Back to Squawk list
  • 192

What Really Happened Aboard Air France 447

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. Read more: Air France 447 Flight-Data Recorder Transcript - What Really Happened Aboard Air France 447 - Popular Mechanics ( さらに...

Sort type: [Top] [Newest]

Wow, this is amazing. It's obvious the lack of experience and the lack of communication of the flight deck while the Captain was not there led to the deaths of so many. If the junior pilot had just stated that he was pulling back 30 seconds earlier, perhaps they would have all been saaved.

A potential flaw is the sidestick - if the two had been connected, then the left-seat pilot would have known the whole time the right-seat pilot was pulling back. You can still have fly-by-wire while having the sidesticks either mechanically or electroncially connected (inputs from one stick activate small motors to move the other stick). Another is the apparent lack of positive notification that the aircraft has departed normal law and entered alternate law.
Very gripping read indeed and a different perspective.
Took me a while to find this again. For some reason it is in "Last Weeks" posts. Anyway, this was a pretty chilling article to read. I hope people stop bashing Airbus now.
Technical glitch, we just corrected it. Thanks!
I can't imagine the ultimate shock that you've gone from cruise altitude on an overnight TATL flight to...

02:14:23 (Robert) Putain, on va taper... C'est pas vrai!
Damn it, we're going to crash... This can't be happening!
Yeah.. I just noticed my post doesn't show up in the 'recent' or 'newest' categories.... odd.
Looks like you posted this story after this article, but you probably didn't see the dupe due to the issue Alistair mentioned. We fixed the problem and bumped the time up on this article so it'll appear in the newsletter tomorrow. Sorry for the confusion and thanks for submitting it.
AF447 has been beat to death on comments as to what happened, both before and after the recorders were found. The transcripts here from the recorders pretty much match, to some degree, some of the causes that were proffered. There are several key things here that come out in the article and need to be looked at. 1.Unless I missed it, there was no clear warning that came out stating that they had gone from DIRECT to ALTERNATE LAW. As the article notes, you can't stall that AC in DIRECT LAW and they apparently thought they were still there.2. Those "asynchronous" Joy sticks. As a Boeing man anyway, all I can say is that is plumb stupid.That computer is only as good as the input to it and if one can pull back and the other cannot tell it or goes the other way, it won't know what to do.#3. I would have thought that since the Captain gave the left seat to the more experienced pilot, that would have put him in charge(PIC) but apparently it didn't happen. Maybe it was assumed when it should have been stated. That is why there is a Captain. Someone has to be in charge and in this case, that would have stopped some of the confusion.
JD345 0
Concerning direct and alternate law... it would seem they didn't have the foggiest idea that they were in "alternate law" and assumed their inputs would be approved by the boss like usual. That's a pretty big deal of a change... you'd think there'd be an obvious warning. Maybe there was.

Some of the chinks in the Airbus systems showed here, but they had a stall horn blaring at them the whole time... with their psychological state and the resulting chaotic chain reaction, I tend to think they were going to bring the plane down one way or another, whether it was an A330 or a 767 or anything else.
This only adds to my suspicion of fly by wire. In certain circumstances it adds one more layer of problems for the pilots to solve. Damn sure better be as good with the computer skills as you are with piloting skills. I'm probably just getting old.
James: it ain't the fly by wire tecnology, it's been around for a good while. Sans joystocks, Boeing is staying with the yoke, the 787 is fly by wire and several other models are partially there or have it in use. The thing is that conflict between those 2 joysticks on the Airbus that myself and Jack have alluded to here in this comment string.
I certainly am not well educated on the system, but I remember the early days at AB and airplanes doing wild things, even crashing. Remember when they taped cockpits and listened to how many times the pilots Said "why is it doing that. etc". If they have joysticks that operate like discussed it would certainly be a good thing for the pilots to understand. As I said part of that other layer. I know very little about airline cockpit training but it's obvious the CRM didn't play well here. The left seat should have made the right seat let go of everything if he was taking control. All ashamed, but sometimes lessons are expensive.
Also I didn't really mean the fact that the control movements in fly by wire is the concern, it's the software integrated into the system that can take absolute control away from the pilots at times. When I hear about these different laws and the fact the pilots may have not known which law was in charge at the time it adds another layer for them to negotiate. Programming joysticks independent of each other seems ludicrous. Computers are great but sometimes you just want to disconnect.
JD345 0
In the end, it was psychology that brought that plane down. But there were plenty of contributing factors -- the one that's killing me is the dueling joysticks.

Is that REALLY what happens... the computer just splits the difference when it gets conflicting inputs? Who in the blazes came up with THAT arrangement? Seriously... someone had to program it to do that. Under what circumstance would that have a positive outcome? Wow. You'd think there would be some feedback... a horn, vibration, even a disagree light somewhere... is there? If the guy could have known when he was pushing down that the other guy was still pulling up, maybe this wouldn't have happened...

