Air France Report: More Questions Than Answers An initial report from the investigation into what brought down Air France flight 447 shows that the plane stalled before rapidly plunging more than 30,000 feet into the ocean. Now, some industry leaders are raising concerns about pilot training and the growing role of automation in the cockpit.
NPR logo

Air France Report: More Questions Than Answers

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript
Air France Report: More Questions Than Answers

Air France Report: More Questions Than Answers

Air France Report: More Questions Than Answers

  • Download
  • <iframe src="" width="100%" height="290" frameborder="0" scrolling="no" title="NPR embedded audio player">
  • Transcript

An initial report from the investigation into what brought down Air France flight 447 shows that the plane stalled before rapidly plunging more than 30,000 feet into the ocean. Now, some industry leaders are raising concerns about pilot training and the growing role of automation in the cockpit.


John Cox, president and CEO, Safety Operating Systems


For three and a half minutes, as Air France Flight 447 plummeted toward the Atlantic Ocean in a stall, pilots continued to pull the nose of the jet up -exactly the opposite of what they were taught to do from day one in a cockpit. All 228 aboard the flight that left Rio de Janeiro for Paris died in the crash.

Now industry leaders and pilots raise questions about the way pilots are trained and about automation. So pilots, how do you train for the unexpected? What questions do you have arising out of the crash of Air France Flight 447? Give us a call: 800-989-8255. Email us:

You can also join the conversation on our website. That's at, click on TALK OF THE NATION. John Cox joins us now from member station WUSF in Tampa. He's a flight safety consultant and a former commercial pilot. And nice to have you with us today.

Mr. JOHN COX (President and CEO, Safety Operating Systems): Good to be here, Neal.

CONAN: And I've read that every pilot is trained - if you go into a stall, put the nose down, pick up speed.

Mr. COX: It's in the primary flight training. You learn stall recovery very early. And as you progress through the training regimes, that's reinforced numerous times.

CONAN: So why didn't it happen above the Atlantic?

Mr. COX: It's a remarkably good question. One of the things that - with the Air France accident is, this is still very early. We've just gotten the data. The analysis is far from complete. The cockpit voice recorder has not been released, so there's a lot we don't know that may answer that question.

But unfortunately, we have seen on occasion pilots not respond properly. That's part of what the Colgan tragedy in Buffalo, New York, that occurred there. And we've seen some other ones, a West Caribbean flight some years ago. The stall training, upset recovery training that pilots go through, can be improved, and I think we should improve it.

CONAN: And it's important to recognize that these pilots are very experienced pilots on a top airline. They were not getting just one alarm or one signal.

Mr. COX: No. The conditions of the Air France flight were confusing. There were a lot of problems that were occurring simultaneously. They had a lot of problem solving to do. The airplane was getting less and less stable over time. It was rolling from one side to another.

So there was a lot going on, but they were not successful in correcting the issue and regaining control of the airplane. And that's where I think we need to focus training for the future.

CONAN: There is also, some say, an issue regarding automation. This is -obviously all new aircraft are highly automated. Airbus, in particular, manufactures planes with particular reliance on automated systems. And they went out of the automatic pilot once they got the alarm, and should they have done that?

Mr. COX: Well, the autopilot switched off. They didn't have a choice about that. The - due to a problem with the airspeed system, there was a problem with - that caused the autopilot and autothrust system to switch off. So the pilots were put into a position of manually flying the airplane now.

I flew the Airbus for six years. I enjoyed the time with it. Yes, it's automated. And the skill sets that you need in a conventional airplane, you also need in a - an automated airplane. So are there some differences? Yes. But once you're manually flying the airplane, it pretty much flies like a normal airplane.

CONAN: So don't you go through, when you're being checked out, when - to fly any particular model of any aircraft, let alone an Airbus or a Boeing, it doesn't matter, aren't - don't you go to a flight simulator and check out all kinds of situations so you know what to do in an emergency?

