Two years after the Airbus 330 plunged into the Atlantic Ocean, Air France 447’s flight-data recorders finally turned up. The revelations from the pilot transcript paint a surprising picture of chaos in the cockpit, and confusion between the pilots that led to the crash.
Just read through the transcript of the pilots. Beyond confusing of why things happened like they did!
From Nova Scotia brings back memories of Swissair..
Highly disturbing. To call it a case of pilot error would be an understatement.
That the yokes function independently of one another amazes me. How can that be considered safe? It’s like having two independent steering wheels.
What’s why I walk everywhere.
mmmmm
Take-off power.
Stick all the way back.
Nose high.
Very slow air speed.
Rapid decent.
Buffeting controls.
Stall warning.
Lower the nose momentarily and regain normal flying speed. Continue to destination.
There has got to be more to this than just incompetent pilots. Even student pilots are aware of the proper procedures – after the first few hours of flying lessons.
Scary read. It sure is sad that they were really close to regaining control, but seemed to be mesmerized by panic.
Here is a similar story with a happier ending.
The Gimli Glider
At the time of this one, I was doing a lot of work related travel.
I hadn’t even realised they’d found the recorders. Was it ever announced in these parts? That was election time, so there were other things to cover, I guess.
That’s a horrifying account. I know there are those that swear they’ll never fly in an Airbus. Something like this gives me sympathy for the position.
What a terrible business.
The most challenging flying conditions of the entire flight
with a wrongly set radar and no sense of teamwork or
narrative by the one incorrectly flying the plane.
FRENCH!
We need to ban all french fries in western Canada again!
That was chilling to read. Some similarities with the Titanic disaster.
As an aside, they didn’t translate “putain” correctly.
I’ve never flown a plane but I’ve watched enough movies to know you drop the nose to pick up airspeed to regain control. You can’t climb out of a stall without a Harrier.
With the modern computerized wonderplanes the pilots no longer have a ‘feel’ of the aircraft. It’s like playing a video game where you actually can crash and die.
“Putain” does not mean “Damn it”.
When I read this a few days ago from an Instapundit link, I immediately thought of the Air France Flight 358 incident where an Airbus overshot the runway in Toronto and ended up next to the 401. The pilots of that one didn’t inspire confidence either. Here’s what I’d like to know about Pierre-Cedric Bonin, the pilot who held the stick back basically from the time the autopilot disengaged until the plane hit the water: how did he get his job at Air France? He didn’t have any relatives there, did he? Now I don’t know that he did, but I’d like to know if there has been any investigation whatsoever into that. Since it’s France, unlikely.
I don’t have much more confidence in Airbus (thanks, Brian Mulroney) than I do in Air France. There’s a great comment in that PM thread that mentions “the standard European procedure of denying that there is anything at all wrong with Airbus and blaming absolutely everything on human error”.
I truly hope someone with Airbus time will comment on this Topic. Pilots have thrashed this accident to death on other blogs.To call this “pilot error” and leave it at that is a bit compressed.
Airbus And Boeing have very different philosophies regarding fly by wire aircraft.
Simplistically stated, Airbus’s design philosophy was that the computers rule and override the pilot.
Boeing’s philosophy is the pilot rules and they can override the computers.
It gets little more squirly with Airbus on how the computers give the airplane back to the pilot when things start crapping out. (a brief explanation here
When that Airbus lost the flight data computers due to different inputs from the air data sensors, it basically threw the airplane back to the pilots. with a completely different set of flying control inputs.
And this happened just when the pilots needed it the most.(they just flew into a brick wall of a CB, some of which in the ITCZ can reach above 80,000ft, well above the max altitude of the A/C
)
They were close to being at what is termed “coffin corner”,,,The altitude and airspeed where an ac can no longer climb or remain at altitude at its max airspeed. In this region the autopilot is usually left on so as to minimize any divergence that would cause an upset while in that flight envelope.
