Say Again? #16:
Flight Data Processing
We know what you think: The AIM is boring and needlessly pedantic, the 7110.65 used by controllers is equally complex, and neither reflects the
I've been writing for AVweb just over a year now. It's amazing the things I've been learning. One of the things that I've learned is that new people start reading this column all the time. And they don't necessarily go back to where I started and "catch up," as it were.
So, at the risk of repeating myself, let's go over a couple of things. First, I'm just a plain-old everyday controller. The only thing that separates me from other controllers is that I'm the NATCA safety representative for a very large facility, which means I spend an unhealthy amount of time reading, thinking, and talking about air traffic control. Second, I'm really uncomfortable with the term "expert" when people try to apply it to me. The only thing I'm an expert at is working the seven sectors I'm assigned to work in my Area at Atlanta Center. I only earn that by default. It's nothing more than the other 50+ controllers I work with can claim.
What I Learned This Summer
In case you're wondering what this literary equivalent of staring at my shoes and saying "Aw, shucks" is all about, it's this: In writing a previous article for AVweb, "Say Again? #2: The GPS Mess," I started browsing through the AIM on a regular basis. To be specific, I started trying to understand the following section. I'm not an expert on the AIM, but I can read:
b. Airway and Jet Routes Depiction on Flight Plan
1. It is vitally important that the route of flight be accurately and completely described in the flight plan. To simplify definition of the proposed route, and to facilitate ATC, pilots are requested to file via airways or jet routes established for use at the altitude or flight level planned.
Sometimes, I even like what I read. When I do, it encourages me to read more:
5. When filing IFR, it is to the pilot's advantage to file a preferred route.
I really liked that one, so I kept reading:
c. Direct Flights
1. All or any portions of the route which will not be flown on the radials or courses of established airways or routes, such as direct route flights, must be defined by indicating the radio fixes over which the flight will pass.
What'd He Say, Ma?
I realize that not everyone has 20 years of reading "governmentese" under their belt, so let me point out the controlling phrase: "must be." That doesn't mean "We'd sincerely appreciate it if you complied." It doesn't mean "maybe." It means your route of flight must be defined by radio fixes. An airport is not a radio fix. So, can anyone explain to me why I see hundreds of flight plans filed from an airport direct to an airport, without any radio fixes described, every day ?
Yes, that is a rhetorical question. I know why. I also know that it doesn't comply with the "orderly" part of my Safe, Orderly and Expeditious mantra. The first problem we run into is that the current flight data processing software we have doesn't recognize every airport in the country. I've already heard all the tirades about that limitation (I've participated in more than a few myself), so let's get over it and deal with the reality of the situation. In order to work around that limitation, many people have taken to inserting latitude/longitude coordinates in their flight plans.
Now when I say "many people," I mean just that. You'll notice that I didn't say just "many pilots." That's because pilots, FSS specialists and other controllers are all doing it. I'll even admit to doing it (in the ignorance of my youth, of course). It's quite understandable why so many people have been and are doing it. It's also wrong.
There's only one place in this section of the AIM where I can find mention of the use of lat/longs:
5-1-7.d Area Navigation (RNAV)
4. Pilots of aircraft equipped with latitude/longitude coordinate navigation capability, independent of VOR/TACAN references, may file for random RNAV routes at and above FL 390 within the conterminous U.S. using the following procedures.
Again, let me point out the controlling phrases (i.e., the phrases you'd just as soon not know about): "may file," "at and above FL390," and "using the following procedures." So I won't be accused of rubbing it in, I'll not detail the "following procedures" for you. Most of you don't operate in those altitudes. For those of you who do, you need to read them. Because very, very few of you are following them.
Don Quixote Rides Again
In case you're wondering why I've decided to tilt at this windmill, it's because there are occasions I'm reminded why we have rules. Invariably, those occasions are unpleasant. For example: I recently relayed a clearance to an aircraft departing Hickory, N.C., direct to New Bedford, Mass. There was, of course, a lat/long in the flight plan to "make" the flight data processing work. The aircraft was cleared to depart Runway 06, left turn, and "on course." I was expecting the aircraft to make about a 10- to 40-degree left turn to New Bedford. The aircraft came off the ground and made a left 230-degree turn to someplace in Florida.
