Friday, October 21, 2011

Shuffle the Deck Again


FAA’s Air Traffic Chief Pledges Execution of Capital Programs

”The largest commitment involves the $2.1 billion En Route Automation Modernization (Eram) program to replace the computer system connecting 20 FAA air route traffic control centers. According to the DOT IG, the program is running four years late and hundreds of millions of dollars over budget.

“This isn’t necessarily a failure of money. This is a failure of management,” U.S. Rep. John Mica (R-Fla.), chairman of the House Transportation Committee, said during a hearing on October 5. Grizzle said developers must introduce 200,000 additional lines of software code to the system and complete 18 more en route centers before anyone can consider Eram operationally ready.”

Here’s a crazy idea. Take the last 100 computer programmers Lockheed has hired -- the kids just out of college -- send them out to academy in Oklahoma City and try to make them controllers. Those that make it through, send them to the various Centers and try to make them controllers. Let them work traffic for 2 years and them give them back to Lockheed. They’ll be there in time to work on ERAM.

Or you could reverse that idea -- like they did last time. Take 100 controllers and let Lockheed make them into programmers.

You could do both at the same time. Well, except for the fact that the FAA is short of controllers. (That’s what happens when you let ideologues run your bureaucracy.) Let me restate it. The U.S. Government could afford to do both -- at the same time. And it would be a pittance compared to what it will wind up costing us in the end.

Meanwhile, enjoy watching history repeat itself. No matter how many times they shuffle the deck, they’re still playing with the same cards.

Don Brown
October 21, 2011


Bob said...

Well put. They did hire controllers to develop NAS stage A, and it was a major achievement. Now we have Lock Mart running Flight Service and ERAM and doing a grand job on both.

radaray58 said...

Great idea brother! Too late for the current implementation, but they would have a solid cadre of programmers to improve the system over time. Ray Spickler, ZKC

Air Traffic Mike said...

“This isn’t necessarily a failure of money. This is a failure of management."

What Congressman Mica fails to mention is that is a failure of the Bush Administration's management.

They didn't NEED our input.

Here's what the government needs to do to get LockMart to get it right. Don't pay them one more dime. LockMart has gone over budget in both time and money. Tell them they need to deliver on this contract or they will be held in breach. If they do breach the contract make the penalty harsh by holding up their other contracts and banning them from any future contracts.

LockMart understands their "purse strings" more than anything else.

If someone went four years and umpty squat dollars over contract in the vaunted "private sector" it would become a monumental lawsuit by now.

XPATCO said...

once again the FAA tries to cookie cut something that has to fit into unique airspace of 20 centers. Without controller input ERAM will never be adaptable for service in the NAS.

Jason Lindquist said...

Speaking from the other side of the fence as a software developer, communicating requirements is one of this industry's big hurdles. The customer doesn't explain them well, or the developers don't ask the right questions to flesh them out, or management doesn't want too many iterations of that process because they want to get moving towards those progress checkpoints that open the money spigot. For something this critical, from both safety and economic perspectives, you've got exactly the right idea. Get developers that can think like controllers, by training them as controllers.

But here's a twist on that idea to make it more effective: You don't want kids fresh out of college. You want people with 3-5 years in the field. We learn nothing in college about collaborative work, project management, assembling requirements, defect tracking, testing, documentation, or customer deployment. If you get us straight out of school, we can't fully appreciate why you're making us controllers, we won't be alert for the subtleties of the job that we'll need when we move back from the scope to the dev PC.

If you would've caught me on my way out of Qualcomm in 2001, I would've jumped at it, and I would have been the right guy for the task. If you'd caught me five years earlier, just out of Illinois' EE department, I would not have been ready--but I would have thought I was anyways!

But it's all academic, isn't it? God forbid legislators or large defense contractors should EVER be that forward-thinking! :)