Airbus really can’t justify a report blaming an inherent system design flaw (IF it is) on this crash, yet other crashes have occurred whereby the Airbus computers have crashed the a A/C.
Air France pilots rioted when they had found out that Air France had been laggardly replacing the inferior Pitot tubes with upgrade models as other Airbus’s had experienced the same icing problems with the deficient sensors
This has become a loaded topic amongst some ATR pilots
PS
To the best of my knowledge there is no “feel”mechanism in the Airbus joystick that would simulate control flutter inherent in a deep stall, giving the pilots a direct feedback to aid in ther situational awareness.(remember,,it was pitch black that night entering heavy storm cloud.
Boeing,,I do believe built that artificial feel into there system.
Also in an Airbus I don’t think one pilot can see the others joystick but I was under the impression there was a command button on each stick to determine the pilot in control. Not sure about that.
Any help from a Bus driver would be appreciated.
Merde, lots of stuff comment on, from pilot errors to design. All of which contributed to the accident. Not being familiar with Airbus design I cannot help but wonder why there is no indication for split control inputs. The article mentions that the computer averages out the differences of the stick but that is plain stupid. If a second stick input is detected then a warning about the original stick having control.
The senior co-pilot should have taken control(“I have control”) as soon as stuff started to deteriorate. Upon returning the aircraft captain should have verified who had control and then taken over.
There was never any reference or mention of check lists or discussion as to what is happening and what to do. Both co-pilots were working alone and against each other.
Upon finding the navigation error they should have reassessed their plan and options before entering the storm.
As a former Flight Safety NCO, I can say that there were several times in this whole incident that the outcome could have been different. An accident is never about just one thing. It is a series of events/happenings(contributing factors) that result in an accident. Discovery Channel has a series, “Mayday”, that does a good job of breaking down what happens to cause an accident and how they try to determine what happened.
“Both co-pilots were working alone and against each other.”
with that comment Tex pegs it.
it’s a guy thing really; males are not fond of the need to commUNICATE their thought processes. the very macho drenched pilot seat of a modern aircraft is a good example where the difference between need and absence of said comMUNICAtion is so stark.
I guess the trick is when travelling to the EU to gloat over their failings, do not fly Air Fwance.
aint hindsight fantastic for fixing things?
you would think that all aircraft, no matter how complex they are, would have manual overides so the crew can see and feel what is happening with the aircraft.
The comments at the article are very good, particularly from Airbus 330 trainers and pilots.
Gimli is referenced above. One main difference here is that numerous pilots have safely landed 447 in identical conditions in the sim. Last I heard, nobody – *nobody* – has landed Gimli in the sim.
Based on what I’ve read previous about this crash, and about the weather and pitot tube issues, I think its a bit disingenuous to simply blame pilot error. The cockpit environment would have been intense, with mixed signals, literally, bombarding the pilots. There would have been significant turbulence too, based on the weather analyses I’ve read. The issue of seat control, is important, though, and Bonin apparently paniced – in a manner akin to a car driving instructor stomping the mat with his foot when he doesn’t think the student is breaking soon enough. Two problems here – failure of systems authority vis-a-vis left and right seats, AND command authority between the pilots.
What Really Happened Aboard Air France 447!
When did Popular Mechanics take over the investigation from the BEA? Based on some book they know exactly what happened!
The final report for that accident will not be released until next year.
“All of the work carried out will be included in the Final Report, which will establish the causes of the accident, whose publication is planned for the first half of 2012.”
http://www.bea.aero/en/enquetes/flight.af.447/info07september2011.en.php
” … When did Popular Mechanics take over the investigation from the BEA? Based on some book they know exactly what happened! …”
No doubt the cacophony of confusing messages did not help , but the fact remains , which is not in dispute … the guy flying the plane pulled up on the yoke, when he knew they were very close the “coffin corner” , and continued to do so even when the system audibly warned of a stall.
It almost sounds like the guy flying the plane froze in fear or panic.