This is not the first time something like this has happened to me. Being the safety rep., I can tell you it's happened to numerous other controllers too. More than one mistake occurred in this example that allowed the error to happen. That's almost universally true anytime there's an incident or accident in ATC. But for some reason, while I was reviewing this particular incident, the light bulb came on and I finally found the words to describe what is really wrong with this bastardized process we're all using.
No Hits, No Runs, No Errors
Controllers are very good at not making errors. It's drilled into us from Day One that we don't get to make any, and there aren't any acceptable excuses if we do. We're human though, which means that we actually do make mistakes. The beauty of the ATC system is not only that it minimizes the chances of making a mistake, but if all the procedures are followed, mistakes quickly become glaringly obvious.
In this particular incident, I obviously made a mistake in looking up the airport identifier. OK, twenty lashes for me. But what would have prevented that mistake from becoming an error?
5-1-7.d Area Navigation (RNAV)
3. Pilots of aircraft equipped with operational area navigation equipment may file for random RNAV routes throughout the National Airspace System, where radar monitoring by ATC is available, in accordance with the following procedures.
(e) Define the random route by waypoints. File route description waypoints by using degree-distance fixes based on navigational aids which are appropriate for the altitude stratum.
If the pilot had complied with the AIM and filed over CAE, AGS, SAV, CRG or any of the hundreds of "radio fixes" that I recognize, my mistake would have become painfully obvious before it became an error. Even if he had filed using "degree-distance fixes," I would have found myself looking at a strip that said he was going from Hickory, N.C., to New Bedford, Mass., via CAE079012..CRG011037 ... what the ...? Why would anyone going to Massachusetts turn due south to Columbia, S.C., and then Jacksonville, Fla.? Maybe I need to put on my glasses and look up that identifier again.
See what I mean about being painfully obvious?
About now, virtually every controller out there is praying I'll stop going down this path. The Center controllers out there see a thousand flight plans in their future that look like this:
... and on and on. The Tower guys are just as fearful. Imagine a full-route clearance with a readback for one of those. Yuck.
Anybody got an alternative? An alternative that works? If you do, be sure to send it to the FAA. I'm sure they'd be glad to hear about it. As soon as you can prove it works, adapt all the databases, and get it published in the AIM, the 7110.65, etc., I'll be sure to use it. In the meantime, I'll work with what I've got.
All this leaves me to deal with quite a quandary. Pilots aren't going to like this, controllers aren't going to like it, so why do it? As I told a friend one time, "What's a safety rep. to do? Tell you not to follow the rules?" That isn't acceptable. Especially when the rules have been proven to work. I recently read a statistic that said ATC was 99.99997 percent error-free. It always irks me when I hear pilots say they want to "beat the system." Is that the system they want to beat? How many of those "9"s do they want to beat? Yeah, that's a rhetorical question too.
Do the DU
I can just hear someone asking, "Well what about DUATS?" What about it? I've never used it. I see it being used everyday. I see the lat/longs the program puts into a flight plan every time the route crosses a Center border. I can't explain the apparent contradiction with the guidance found in the AIM.
Oh OK, actually I could. I have asked around about it. I'm personally satisfied that the story I've heard is true. It fits with what I know. But I'm not a reporter, and I'm not going to pretend I'm one either. So instead of trying to verify a story and finding an independent source and all that other cool stuff you see reporters do on TV, I filed a UCR about lat/longs in flight plans.
A UCR is an Unsatisfactory Condition Report (FAA Order 1800.6A). Most controllers call it an Unsafe Condition Report. That misnomer tells you pretty much all you need to know about the program. Controllers have to feel a situation is unsafe before they'll bother filling out a form, which is rare enough that they don't know the proper name of the form. Oh well. You'll probably never see a movie made about safety reps. It's not exciting enough. But that's what we do. Fill out forms that no one else wants to bother with, about issues most people wish we'd leave alone.
