The Coast Guard seems to be joining the chorus that fixing the government's acquisition system isn't about large overhauls and new rules. On Tuesday, I had the opportunity to interview Rear Admiral Gary Blore, the Assistant Commandant for Acquisition at the Coast Guard, as part of Department of Defense's Bloggers Roundtable. Our conversation focused on the Coast Guard's troubled Deepwater program, which Blore cited as an example of an innovative concept that was exciting "academically" as a possibility for transforming how the Coast Guard does acquisition, but never demonstrated maturity or innovation. "We would've been better off just using the usual techniques," he said.
This isn't a groundbreaking conclusion. In his March testimony before the House Subcommittee on Coast Guard & Maritime Transportation, he talked about how the Coast Guard is trying to depart from the Lead Systems Integrator concept, and heralded the eight cornerstones for successful acquisition: checks and balances, relying on the Coast Guard for final certification, a reliable and standard reference for acquisition management, robust strategic planning, transparency, avoidance of duplication, independent validation, and renewed oversight.
And on that last point of the need for renewed oversight, it seems like there is still work to be done. One of the primary issues POGO has focused on is the program's past failure to meet government-wide communications protection standards, commonly referred to as TEMPEST. Admiral Blore told me that both the visual and the instrumented tests for the TEMPEST were completed in April. But follow-up questions on that statement suggest that it isn't the case at all.
An important element for testing the TEMPEST is the Sensitive Compartmented Information Facilities (SCIF), which have not yet been installed. Currently, the plan is to have the SCIF installed and operational in March 2010. When I asked a Coast Guard representative if they would have to conduct more testing following the SCIF's installation, they confirmed that they would.
Which can only lead me to conclude that depending on your definition of "done"--and I favor a definition of operational testing, in which a system has demonstrated stability in an environment that closely parallels how it will have to actually function--the Deepwater program is still missing the key technology whistleblower Mike DeKort raised concerns about in 2003. The GAO is also concerned that the program is continuing procurement without completing operational testing.
The recent defense acquisition reform work demonstrates a recognition that major defense acquisition programs (MDAPs) experience more cost and schedule success if they don't wait until technology is tested and proven before ramping up production. The Coast Guard should cross-apply that analysis and not only be more transparent about the problems that exist, but also be wary of ramping up other Deepwater assets until the TEMPEST is proven. Sources tell POGO that there are still significant design problems that Congress should investigate and consider holding people responsible for. Without action, the problems and the fixes necessary for TEMPEST will only carry on to the other assets.
-- Mandy Smithberger