If it ain’t Boeing I’m not going. Airbus’ design phylosophy is dead wrong. The pilots are trained to always be in the automated mode. Manual flying is strongly discouraged. Even during an emergency situation automation is used. Have a single engine failure or missed approach? Use the autopilot, do not fly manually.
The reason why that plane crashed on the hudson river was because it was an Airbus. In a boeing the pilots could have kept the engine going until it melted but the airbus shut them down to protect them. As for 447 well in a boeing there’s a section of the manual that tells you thrust and pitch settings for situations when the instruments malfunction.
Air France is better know as Air Chance, and that I learned from an Air France Employee.
The computer averages the different inputs on the two sticks!!! Geeze, flying by committee…
A chilling read.
Who’s read AirFrame?
“Who’s read AirFrame?”
Thanks for reminding me. It was very highly recommended to me by a pilot friend back in the summer, but I forgot. Available now in my local library (soon as I finish REAMDE!).
By Michael Crichton for those who do not know of it.
Airframe? I’m looking at it right now. It’s sitting on the bookshelf right next to my Mark Steyn collection.
BTW It’s a great read.
“That the yokes function independently of one another amazes me. How can that be considered safe? It’s like having two independent steering wheels.”
When a pilot takes control, he is supposed to announce that he is flying the plane. There are also lights to indicate which side-stick is being used. That said, perhaps this design should change so instances like this where one of the operators has a brain freeze (or is suicidal) don’t happen.
One feature on Boeing aircraft that I like is the stick-shaker. Basically, when the plane enters a stall, in addition to an auditory warning, the yoke shakes. That way if the pilots are distracted by multiple system failures, tubulence, etc. it is unlikely they will overlook the stall.
As the article points out, if Bonin did not bother touching the controls at all, the plane would have been fine. Instead, he was pulling back on the stick the entire time while his team pushed down on the other stick. The article speculates that Bonin did not realize the plane was in Alternate Law, which in contrast to Normal Law, allows the plane to stall. Even so, with all of their instruments working properly (and the stall warning going off loudly!), he should have realized the plane was in a stall and made a correction.
“If it ain’t Boeing I’m not going. Airbus’ design phylosophy is dead wrong.”
They’re both exceptionally safe. I prefer Boeing’s design philosphy re:pilot control vs automation, but there is little practical difference between the two when it comes to safety.
Just another example of multiple human errors leading to catastrophe.
“If it ain’t Boeing I’m not going.”
Yep……..
Back when Scare Canada adopted the A320….I swiftly determined that a fly by wire kite should never be ridden without an ejector seat.
The prototype demonstrating a touch and go at the Orley Airshow….lifted up again and then promptly, automatically did a perfect wheels up landing in a woodlot….Scare India..an early adopter…an A320 did an imperfect landing into a golf course 1 1/2miles short of the runway in India.
3 computers that ALL MUST WORK to fly the plane….the controls are never connected to the control surfaces without the computer interface.
40 years ago somebody anticipated this:
The airliner had just taken of when the PA announced (metalic voice)…
“This is the first fully automated flight…there is no human flight crew…hence there can be no human error…so nothing can go wrong…go wrong….go wrong…..”
I bet the pilots where well versed in multiculturalism & social studies. Flying an airplane. Not so much.
The schools they went to probably thought Women studies & sensitivity courses where more important than leadership or mechanics.
They let Air France training system off far too lightly on this. They suggest that people will learn from the incident, but all these things were already known: pitot blockages have happened, people have crashed Airbuses by not knowing the logic of the fly-by-wire controls. This should never have happened.
My colleague was given an identical failure in the simulator today (not an Airbus), our third session on our first type rating – it should be an exercise practised occasionally in the simulator. I had just “died” in the right seat, so he was already unexpectedly flying solo. He nearly stalled, but didn’t and immediately understood the reason for near stall and easily flew the unfamiliar aircraft through an emergency landing.