Decisions, decisions. Which way should we go now? Let's see. If you remember "The GPS Mess," you'll remember my example about the aircraft riding a sector boundary and all the problems it creates. If you file via the airways, that problem is eliminated. If you file radial/fix distances, we may still be stuck with that problem. Here's another problem it leads to that you don't know about.
We occasionally get a data-processing error when an aircraft transitions two or three facilities in a short period of time. As an example, twice last week I had an aircraft that went from Greensboro (GSO) Approach airspace, to Roanoke (ROA) Approach airspace, back into GSO, then into Atlanta Center (ZTL) airspace, and finally into Indy Center (ZID) airspace. All within about 30 miles.
GSO is a "split" facility in regards to flight data processing. Half of their airspace lies under Washington Center (ZDC) airspace and half lies under Atlanta Center (ZTL) airspace. In that the HOST computer at each Center does that flight data processing, the flight plan went ZDC-ZTL-ZDC-ZTL-ZID. One of the "logic checks" our software runs is to check for duplicate flight plans. There can't be two "active" flight plans with the same callsign.
Now, I can't explain all the technical details behind that, but in layman's terms, this is what happens. The ZDC computer calls up the ZTL computer and says, "Here's a flight plan on N12345." ZTL's computer takes it and processes it. In doing so, it sees that the flight is going back into ZDC's airspace. So ZTL's computer calls up ZDC's computer and says, "Here's a flight plan on N12345." ZDC's computer says, "I've already got one of those, leave me alone and go buy a clue." There's more to it than that, but that's about the limit of my knowledge on the HOST programming, and I'm boring the half of my readership that doesn't really care.
What it means to that one pilot is absolutely nothing. He'll never know. What it means to me, the controller, is I have to get another controller to come help me fix the data processing on that one flight. All the other flights in my sector are going to suffer while we figure out what went wrong, and try to get ZID some flight data on that one flight that won't process. Management types would call that a poor allocation of resources. Controllers call it one guy trying to cut five miles off his route and sinking the entire sector.
Easy to Be Hard
Now, all you software engineers, computer gurus and techno-geeks who are mentally composing emails to tell me how simple all this is, just stop what you're doing. Unless your name is Steve Jobs, I don't want to hear about it. It isn't easy. The biggest — the very biggest — names in the computer industry have been working on all this for years. If you think you're smarter than them, send your ideas and job application to them.
Billions have been spent upgrading the ATC system. We're finally achieving a measure of success. Not because we've gotten smarter or because technology is so advanced, but because we've learned from our mistakes. Instead of taking the "Big Bang" approach to upgrades and blowing a billion bucks, we're now in the "build-a-little/test-a-little" mode. That's not very exciting either but it is working.
I think I've drifted far enough off course. Let's see if I can't get back on the airway.
A Fine Fix We're In
Once again, from the AIM:
5-1-7.c Direct Flights:
1. ... Fixes selected to define the route shall be those over which the position of the aircraft can be accurately determined ...
In case I haven't made it plain, lat/longs don't qualify in my book. I know right where the SPA360021 fix is. I don't have any idea where 3333/8400 is located. From the AIM:
(c) Plan the random route portion of the flight plan to begin and end over appropriate arrival and departure transition fixes or appropriate navigation aids for the altitude stratum within which the flight will be conducted. The use of normal preferred departure and arrival routes (DP/STAR), where established, is recommended.
We can debate about what are "appropriate arrival and departure transition fixes" if you like, but I think this makes it a little clearer:
(c) Plan the random route portion of the flight to begin and end over published departure/arrival transition fixes or appropriate navigation aids for airports without published transition procedures.
As I said way back at the beginning of this article, I'm not an expert on the AIM. I don't know why there's a slight difference between those two paragraphs. But instead of remarking about how stupid it seems, I'm going to tell you that there's probably a very good reason for the difference. That's the lesson I've learned about this business. Just when I start thinking the other guy is stupid, it turns out that I am.
