Say Again? #6:
Decisions, Decisions …

Ever seen a controller stop off at 31-Flavors at the end of a shift? No? Ever wonder why? By the end of the day, evening, or night (as the case may be), that friendly voice on the ATC frequency has been making several decisions a minute for eight hours or so. The ATC decision-making process has a profound impact on those of us who fly within the system. Don Brown, the NATCA Safety Representative at the Atlanta Air Route Traffic Control Center, fills us in.


Since I’ve been writing this column I’ve been reading various and assorted publications about aviation just as I always have. I try to stay abreast of what is going on the the industry and how it might affect ATC. Writing this column has given me another perspective as I read.

I’ve always questioned whether most pilots understood why controllers have certain procedures that we follow. You’ve heard me and other controllers mention that controllers can train for up to five years before earning their ticket. Putting those two things together, I’ve never had a lot of faith that pilots ever would really understand what controllers are up to. But my new perspective has led me to believe that it goes even deeper than that. I don’t think pilots even understand the decision process that controllers use as we operate under our procedures.

Decision making is a very personal process (of course). You can’t really jump inside someone’s head and see all the thoughts that are processed to reach a decision. Therefore, most of what follows in my personal decision making process. But we controllers do have to work within the same procedures and rules and we do work in the same physical environment. So maybe this will be helpful to you in understanding the controllers that you work with.

Safe, Orderly and Efficient

The first thing to influence any decision I make regarding ATC is the phrase you’ve probably heard a dozen times: Safe, Orderly and Efficient. It sounds simple enough, but there is a depth in that phrase that few people take the time to explore. I thought it was important enough to explain to a senator one time, so I’m going to explain it to you. Whether you are talking about giving a shortcut to an airplane or purchasing a billion dollar piece of ATC hardware, it’s worth your time to work though that simple phrase.

I use it just like a flow chart. Is it Safe? Does it contribute to the safety of the operation? Does it make it safer? If, and only if, the answer is positive do you get to move on to “Orderly.” When you reach Orderly you ask the same questions. Is it orderly or will it create chaos? Will it make the situation too complex to handle safely? Again, only if the answer is positive do you get to proceed to “Efficient.” Even if I reach “Efficient,” I ask more questions. Efficient for whom? The system? One user or all the users? As I said, decision making is a personal process. My decisions are highly influenced by the fact that I’m a safety rep.

Slow Airplanes at Play

King Air 200Before this turns into something that sounds like “The Zen of ATC,” let’s get down to specifics. Some decisions are easy to make for controllers. A well-accepted axiom in the enroute environment is: “always turn the slow guy.” If both aircraft are enroute, it’s normally easier to get the slow guy behind the fast guy. Now that I think about it, I guess that applies to the approach environment too. If you’re vectoring for the ILS and have two aircraft tied at the turn-on point, it’s only logical to put the faster guy in front. You can turn the slow guy in tight behind the faster traffic and your spacing will increase as they fly toward the airport.

The thing to understand is that “slow” is relative. If you’re flying a BE20 in the lower flight levels, you’re probably one of the faster aircraft around. If you fly the same BE20 at FL310, you’re going to be the slow kid on the block. Fly the very same BE20 at FL310 with a 100 KT headwind and, for all practical purposes, you’re standing still. Please (before the Citation drivers panic and try to sell their airplanes), read the next paragraph before you start jumping to conclusions.

Altitude and Attitude

The primary method Center controllers use for separation is altitude. That’s because of the volume of airspace that we work. Our smaller sectors are the size of an entire Approach Control. We’ll work that sector with one person 90% of the time. Contrast that with 3 to 15 controllers working your typical Approach Control. Center controllers separate airplanes as soon as they can and keep them that way. Altitude is just the easiest way to do it. I’ve seen controllers who like to vector, trying to have a deal in three different States simultaneously. That is a lot of airspace to watch. After a lesson like that, it’s not hard to convince a controller to “use altitude.”

So whose altitude do we change? That depends on whether the situation requires a climb or a descent. A jet inbound ATL is crossing a jet going to all the way to California? Descend the one that’s going to ATL. If we need to climb and we think the jet going to LAX has burned off enough fuel, we’ll ask him if he’d like to climb. As you can see, there are many factors that can enter into the equation. Got a Cessna 150 and a Baron tied at 6,000? It’d take forever for the 150 to climb. We’ll climb the Baron.

Simply put, that’s all controllers do all day long every day: We make decisions. We make so many decisions that once we get off work, it’s hard to get us to make any more decisions for the day. According to my wife, it’s hard to get us to do anything for the rest of the day. But that’s another story.

Just Because

Controllers At WorkThe point being, controllers have to make so many decisions that we search for ways to simplify the process. Especially the arbitrary decisions. For example, you have two airplanes on top of each other inbound to ATL. Say one is at FL260 and one is at FL280. They’re the same company, the same type and they are the same speed. Which one goes first? If they have a headwind, the lower altitude usually has less wind and you might put him first. If they have a tailwind, you might run the higher one first. Do you vector or do you use speed control? Is there a dogleg in the arrival where you can shortcut one of them?

Do you see what I mean? A controller can decide to run his traffic a number of ways. There are a dozen decisions you can make with just two airplanes. It’s a rare day that we’re just running two airplanes. If you’re running arrivals into a hub like Atlanta, you’re working dozens of inbounds. And you have to decide where you are going to place each and every one of them.

In this case, I’d run the lower guy first. That’s my personal preference. I also prefer speed control over vectors when the traffic is inbound on the same course. Let’s say I have plenty of time to work on these two. I’ll assign the front one 300 KTS or greater and I’ll assign the back one 280 KTS. Pretty soon I’ll have my five MIT (miles in trail). Simple huh?

Five ! Ten ! Do I Hear Twenty?

As you may have guessed, it’s rarely quite that simple. We rarely have the luxury of time. If I need to get the 5 MIT in a hurry I’ll assign the back one 250 KTS or I’ll vector him. I’ve even run into situations where I’ve had to do both. For instance, you just got five of them spaced out 5 MIT and Flow Control calls and says go to 10 MIT. ARRRRGGGGGHHHH ! Now you’ve got to scramble. You’ll have to slow the back four to 250 KTS and vector all four of them. Don’t forget that you get to run all your enroute traffic at the same time.

I think right here would be a good place to start mentioning some numbers. Let’s pick a nice easy number like 10. If you’ve got 10 arrivals and you’re running 5 MIT, that’s 50 miles total. Assuming you’re perfect that is. Five miles is the standard separation at the Center. Let it go down to 4.9 miles and you’ve bought an operational error. Let’s make it 60 miles total for career enhancement. Now you’re going to 10 MIT. That’s a line of airplanes stretched out for 100-120 miles. If we have to run 20 MIT you’re now looking at 200-240 miles. Do you see how easy it is to be slowed to 250 KTS 200 miles from the airport?

Hurry Up And Slow Down

Keep in mind that, in effect, you’re trying to keep an airplane at FL330 with a 100 KT tailwind in trail of an aircraft that is down at 12,000 and 250 KTS. At some point, a controller starts running into the point of diminishing returns. Sooner or later it’s going to be easier (and safer) to go into holding instead of giving five airplanes an aerial tour of the great state of North Carolina.

While I have you held captive in the holding pattern, let me give you another sermon on the rules. You know that silly rule that says you’re supposed to state your altitude leaving when you’re assigned a new altitude? You knew it was in there for a reason, didn’t you?

AIM 5-3-3. Additional Reports

a. The following reports should be made to ATC or FSS facilities without a specific ATC request:

1. At all times.

(a) When vacating any previously assigned altitude or flight
level for a newly assigned altitude or flight level.

Now we plug in the controller side of the equation from the FAA 7110.65


b. Assign an altitude to an aircraft after the aircraft previously at that altitude has been issued a climb/descent clearance and is observed (valid Mode C), or reports leaving the altitude.

There you are in a holding pattern at 12,000 with a dozen other airplanes stacked above you and Approach starts running them again.

ZTL: “Airliner 123 cleared over the SHINE intersection to CLT via the SHINE5 arrival descend and maintain one one thousand.

AIR123: Airliner 123 cleared to CLT down to one one thousand.

ZTL: (sigh). Airliner 123 say altitude leaving.

AIR123: Uh, we’re out of twelve.