Mr. COX: The training, particularly initial training, is very rigorous for commercial airline pilots. But it's important to note that one of the things we don't teach is stall recovery. We teach avoidance, we teach the recognition.

And this is one area that the industry is beginning to change the way that they view things, is to actually take those very complex flight simulators and utilize them in new ways so that you not only teach the avoidance and recognition but you also teach recovery. And that's going to be something that we're - that is changing for the future for airline pilots and transport pilots overall. I think it's a good step forward.

There have been a number of us in the industry that have argued for this for a - decades. And I'm very hopeful now that we're actually going to see those changes go into place.

CONAN: It's important for those of us who are not pilots, when we say stall, we're not talking about the engines, not in terms of an aircraft.

Mr. COX: No. The engines actually have no part in a stall, either in the stall condition or the stall recovery. A stall in this definition is when the air over the wings is no longer attached to the wings. You're too slow. The angle of the wing is too great. And so the - that wing is no longer producing lift, and the airplane starts down.

Oftentimes, usually, you will see the nose will drop and then the corrective action by the pilots is to keep the nose down, reduce that angle and get the flow of air re-attached to the wing where it can produce lift, and then you can fly out of it.

CONAN: And, indeed, that was a crash you referred to earlier and one we know more about, the Colgan crash in Buffalo. That's the conclusion they reached. If he'd just put the nose down instead of up, he would have been all right.

Mr. COX: Colgan was a classic stall accident that at the onset of - the stall warning device, known as a stick shaker, the pilot responded with a very abrupt aft pull of the yoke. And in so doing, not only stalled it but stalled it quite severely, and then did it twice more. By the time the third stall occurred, they no longer had room to recover. And unfortunately, they went into the ground.

The Air France case is more complex, and we don't have all the information yet. But we do know, according to the French investigative agency, the BEA, that the aircraft was very, very highly stalled - more than twice the normal stall condition as far as the angle goes - for over three minutes. So it's a very long period of time.

CONAN: And when you say the yoke pulling aft, you're putting back on the yoke to raise the nose of the aircraft. That's the thing wheels attached to as well.

Mr. COX: That's correct in the case of the Colgan. Now, the Airbus uses a different control system. It's a side stick. But it's designed pretty much the same way. So that if you pull back, the nose will normally rise and that is not what you want to see done when you got a stall warning.

CONAN: We want to hear from pilots in our audience today. What questions do you have arising out of the flight of 447, the crash of Flight 447? 800-989-8255. Email:

And Ron(ph) is on the line, Ron calling from Eugene, Oregon.

RON (Caller): Good afternoon.


RON: I'm a private pilot. And a couple of years ago, I was in a situation just like this in - my Cessna 172 going into a mountain airport, had a tailwind come in that took all the lift out from the plane. And it was only from really good flight instruction that I was able to put that nose down in the face of being only 300 feet above the end of the runway, got enough air over the wings and was able to land it safely. And my two kids were in the backseat of the airplane. I mean - and I also had the experience, as a student pilot, of seeing someone approaching an airport, lift the nose up instead of putting it down, and they went wing over and crashed into the ground and died.

My question here is that this is very basic for, you know, single-engine pilots like myself, and it's drilled in. Can this get lost, this basic instinct, because I only had about a tenth of a second to make that decision, to put the nose down before the plane fell? In all the rest of the training that they get as airline pilots, do you think that they might be losing the basics?

Mr. COX: There's quite a bit a difference between the stall characteristics of a light aircraft and that of a transport jet. And as pilots progress to transport jets and transport airplanes in general, the focus becomes more on avoidance and to a slightly lesser degree recognition. And some of the reasons for that is it is very difficult to replicate in a simulator the stall characteristics of a jet.

A swept-wing jet, when it stalls, can be a very violent maneuver. It, in some cases, can roll off on a wing very rapidly. I've done it. And it is something that you don't want to do in, obviously, in the real airplane for fear of an actual loss of control event.

