I'm on Twitter! More Must Reads

    follow me on Twitter

    Friday, January 16, 2009

    313 Ship Navy? Never going to happen.

    National Defense Magazine this month discusses the Navy's efforts to get to 313 ships. Their conclusion: It ain't gonna happen. In fact, they think the odds are quite good that we're headed down from the 280 odd ships we have today, and they recommend going from 14 to 10 SSBN's and reducing the number of carriers to 10.

    I have been discussing this with some surface warfare friends of mine, and they agree that efforts like the LCS, the DDG-1000, and the CG(X) are all flawed. The Navy has serious issues with costs and delivery schedules on these projects. Interesting that, just like the Seawolf, the SWO's are going to get 3 DDG-1000's. The LCS, slated for a class of 55, is in serious danger, as well.

    Is the Burke design so old that it can't be used again? Are our enemies so sophisticated that the core platforms (hull design and propulsion plant) can't be reused? Just a thought.

    What is to blame? Well, just as most I/T projects are derailed by an inability to hold the line on new requirements after coding begins, the Navy has been unable to restrain itself in issuing new requirements after construction has begun. The Navy is right that it is impossible anymore to wait until a design is finalized before beginning construction, but... a little discipline has to be exercised.

    For example:
    On the Seawolf, only about 6 percent of the design was complete before construction began in the late 1980s. About 68,000 changes were made on the ship as it went through the construction process. When EB later set out to build the Virginia-class submarine, its engineers had completed about 50 percent of the design before bending steel. That ship experienced 12,000 changes.

    This is all pretty basic stuff, whether you're delivering custom software, integrating large information systems, or building ships.

    Amazing that in a $500B DoD budget, we can't get but $15B for shipbuilding.


    No comments: