$213 million spent.
“Until the problems appeared this month, the security upgrade appeared, based on public reports, to be a rare example of a project being completed on time and within its budget in an agency plagued with project management problems. It was not until Oct. 17, according to the NNSA memo, that the lab formally informed NNSA of the project’s problems.“
That’s the story for the new security system at Los Alamos National Laboratory’s most important nuclear weapons manufacturing site.
Last month it looked like it was on time and within budget, but suddenly it’s not?
How does that happen?
Reading the story, I’m going to propose a single root cause of the failure here. I know nothing about this project other than what’s been reported, so take my analysis with a large grain of salt.
“The NNSA, the federal agency that oversees Los Alamos, is dispatching a team of accountants next week to review the project as part of an effort to hold the lab “accountable for poor project management,” according to the memo.”
Oh good, let’s hold them accountable now that it’s really too late.
And instead of sending in some veteran Construction Project Managers to review what went wrong, let’s send in a team of bean counters.
(Hey, I like bean counters. But let’s send a couple of those folks along with some people who can actually assess the management of the effort itself.)
Plan, Do, Check, Act
Yep, it’s the good ol’ Deming cycle.
Quality isn’t something you assess at the end of a project.
“The faulty work was done by a lab subcontractor in 2010, according to Roark, but was only recently noticed during final ‘commissioning’.”It’s part of the process (or at least it should be).
So it’s been 2 years or more since the faulty work was done.
Perhaps some administrative manager was the one who signed off when that work was completed, without having any clue about whether or not it was done right.
Perhaps no one signed off on anything.
Either way, it sounds to me like this is what they did:
Plan, Do, Cross Your Fingers
Does that sound like a good project management methodology to you?
I didn’t think so.
Verification and Validation is so important.
And so important to do early and often.
And this is why, folks.
End-to-end testing of the fiber optic installation was possible back in 2010. Likely, there was the possibility of an iterative deployment approach with verification and validation built in to each iteration.
Furthermore, the fact that the ‘official’ status of this project was on time and within budget prior to this month is a big red flag that either:
- Little to no real verification and validation is occurring
- Verification and validation is all being deferred until the last possible moment for a big monolithic milestone that so many government agencies seem so fond of, and/or
- There is a huge communication problem going on where problems can be covered up for long periods of time.