But - so the concentration is on the recognition of the stall warning system and a response to that. Historically, it has been done to lose minimum altitude. And while that sounds good, it didn't emphasize lowering the nose. So there's the transition in the past that pilots have made from the light aircraft to transports that - and how stalls are dealt with.

And you're quite correct in the small aircraft. You had experience with it. You've seen the airplane stall. And even while it's counterintuitive, you pushed the nose down out of training. With the transports, unfortunately, they were - oftentimes, pilots were taught just to push the power up and to let the power pull the airplane out of it. And in the case Air France or in Colgan, they were not in a position to do that.

CONAN: Hmm. Ron, thanks very much for the call.

RON: My pleasure. Thank you.

CONAN: Let's see if can go next to - this is Jeff(ph), Jeff with us from Ashland in Wisconsin.

JEFF (Caller): Yeah. Good afternoon. I, unfortunately, tuned in a little late. I am a private pilot. I have 3,600 hours, 37 years of experience, no heavy time at all. But what I don't understand is this is such a fundamental mistake and such a blatant mistake. I can't imagine why these guys did this.

In terms of stall characteristics, I will tell you I saw a Mooney and it rolled me over my back. So that was fairly violent. I've had (unintelligible) birds. But what we were taught and what has saved me once or twice, if you even think you're stalling, you push the nose forward. I was in my Mooney, and about 20 years ago, and I was in the clouds, (unintelligible), and a stall warning (technical difficulty) went off, released me.

CONAN: Your phone is betraying you, Jeff, but I think we've got your question. And let's put it to John Cox.

Mr. COX: As with the previous caller, it's important to realize that with smaller aircraft, this type of stall training you do includes recovery, where with transports, because of the difficulty in replicating it in simulators -and certainly we don't want to be doing this in actual aircraft - the transport pilots are more highly - or trained more frequently for stall avoidance and recognition. So that's - and I think the focus of that is going to be changed over time. And I'm very glad to see that actually.

CONAN: And the - you raised a point in one of the interviews I saw that you did. If this was a mechanical or an electronic problem, there would be instant demands that it be fixed. The pitot tubes, the airspeed indicators, there was a problem with those, they knew about it, they're going to be fixed. Is this a training problem that needs the same kind of focus and attention as a mechanical or an electronic problem?

Mr. COX: Well, I said that the same methodology, I think, should apply is that if we had a mechanical problem, we would derive a fix for it. I think I would characterize this as a training opportunity so that we can improve the effectiveness of training for professional pilots in transport aircraft. And in so doing, we can lower the accident rate due to loss of control, which is the leading cause of accidents.

So this is an area where we should be focused. There's a lot of work being done by a lot of very skilled individuals to make the recommendations in the training programs. And now, I hope the regulators will not only adopt it but mandate it.

CONAN: It's also important to recognize, we've been talking about the reactions of the pilots. Until we get more information, until we know more, it is not fair to conclude that they messed up.

Mr. COX: That's correct. And I've been an accident investigator, aircraft accident investigator, for about 30 years, and I know we don't know everything about that. I know also that we don't have the analysis done. And until the BEA provides us a complete report and lets us dissect the accident piece by piece and extract as much information as we can, then and only then can we fully utilize this tragedy to prevent a reoccurrence and - across the industry, not only for the individual airline or aircraft type, but for the industry overall.

CONAN: John Cox, thanks very much.

Mr. COX: My pleasure.

CONAN: John Cox, an airline safety consultant. He serves as president and CEO of Safety Operations - Operating Systems. He joined us from member station WUSF in Tampa, Florida.

And you're listening to TALK OF THE NATION from NPR News.

Copyright © 2011 NPR. All rights reserved. Visit our website terms of use and permissions pages at for further information.

NPR transcripts are created on a rush deadline by Verb8tm, Inc., an NPR contractor, and produced using a proprietary transcription process developed with NPR. This text may not be in its final form and may be updated or revised in the future. Accuracy and availability may vary. The authoritative record of NPR’s programming is the audio record.

Related NPR Stories