If you read last month's column on airspace stratums, it might dawn on you why pilots should file "transition fixes or appropriate navigation aids for the altitude stratum within which the flight will be conducted." Did you notice that it said "departure/arrival transition fixes"? Here's another tidbit about flight data processing. The data processing will send a flight progress strip to every appropriate sector as the aircraft climbs through all the stratums on departure, and it will generate strips to all the stratums as the aircraft descends to its destination.
When to Hold 'Em
What happens every second of every day is that as soon as a climbing aircraft checks on a new frequency, they ask for direct "somewhere down the road." All too often, controllers are willing to accommodate and clear them direct. The controller amends the flight plan in the computer, and a "Remove Strips" message is generated to every stratum between the aircraft's current altitude and its filed altitude.
That's probably a little tough for some pilots to follow, so let me give you an example. You're filed from CLT to ORD on an appropriate route requesting FL350. CLT approach switches you to ZTL, and you check in leaving 11,000 for 14,000. ZTL climbs you to FL230, and leaving 14,000, we turn you to a heading of 340. You're out of 15,000 and ask for direct OKK. The controller clears you direct OKK and puts the amendment into the computer. As soon as the controller enters that amendment, the computer assumes you are level at FL350 and starts processing your flight data accordingly. Every sector below FL350 receives a Remove Strips message and the computer will not forward the new routing to those sectors.
Assuming the controller who entered the amendment follows the rules (whoops, too late), he now has to call the next sector above him and manually coordinate the amendment. He has also committed that next controller to call the sector above him and coordinate the route amendment. In other words, he's making work for the next controller by operating contrary to our procedures. We have names for controllers that habitually "make work" for other controllers. None of them are printable.
You'll need to think about all the implications behind this software "feature," but to get you started, I'll give you a hint. The system isn't designed to reroute airplanes that are transitioning (climbing or descending) through a lot of altitudes. If I haven't convinced you in all my previous articles that it's counterproductive to ask every controller you talk to for "direct somewhere," maybe I can convince you to limit yourself to the times that you are in level flight.
In the process of writing this article, I had two strips show up on the sector at the same time that epitomize the larger point that I'm trying to make in this article. One strip was on a King Air that had entered ZTL's airspace on the west side of the facility and had flown all the way across. The other was on a Seneca that had done basically the same thing. The King Air was supposed to be on a preferential route into INT (Winston-Salem, N.C.), but had been short-cutted two or three times. The Seneca, who could have filed direct, chose to file via the airways, even though he was equipped for advanced navigation.
The King Air's flight plan had been amended six times. The Seneca's flight plan hadn't been amended once, even though he flew just north of the CLT hub. The flight data processing on the King Air had generated 41 flight progress strips. The Seneca's flight plan? Six strips. And the kicker? The King Air's last strip, after being turned for traffic and rerouted again, showed the BROOK intersection direct INT. The BROOK STAR is the preferential arrival for INT. Six amendments just to wind up where he should have been to start with.
There are many, many more items I'd like to discuss, but I need to wrap this up. I hope that the few items I've been able to include will give you a few more pieces to plug into the puzzle we call the National Airspace System. It really is an extraordinary system, although it can be maddening trying to figure out exactly where you fit into it.
Speaking of maddening: Yes, I realize how unpopular most of the article will be (just in case you are wondering). I spent the better part of an hour arguing with a controller friend (who made the mistake of asking what this month's column was about) on whether I should write all this or not. He's one of our best and brightest. You may be too. But this system wasn't designed for just the best and the brightest. It was designed to accommodate everyone. That includes the junior controller and the novice pilot.
I hope you'll keep that in mind the next time someone asks for your advice about how to fly, or (if you're a controller) how to work airplanes. I'm all for passing along the wit and wisdom that experience has taught you. But do me a favor. Pull out the book and show them how your experience builds upon the foundation laid out in the AIM and the 7110.65.
If your particular knowledge isn't in the book, maybe it should be. There's a process for adding it in the AIM and in the 7110.65. If you're not eager to share your knowledge with everyone, think twice before you share it anyone.
Have a safe flight!
Facility Safety Representative
National Air Traffic Controllers Association