Air France Flight 447: Lack of training led to crash, report finds

Disaster could have been averted, French authority says


More than two years after Air France Flight 447 fell out of the sky over the Atlantic, French investigators have determined that a lack of pilot training played a key role in the harrowing nighttime crash that killed all 228 aboard.

This week the French BEA authority released its final report into the accident, which occurred on June 1, 2009 as Flight 447 was en route from Rio de Janeiro to Paris. The 224-page document concluded that the plane plunged into the ocean after a series of events led the crew to stall the plane near its cruising altitude and “the lack of any actions that would have made recovery possible.”

The long-awaited report confirmed what many in the industry had suspected: the current training regime for pilots is insufficient for dealing with so-called “loss of control” incidents that result from aerodynamic stalls, which are caused when a plane’s wings lose lift—usually because the plane is flying too slowly or trying to climb too steeply.

Needless to say, such accidents aren’t supposed to happen. Stall prevention and recovery is part of a pilot’s basic training. In general, the rule is to point the nose of the aircraft down and increase thrust until the plane’s airspeed is sufficient to restore lift to the wings. However, a Maclean’s investigation last year found that many pilots faced with a fatal mid-air stall not only failed to respond properly, but often appeared to do the exact opposite of what they were supposed to. The Maclean’s investigation also revealed that a combination of factors may be responsible for the seemingly unthinkable behaviour, ranging from the way pilots are trained in flight simulators to the increasing complexity and automation of today’s modern jetliners.

In the case of Air France, the sequence of events that led to the crash began when the plane flew through a band of thunderstorms. The pilots slowed the aircraft in anticipation of turbulence and then suddenly lost their airspeed indicators after the plane’s external sensors, called pitot tubes, failed—likely because they had become clogged with ice. But while the loss of accurate airspeed readings presented the pilots with a mid-air crisis, it should not have been fatal. French accident investigators found that the Air France pilots responded to the emergency in a disorganized fashion and unwittingly put the plane into a stall after the autopilot clicked off, as it was designed to do in the absence of accurate airspeed readings. More importantly, the co-pilot who was at the controls when the plane went down (the captain had left the cockpit on a rest break, only to return in the final minutes before the crash) repeatedly tried to point the plane’s nose upward despite the fact that the plane’s stall warning system kept sounding, albeit intermittently. The aircraft might have been saved had he done the exact opposite.

Among other things, French investigators recommended key changes to how pilots are trained to recognize and respond to stalls, including improved training when it comes to flying an aircraft manually at high altitude. The report also suggested changes to the design of some of the Airbus’s cockpit safety systems to make it easier for pilots to recognize when they’ve entered a stall. That includes the addition of an angle-of-attack indicator that shows the angle of the wing relative to the air flowing around it. “Only a direct readout of the angle of attack could enable crews to rapidly identify the aerodynamic situation of the aeroplane and take the actions that may be required,” the report said.

Filed under:

Air France Flight 447: Lack of training led to crash, report finds

  1. I’ve spent a lot of time on aviation forums discussing this one… The consensus has long been that the pilots responded to a non-catastrophic situation in a way that created a catastrophe. It’s hard to castigate pilots who lost their lives, but there was a glaring lack of basic airmanship in those final moments. They failed on Airmanship 101, not on some complicated technological problem.

    It’s become obvious that pilot training (and maybe cockpit design) has not kept pace with the complexity of modern airliners. Computers are great until they do something you don’t want them to do (as was the case here). Then the pilot must be able to revert to his basic skills and correct the problem. Unfortunately, this is proving very hard to do at high speed and high altitude when you go from leisurely, automated flight to stressful, manual flight in the blink of an eye. The general public may not know, but there have been a number of near-misses where, fortunately, the pilots responded appropriately.

    AF447 will go down as one of those seminal aviation tragedies that changed everything. It’s already begun – airlines, pilots, aircraft manufacturers and regulators have started talking a lot behind the scenes. Too bad it took a tragedy for this to occur.

    • Totally agree!! Well said

    • Sorry, cannot agree with you on the Airmanship 101 comment. Please try to place yourself in the exact same environment before concluding a failure took place:
      1-the fly by wire system switched from normal mode to direct mode without a clear indication to the crew
      2-you are flying in a violent storm, at night – so no external visual clues
      3-you are at high altitude where the airspeed margin between stall and cruise speed is quite small
      4-your instruments have been giving you erratic and conflicting information
      5-you have never flown the A330 in direct mode before
      My friend, which one of these situations are thought in Airmanship 101?

