI International.

-- Michael Smallberg

Comments

[Mike]

What is it about those doing procurement and contracting in the 'IC' (Intel Community). They consistently fall into this trap. Simple math suggests that more building space and more people equal significantly more dollars, not to mention that history suggests that this approach tends not to work more than it tends to work.

We don't need a massive overhead approach to solve our technology problems within the IC, but rather a smaller, leaner, smarter group of the right professionals that know their business to solve these problems and put real solutions on the table within reasonable time constraints. We have proven that this is the right approach, but we haven't advertised those wins, so nobody seems to know that it can be accomplished without the excessive overhead. MM