The lesson learnt is that JAA/EASA should immediately stop all courses for the MPL (Multi-Pilot Licence). This is a course to fly only multi-crew aircraft, into the right (First Officer’s) seat of an airliner with very little time flying a real aircraft. What the handling pilot showed beyond the CRM mentioned in the article was a lack of three things: real, broad, unplanned experience to fall back on, pure handling skill and airmanship. The first two can only be gained in real aeroplanes, the last only be developed and then demonstrated to oneself and others by commanding real aeroplanes.
Certain training courses have always given pilots too little experience of real command; the MPL has made this worse.
Just another example of multiple human errors leading to catastrophe.
Beginning with the design of the machine…
There is a “Take-Over button” on the side stick controllers on Airbus’s but its function is not as straight forward as one might expect.
Simultaneous inputs form both sides are actively discouraged during training. If they are made, both inputs are simply summed and result passed to flight control computers, limiting value being 100% displacement. E.G. both full left sidesticks give maximum left roll rate command that is achieved with one sidestick in neutral and other fully deflected. If sidesticks are displaced at the opposite ends, sum iz zero and no command is given to FCCs.
When practicing incap in the sim, procedure was for the pilot simulating incapacitated pilot to push stick fully outboard and forward. Pilot taking over would have to press priority button and correct the attitude,. if he forgot the appropriate sequence and used dual input to stop the excursion, nothing much would have happened when the priority finally got pressed. Both controls and aeroplane have a lot of inertia and G-protection would prevent excessive load anyway.
IIRC, if one stick was latched out by holding the other takeover button, it could be restored by pressing its button, not any.
Oh, and PIC = pilot in command, independent of who flies. The gal/guy that handles the aeroplane’s controls is PF(lying) or H(andling)P.
There are also visual warnings (left or right green arrows indicating who has priority) as well as audible indications as to which pilot has assumed sidestick priority.
The other pilot did not take control because he was also overwhelmed by conflicting instrument readings and a computer screen showing multiple and cascading computer failures.
Doubting Rich at 5:57 PM
I was posting at the same time you were,,, and I concur as other ATR pilots I know have said the same thing.
Aviate,Navigate, Communicate is the primordial sequence in aviation. It’s obvious where this crew lost it. I just don’t think this A/C give them the tools they needed when they needed it, nor did their training.
JMHO
And I’m still wondering if they lost the vertical stab during that horrendous decent, and the had no way of knowing it. Airbus says it didn’t come off until impact. Others have questioned that since it has happened before due to excessive rudder input.
I’m not trying to malign Airbus,,, its just that its a different beast and was engineered that way from the start.
It’s not like I won’t get on one.
I’d just prefer a B737NG, that’s all
Cheers, G.
I haven’t been on PPRuNe for years!* Don’t really like the sound of that Airbus control, but then with over 3000 hours in pistons, controls firmly wired to the surfaces, I am perhaps too died-in-the-wool to give up that much authority!
With you then on the 737 – as is my father’s friend, the first person ever to fly me when I was 7. An ex-Mosquito pilot who would only ever travel British Airways on Boeings, and must have despaired when his son became a Virgin A340 Captain!
*although then I was in the top ten prolific contributors!
RFC
“From Nova Scotia brings back memories of Swissair.”
That is rather unfair on the Swissair crew. They did one dumb thing, albeit a rather understandable one, and then they could not have recovered the situation.
This crew probably believed the hype about their aircraft (yes, I do blame Airbus to this degree, but the crew should not just accept their nonsense!).
From that mistake they failed to believe that they were in a situation they had been taught to deal with since probably their fifth or sixth flying lessons. A situation they had seen dozens if not hundreds of times each in small aircraft and in simulators, training to recover from it.
For much of the time they did not even have a fault on the aircraft. They had all the data they needed, and the solution is one drummed into them from the start of their training, and one that could not have harmed them at that altitude even if they were wrong and it was inappropriate: smoothly apply forward pressure to the control stick and increase power.
