Speed to Market vs. End User Needs: Homeland Security Edition

(Via Good Experience)

I remember Sen. Trent Lott once crowing about bringing federal dollars to a Mississippi military base. With the money, the military would be able to build two new buildings and Sen. Lott was there for the ribbon cutting.

I remember wondering what the buildings were for. And none of the news accounts went into a lot of detail. I assumed it was something classified or at least for general consuption. Mississippi has communications units on the coast - and it doesn't take a lot for that kind of gear to be something that is kept out of press releases.

The truth was considerably less impressive. The military had no idea what to do with the buildings. They hadn't asked for them, they didn't want them - but an act of congress had ordered them to build two buildings. The order was duly followed and the base was going to relocate some operations there to keep the buildings occupied and maintained - but they had no pressing need for additional structures.

Nobody had asked them, and yet a few million dollars of taxpayer money went into unnecessary buildings. Then the base had to absorb the additional cost of maintaining them.

If the military had been asked, they would have told congress to spend the money somewhere else. In fact, the military would likely have told congress to close the base in question - or at least dramatically scale back its function.

But Sen. Lott and congress knew better.

60 Minutes tops this story in spades with a discussion of how the US/Mexican border is supposedly being secured with an "electronic fence." Some brainiac had the idea that existing security equipment could be mounted in the desert and still be reliable.

It can't.

They also had the great idea that border agents could carry laptops around and connect to border surveillance feeds via cellular modem.

But the border is 1,200 miles long and doesn't get good cellular reception. Not only that, but the border agents have pointed out that operating laptops in hot cars in the dusty desert ensures that equipment will fail and fail often.

The parallels to I/S development are classic. The vendor (Boeing) promised that it could meet the specifications and the timeline easily.

Then their assumptions about their system were proven horribly wrong. Then their initial 28 mile setup was delayed by years. Then the costs went up.

It's too much of a simplification to say that end user research would have solved the current mess, but when the man in charge of the project is saying they should have done iterations and refined it based on what they learned.

But the client wanted it fast - so they've lost billions and years.

Go figure.