      • 5-you have never flown the A330 in direct mode before

        Then maybe you should have stayed at home.

        • Okay then all Airbus pilots should stay home. Statistically you are never supposed to end up in direct mode, so far as I know, nobody trains for it. Five independent airspeed sensors are not supposed to go down together either, but they did. It’s called common mode failure, not a regulatory requirements when the a330 was first certified. It is now becoming a serious consideration for all new designs.

  2. I wonder why the computer cannot talk to tell pilot to point the nose down? It is simple. Warning sound is not enough, in panic time you don’t know what to do with flashing light and sound. If the computer said ” push nose down”, I am sure pilot would have done it without thinking.

    • It did. But the pilot turned his nose down instead of the plane’s.

  3. Pilot error is our typical fall back position whenever we cannot fully appreciate the complexity of the problem at hand and how we got there to begin with. Think about it for a moment, you have three highly experienced, highly trained professionals. They each have families, lives, and responsibilities. They have been through one of the most rigorous selection processes on earth to be allowed to occupy that seat in the front of a two hundred million dollar airplane with over two hundred souls in their care. The real question is not whether they responded correctly. That is easy for us to conclude, especially, after the fact. The real question is, given the exact circumstances, would you, or any other trained professional have responded any better? We need to have the courage and determination to dig deeper, to fully comprehend the issues, instead of taking the easy way out and blaming the three that are not here to defend themselves.

    • Yes, as an incredibly untrained individual who has been on two airline flights and otherwise spent a fair amount of time playing in flight simulators, I would have responded to this situation completely differently. The pilot froze from fear when his autopilot disconnected. I can understand being a little nervous, I suppose, flying manually in general, but not at 35,000 ft. There’s no reason any pilot should be afraid to fly manually at 35,000 ft. This pilot froze from it and stalled the aircraft. If these people were professionals, I’m a goddamned flight instructor.

      I’ve have nothing but sympathy for the families of those involved, but not facing incompetence when it stares you in the face is a serious detriment to public safety.

  4. Balderdash! George Orwell is Laughing out Loud from above. The best scripting of NewSpeak I’ve heard, well, since last week. The hardware FAILS and they blame the Pilots!

    • You don’t know what you’re talking about. JP is right. Pilots, and to an even greater degree, co-pilots, have lost a significant degree of basic airmanship due to automation. After the airspeed indicators failed due to icing, the autopilot disengaged and manual control of the aircraft went to the co-pilot Bonin. There was nothing wrong with the engines, altimeter, artificial horizon, throttles, or the control surfaces of the aircraft. There was an audible “stall” warning repeated some 75 times during the descent. The fact of the matter is that Bonin was inexperienced, became confused, and panicked after taking manual control, and IMO his actions bordered on incompetence for a professional airline co-pilot. Bonin flew a perfectly mechanically functioning airplane straight into the Atlantic with 228 people on board. If the same guy had been in the cockpit when US Airways flight 1549 lost both its engines during takeoff and ditched in the Hudson River, those people likely would all be dead.

      • You are absolutely correct in saying that a fully functional aeroplane was flown into the Atlantic Ocean by the acting PIC. You are also correct in saying that our current breed of airmen are increasingly losing their ‘seat of the pants’ skills to automation. However, you do need to appreciate the vast difference between the autopilot switching off and the fly by wire system moving from normal mode, skipping the alternate mode, and going to direct mode without a clear notification to the flight crew – therein lies the rub…

        • Sir, I understand, and can appreciate the fact that the copilots were unaware of the switch to direct flight mode initially. In reading the narrative from the cockpit voice recordings, it also is tragically clear that neither the other CP, Robert, or the captain, when he finally returned to the flight deck, realized that Bonin had taken control and been pulling back on the stick practically the entire time causing the fatal stall. The captain recognized this after Bonin mentioned it but by then it was too late. As with most of these tragedies, multiple failures occured(not the least of which was just plain miscommunication in the cockpit), it wasn’t one thing, but because Bonin was inexperienced and clearly became rattled, he made irrational decisions with regard to maintaining control of the aircraft. Flying an airliner, or any plane, for that matter is serious business, and the consequences for getting it wrong can be far more serious than in most businesses. That, I think, is what we’ve witnessed here. Its real easy to hit the autopilot button, not so easy to fly in pitchblack night and/or in a storm. its time the airlines start training their flight crews more rigorously for emergencies and all facets of manual flight (i.e. takeoff, landing, high altititude, low altitude, etc.). Their passengers and flight crews lives depend on this ability.