ZTL: Airliner 234 descend and maintain one two thousand.

AIR234: We’re descending to one two thousand Airliner234

ZTL: (big sigh) Airliner 234 say altitude leaving.

AIR234: Airliner 234 is leaving one three thousand for one two thousand.

ZTL: Roger, Airliner345 descend and maintain one three thousand.

Have you ever seen a radar scope with a full holding pattern? All those targets dragging those big data blocks over the same spot? Ever tried to observe someone’s Mode C when it’s like that? We’re lucky if we can find your data block, much less observe your Mode C. Getting the picture?

Radar? Who Needs Radar?

RadarAs strange as it may seem, there are times when non-radar procedures will actually allow us to move airplanes faster than radar. That must be the reason this is in the book.


c. Use nonradar separation in preference to radar separation when the situation dictates that an operational advantage will be gained.

A holding pattern situation isn’t the only place that rule is useful by the way. And in case it hasn’t struck you yet, you might want to move smartly when changing your altitude in a holding pattern. Not too fast and certainly not too slowly. As a matter of fact, you might want to stick to the book. Am I repeating myself again?

Smart Move

Speaking of moving smartly, it’s time I brought this up. I mentioned that Center controllers prefer using altitude separation. I also mentioned that we like to get airplanes separated quickly. A 1,000 FPM descent rate for a jet doesn’t fit into the plan. I guess no one asked a controller about that when they programmed the computer to fly the airplane.

Go back to the scenario where we’re running intrail to the hub. You’re toward the back of the pack and the controller says, “Airliner123 descend and maintain one two thousand.” You put down your doughnut, then you put down your coffee, and then you reach up and dial in 12,000. “George ” wakes up and says, “We’re way too far out and to maximize fuel economy (not to mention this wonderful tailwind we’re enjoying) I’m only coming down at 1,000 FPM.” Twenty miles later the controller looks up and sees that Airliner123 has only come down 500 feet. Said controller is up to his eyeballs in airplanes and just loves it when somebody simplifies his decision making process. “Airliner123 turn thirty degrees left vectors intrail.” How’s the fuel economy looking now George? (note: “George” is autopilot.)

It’s ATC101. Controllers can separate airplanes vertically or laterally. If you’re transitioning through a bunch of altitudes, the longer you take to do it, the longer you’re exposing yourself to conflicting traffic. If we can’t get you to descend (or climb), the only other option is to turn you. When we’re running intrail, sometimes we want to get you out of that tailwind. Believe me, controllers have been lectured over and over about fuel consumption at the lower altitudes. Nobody has to lecture us about how much fuel you’ll burn in the holding pattern if we can’t get the intrail we need.

Human Relations

There’s one more issue regarding decisions that I want to talk about before I wrap this up: Controllers are human. That means that we have many of the same attributes that everyone else you know has, both good and bad. We can have good days, bad days and in-between days. We can be kind and we can be mean. We can be short and we can be patient. Well, okay, we’re not a very patient lot but you get the general idea.

Controllers don’t just make decisions all day, we make instantaneous decisions all day. Many of those decisions are made about you, instantly. Can I trust this pilot? Can he handle a complicated clearance? We make those decisions on the way you sound, the way you fly and the way you file your flight plan. Sure, the local guys develop long-term reputations. So do the companies and airlines. But even the guy who’s never been in my airspace before starts making a reputation the second he enters the system.


controllerYou shouldn’t stay awake at night worrying too much about what controllers think about you, but you shouldn’t ignore it either. I’m a stickler about good phraseology. The next controller down the line probably won’t notice. The trick is, if a controller has to make a decision about your flight, you’d rather have him in a positive frame of mind concerning your flight. Making the same request for a shortcut, day after day, at the same spot won’t endear you to controllers. Having every pilot in your fleet make the same request, at the same spot, every day, certainly won’t either.

Like I said, I wouldn’t stay awake worrying about it. On the other hand, I don’t know anybody that dislikes it if you use standard phraseology and file the STAR. Who knows? If a controller perceives you as a conscientious, professional pilot that he can trust, he might just put you first the next time he has to make a decision.

Have a safe flight!

Don Brown
Facility Safety Representative
National Air Traffic Controllers Association
Atlanta ARTCC