The FAA Follies

Hey, FAA-watchers… remember this?

September 22 – Passenger delays are being reduced thanks to a program that increases the efficient use of runway capacity at major airports.

Airlines are reporting major savings in time – and money – because of Traffic Management Advisor (TMA), which analyzes traffic approaching an airport hundreds of miles away and calculates scheduled arrival times to maximize arrival capacity.

…

TMA is also expected to be up and running at John F. Kennedy and LaGuardia airports by the second and third quarters of fiscal 2009.

But TMA did reach a major milestone recently. New York, Indianapolis and Kansas City Centers declared “planned capability achieved” on July 25, Aug. 13 and Sept. 10 respectively. That means all 20 en route centers have reached this milestone.

This milestone marks the point in time when TMA has been used and checked out with all configurations of an airport. On average it took about one year for this to occur from when a center initially began using TMA on the floor.

That’s from a news item on the FAA’s web site talking about TMA. TMA is a pretty neat tool; for all the complaining we do about the agency around here, the reality is that the FAA does get stuff right from time to time. And TMA is said to be saving the airlines tens of millions of dollars a year, maybe even hundreds of millions; one estimate says it saves over $20 million at LaGuardia alone.

At Seattle Center (ZSE), we’ve done time-based metering for years and years. Many centers used to only do “in-trail” metering restrictions to airports, but in theory and in practice once you get the hang of doing time-based metering, it works out quite well.

(For anyone new, the point of time-based metering is that an airport can accept a certain number of airplanes in a given time period. We usually use an “arrival rate” by the hour. If an airport can take 30 landing aircraft per hour, we say “the rate is 30″ and the ATC enroute center needs to feed the approach control one airplane every two minutes.)

TMA is a whole new software tool that has some added advantages over the older version of time-based metering that we used to do. Perhaps the biggest is that it can integrate flows from multiple centers into a given approach control (TRACON); this means that metering can work much better in some of those congested airspaces back east.

TMA is also a technology that we’re counting on for the NextGen project. From a NextGen fact sheet:

A new software tool called Traffic Management Advisor (TMA) helps controllers sequence aircraft through high altitude airspace and into the airspace around major airports by calculating their precise routes as well as the minimum safe distances between aircraft. TMA is deployed at all 20 of the nation’s en route centers in the continental U.S. and 33 of the top 35 airports.

There’s an interesting web page about TMA on NASA’s site. It’s pretty technical, but anyone curious enough might want to have a look at it.

There is one little problem with TMA, though. When the 20 enroute control centers get their new software package, ERAM, up and running… TMA won’t work.

Apparently the TMA package was only built to interface with the old HOST computer system that we currently use in the control centers. The new system, ERAM, doesn’t interface with TMA.

The FAA spent tons of money over a decade developing TMA. They installed it and got it up and running at all the centers. It’s a big part of ensuring that the flows into airports are as efficient as possible.

And the new system doesn’t support it.

Apparently, a recent emergency telcon was held where the ERAM program office (which is sounding more and more dysfunctional, with a bunch of people who all know perfectly well how screwed up the ERAM program is, but they’re all afraid to tell the big boss that it’s not working out) told the Traffic Management Units that they’re not going to be able to use TMA anymore, sorry. Oh, and they have three hours to come up with their backup plan.

Of course the TMU guys basically told the ERAM program office people to blow it out their ear and that this was completely unacceptable. Time for another emergency telcon! Now the whole thing is in a big tizzy, TMU is furious, the ERAM office looks like a bunch of clueless jerks, and the ERAM program is almost certainly going to be delayed… again.

Remember, ERAM was going to be up and running in April of 2009. That’s what the FAA’s PR machine got Aviation Week to publish, back in March.

Here on the Follies we reported that was a lie, and pointed out that it was BS and that ERAM had already slipped. Well, it’s sliding more, and more, and more.

The reality of ERAM is this: The NEXT big software drop, with even more New and Improved bug fixes (really! They’re fixed this time!) is said to be delayed. Even when it comes to Salt Lake City Center and Seattle Center, everyone expects that it will be like the last who-knows-how-many software drops we’ve had; most fix 3 bugs, but 2 new ones are found (and they’re often absolutely terrible and must-fix types of bugs).

Here at Seattle, we’re rapidly approaching the time for yet another round of ERAM refresher training. We’ve spent great gobs of money on overtime for this type of training, even though everyone knows it’s unnecessary because ERAM will be delayed again. We’re also rapidly approaching the time frame where we’re not going to be able to begin to implement ERAM, because the Winter Olympics are happening in Vancouver, British Columbia; our center feeds Vancouver Approach and we cannot take the chance of using some new system on the large amounts of live traffic that are expected for the event.

What’s more, since ERAM doesn’t work with TMA at all, either TMA has to be redone to integrate ERAM (which is probably a bad idea since ERAM keeps getting updated and changed and “fixed” so it’s a moving target) or ERAM has to have a TMA interface integrated into it.

Either solution is going to cost a bunch of money and need to be tested.

Now, I don’t know who the rocket scientist was that allowed the ERAM contract to be done up without requiring that ERAM actually, you know, work with the other software tools that we use in ATC… but since it wasn’t in the thing in the first place, to make ERAM “backwards compatible” will no doubt cost the FAA a bunch more money and time.

I hate to be a bragger… but I would like to refer you to one last post that appeared here on the Follies. It was on April 18, 2008; over a year and a half ago, when I said this:

Remember this post a year or three from now, folks, when we are reading about either delays or about how the ERAM program isn’t working out right and wasn’t ready to be deployed.

Told.

You.

So.