From a AT pilot. (Not A-Bus)
[ I am told that is exactly what you are trained to do on the Airbus. They believed they had a computer malfunction and the equilivant to control=alt=delete is stick full back and full power to reset the computers. Problem was they were actually stalled.]
ron in kelowna at8:23
Buddy ,,, you’re on to something and it’s truly bizzare. Thanks
The Airbus maneuver instructs pilots to climb at a five- degree pitch attitude — the aircraft’s angle above horizontal — when airspeed readings become unreliable anywhere above 10,000 feet (3,048 meters). Only later in the procedure are they told to check whether it’s safe to level off.
[…]
New Instructions
Four days after the crash, Air France gave its pilots new instructions that contradict the Airbus procedure for coping with airspeed-data loss.
When the problem occurs at safe cruising altitude, pilots should “maintain the same pitch attitude and engine thrust,” according to the June 5, 2009, memo signed by Lichtenburger and three other executives. Crews should then troubleshoot “without carrying out the emergency maneuver.”
Most pilots realize that there is no need to climb when already at cruising altitude, Lichtenberger said. Air France issued the memo because it nonetheless “felt there was a risk that pilots might follow the Airbus procedure to the letter.”
Now I’m sick to my stomach.
This was actually taught to crews,,,
The little planes I used to fly give me no qualifications to speak on the intricacies of flying large transport aircraft so I won’t bother but what I do find interesting and I have read about it in many aircraft publications is the divergent opinions on aircraft design Boeing versus Airbus. Over and over again one can see that the Airbus is a socialist designed aircraft and the Boeing is a free enterprise designed aircraft. Airbus thinks the guy with his hand on the joystick is an idiot and does everything it can to keep the idiot from messing up the airplane. Boeing thinks the machinery is too stupid to fly itself and it needs the tender loving care of an informed pilot to keep the aircraft doing what its supposed to do. Airbus believes the ‘system’ is greater than the ‘individual’ while Boeing believes the ‘individual’ is greater than the ‘system’.
The Euros speak poorly of American engineers accusing them of taking shortcuts and leaving critical detail to the vagaries of manufacture. That conjecture seems to be exemplified by the differing control approaches taken by Boeing versus AirBus. Boeing designs their planes to work in concert with or in crisis rely on the skills of the pilot. AirBus sees their pilots as interchangeable monkeys. It seems to me Sully’s actions speak to the truth.
Response I received from Airbus Captain flying for EK
My 2 cents on this issue,
1.There is some truth to the hype about the Airbus flight controls laws etc… With over 6000 hrs on the 319/330/340, I have to say, Airbus is so much nicer to fly that any Boeing, but it’s a personal preference. Again, Airbus is so easier to fly but in this AF case would have been easier if the 2 FOs / pilots reverted to normal flying techniques, but it seems that we all have forgotten that those 2 guys have no basic flying skills !!!!! those are 2 cadets who got their 250 hrs on GA aircrafts and all they have been flying since then is Airbus and with no training on manual flying, how does anyone except them to be able to fly pitch and power when they have never done it before? (also, take this into consideration, that most of those 2 FOs flying is cruise time)
2.Airbus does have standby instruments and could have been used. Those 2 guys lacked basic system knowledge and all they should have done was disconnect the Auto thrust, maintain the same thrust (N1) setting / same pitch with is 2/2.5 deg up and reverted to the QRH for unreliable airspeed, but we all know what they did instead.
3. At EK, we have incorporated Stall recovery in our recurrent training , and I have to say, if you don’t recover from the initial stall indications, you will not be able to recover once those clean wings go into a deep stall.
4. I agree that Airlines didn’t train for this but the bottom line, those guys screwed up and it was a learning lesson / wake up call to everyone.
5.Did I mention to you that I did have an unreliable airspeed case in A340 about 6 months ago on takeoff once we got airborne and all worked out fine, the procedure is great if followed.