Software is…well it’s a funny thing. In general, we’re still not very good at it. Don’t get me wrong, we’ve been able to do some amazing things. But the field is still very much in it’s infancy. And bugs, cost/schedule overruns, failed projects, etc are more the rule than the exception.
But even with all the examples out there, it’s still entertaining to read about some of the more colossal clusterfucks (that’s a technical term some of you might not understand). For example, the FBI’s $170 million failed system…
It was late 2003, and a contractor, Science Applications International Corp. (SAIC), had spent months writing 730,000 lines of computer code for the Virtual Case File (VCF), a networked system for tracking criminal cases that was designed to replace the bureau’s antiquated paper files and, finally, shove J. Edgar Hoover’s FBI into the 21st century.
730K sources lines of code (SLOC)??? Wow. For essentially a digital file cabinet?? For comparisons sake, I work for a similiar company, and in particular one of the system I work on is an air traffic control system (if you’ve stepped on a plane in the western hemisphere you’ve been controlled by it). How many SLOC??? Roughly 450K. Seems a little off, but who knows. Perhaps the system is more complex than I think it is.
Nevertheless, the system was a colossal failure. The first sign of trouble…
Software problem reports, or SPRs, numbered in the hundreds, [Zalmai Azmi, the FBI's top technology officer] recalled in an interview. The problems were multiplying as engineers continued to run tests. Scores of basic functions had yet to be analyzed.
“A month before delivery, you don’t have SPRs,” Azmi said. “You’re making things pretty. . . . You’re changing colors.”
Never in one single statement has a one person demonstrated so completely that they themselves were probably a big part of the problem. It’s comforting to see that the person in charge of technology at the FBI appears to be completely oblivious to how technology is developed.
Not that this is a surprise. Anyone that has worked on a contract for the government can give you countless tales where they obviously have no clue what they are doing. It’s an epidemic within government agencies. They want you to build them the moon, made of gold, and populated by Ooompa Loompas living in gumdrop houses. At first it’s kinda cute. Then you realize that is your money they are wasting and you get depressed. Then you move past that and you’re just annoyed at how they can ask for something when they have no idea what that something is.
So here is a lesson I’ve managed to learn in my brief time working with government agencies. Managing customer expectations is key. They want you to give them the world on a silver platter. You have to make sure they understand they can’t get that, but here are some other options that might better suit their needs for varying prices.
Be ready to give a recommendation on which one you would pick, because they WILL ask. Remember they haven’t a clue what they are doing. The temptation to just point to the most expensive (and profitable) one here will be enormous. But remember, if they are happy with the end result, they come back for more.
This project didn’t fail because of inept leadership at the highest levels of the FBI. It failed because the contractor failed to explain to that same inept management in the FBI exactly what they were asking for and why it wasn’t going to happen.
The contractor got the government signed up for something they knew they couldn’t deliver. As a result they pushed blindly forward with little thought to a quality robust design. Basic functions were left out of the design because the contractor did not correctly analyze what the customer (the FBI) needed. The contractors claims that the government kept changing it’s mind is ridiculous. Of course they did, they don’t know what they want because you haven’t helped them figure that out yet. Had you done so they still may have occasionally done so but you would have been able to anticipate these and marked them as risk factors and planned accordingly.
Now don’t get me wrong, the FBI was at fault here as well. They remained blissfully unaware of how immensely they were wasting money and getting nothing for it. But they are the government….that’s their job. In this case the fault goes to the contractor, for failing to fulfill one of the most basic duties of any government contractor.
But heh, it’s not the end of the world, and the government has bottomless coffers of your money. So they’re gonna give it another try.
Last year, FBI officials announced a replacement for VCF, named Sentinel, that is projected to cost $425 million and will not be fully operational until 2009. A temporary overlay version of the software, however, is planned for launch next year.
The project’s main contractor, Lockheed Martin Corp., will be paid $305 million and will be required to meet benchmarks as the project proceeds. FBI officials say Sentinel has survived three review sessions and is on budget and on schedule.
SAIC is not involved. FBI officials say they are awaiting an audit by a federal contracting agency before deciding whether to attempt to recoup costs from the company.
Ahhhh, the circle of life in action…