      • It started with a hardware fault. Period. The Pilots didn’t fail. They were simply unable to determine the correct response to a hardware failure. That they may or may not have had the proper training would not alter the fact that any judgement or decision they would have made was predicated upon a situation NOT of their making, completely out of their hands. Machines break and humans are fallible. What Pilot can say they know the exact response to every situation, imagined or unimaginable.

        • So it started with a hardware fault, so what? News flash – thats the way most of these situations start – with a systems failure of some kind. Although the computer fault, predicated by the icing of the pitot tubes, was not of their making it’s their business to follow protocol and procedure and use basic piloting skills, which includes flying off the instruments at night and/or in bad weather. Yes, and I agree with you, the “correct” response was not to pull back on the stick while losing airspeed, pitch indicator up, and the audible warning system yelling “stall” repeatedly. The pilots did fail, specifically the co-pilot. Bad decisions were made, first by the other co-pilot, who was more experienced ceding control to him, and by the captain deciding to take a little rest before flying into a huge thunderstorm.

          You know, thats why they have these guys called pilots. They are supposed to know how to fly and there are systems checks that they do when they believe the aircraft is behaving abnomally. Thats why they train in a simulator. For emergencies. Saying the situation was “completely out of their hands” is ridiculous.
          Your statement is akin to saying something like “my turn signal indicator broke on my car so I ran headlong into a brick wall because I didn’t realize I was supposed to turn”. Rubbish.

  5. I think the greatest problem was that of Bonin not telling Robert in time that he had sustainably pulled back his side stick for the entire time they were trying to figure out what was wrong. Had he told him on time, the situation might have had a positive outcome. Team work is incomplete without communication and total openness.

  6. This Airbus A330-203 did not have multiple independent systems for detecting speed of the aircraft such as a GPS based system that would at least cross check the readings being given by the pitot tubes and then provide a cockpit warning that the airspeed could be wrong, or another safety mechanism whereby the pitot tubes are heated as long as this would not impact the reading so that ice could not occlude them.

    Thw accident was caused by the co-pilot induced Deep Stall condition and remained in that condition until impact.

    To recover from deep stall is to set engine to idle to reduce nose up side effect and try full nose down input. If no success roll the aircraft to above 60° bank angle and rudder input to lower the nose in a steep engaged turn.

    Pilots lack of familiarity and training along with system malfunction contributed to this terrible accident. Also the following contributed to the accident

    (1)the absence of proper immediate actions to correct the Deep Stall

    (2) Insufficient and inappropriate situation awareness disabling the co-pilots and the captain to become aware of what was happening regarding the performance and behaviour of the aircraft

    (3)lack of effective communication between the co-pilots and the captain which limited the decision making processes, the ability to choose appropriate alternatives and establish priorities in the actions to counter the Deep Stall

    Practicing recovery from “Loss of Control” situations and improve flight crew training for high altitude stalls (simulator training usually has low altitude stalls which are significantly different due to energy status of the aircraft) should become the mandatory part of recurrent training,

    • All modern airspeed sensors, including the ones found on general aviation aircraft, have a means to heat them to prevent ice build-up. The five independent airspeed sensors on this aircraft were the subject of an AD, and slated for an eventual replacement – as fate would have it, not soon enough!
      Keep in mind that GPS gives groundspeed whereas the wing flies on airspeed, and the two are usually not the same unless we are operating in zero wind condition – certainly not the case in a storm like the one encountered by AF447. You would need a very intelligent GPS based backup system – one that is programmed by people that are fully conversant with aerodynamic AND aeronautical principles.

  7. When the pitot tubes failure caused the computer to suddenly dump the plane into Alternate Law wherein a pilot is now flying the plane, the assumption is that the pilot knows how to do this. Unfortunately “seat of the pants” flying experience is a thing of the past. The average passenger has more experience flying an airplane than the pilots do. Recognizing a stall and recovering from a stall is usually the third lesson for a student pilot in a small plane, but these “button pushers” lacked basic skills simply because its too expensive to let pilots actually fly airliners anymore.

Sign in to comment.