The fact remains that regardless of the way the controls work, you can't fly one airplane in two different directions... that the situation even got to the point of disagreeing controls says enough... seems they weren't going to take "yes" for an answer.
Jack, I totally agree -- this "feature" is not only idiotic, but had the system notified the pilots that it was receiving conflicting instructions, then Robert would have realized that Bonin was flying the plane into a stall. Just that one thing would have saved them. But yes, the problem was that Bonin panicked, and he should have been kicked out of his seat at that moment. This is the exact same panic reflex that crashed the US Airways/Colgan flight at Buffalo.
These 2 " Pilots " seems like they can't even fly a Cessna! Why in the world would you pull back the stick when you know you want to go around it especially a 7000 VS CLIMB!
JD345 0
that's exactly it... some of the nuances of the Airbus systems were contributing factors, or at least didn't do much to mitigate the situation... but with their mental state they weren't qualified to fly any airplane... and didn't, which was the problem.

Bad things happened, and for whatever reason, it was at just the wrong time. They were playing catch-up the whole time in a trade where you don't get a particularly large margin for error.
Rule Number One.... FLY THE PLANE!!
Great read,very well laid out,i enjoyed it.
Some of the "translation" has been sanitized.

They ignored some basic flying rules. If one suddenly loses airspeed indication, look at the GPS ground speed ! If G.S. is still reasonable, it is the A.S. indicator malfunction.

If plane is "falling" in a full wing stall; push stick forward ! DUH !

Years ago, I got to visit the cockpit of an AirFrance 747 enroute K-CDG to K-IAH. Mid Atlantic, at altitude, the flight crew had everything turned off and laughed when I inquired of the INS (Inertial Nav).

They told me they did the crossing using a wristwatch and a whiskey compass, approximating a "great circle" route.

This plane had a loaded instrument set like I had never seen and they weren't using any of it.
I don't think the airline would have liked that! Just like why most pilots, I assume, don't hand fly jetliners at 38000 feet, its an insurance aspect. Turbulence, they get disoriented, that's a big price to pay.
npog99 0
That's the attitude that gets people killed. Flying can be fun. But, arrogant attitude spoils everything.
Maybe I'm a chicken, but I fly with everything on ALL the time, except I may turn the RADAR on standby if there is no weather and I have 30 miles viz. I have seen too many things go wrong in 45 years of left seat.
Richard, I am just like you; I had 37 years and still do fill in. All that stuff is there for your use; the "no old bold pilot" saying comes to mind.LOL
As has been pointed out by many others,the first duty of a pilot is to FLY THE AIRPLANE.
Something wrong ? Check arirspeed, attitude, level wings, control airspeed, manage power carefuly return to heading & altitude. Everyone on here knows this well.
I was not familiar with dis-coordinated sticks and can't imagine which idiot thought this up. The pilot must be able to know what the fo/(co-pilot) is doing w/o trying to figure out a puzzle.
Please tell me that their GPS and other long-range equipment wasn't turned "OFF" ! If g.s. is 3 x VSO, the problem isn't airspeed. I don't know what the stall speed ot the AirBus is; but,they surely weren't supposed to be driving around over open ocean at 100kts ias. Or, 100kts g.s. either.
We all know that a situation can develop and accelerate very quickly, may GOD Bless these dead pilots and their pax.
Now, investigate the AC and the AirFrance procedures.
Jack alludes to all the confusion, patly caused by the AB systems, BUT, for whatever reason, there was a total loss of situational awareness, probaly caused by the Tstorms and all the turbulence, and ckpit confusion. The started somwhere around FL350 and did not realize they were about to crash til the got down around 2 grand. At their fall rate, it took them close to 10 minutes to realize that. That's a long time.
npog99 0
I am glad you mentioned it. It appears as if most of us forget the lack of CRM in the case. CRM is proven to halt the chain reaction leading to accidents. These guys were too confused and scared to function logically. Their training didn't help them much in this instance. The hand pulling back on the stick was probably frozen in place, and ignored. Just STOP. Reassess the situation, loudly and clearly state your actions, use CUSS words...

This was a perfectly flying machine, except for the pitot tubes--and the fact that it would make sense if the unused stick would somehow reflect what the other is doing--the plane was flying as commanded. When the captain walked in the cockpit, good implementation of CRM would have noted the captain about the stick being pulled back, and he would have probably correct it.

Besides that... If I were a pilot, wouldn't I want to know the different statuses the computer flying the plane with me has? Wouldn't I want to know how the computer would react in extreme situations?
If nothing else, relax the controls !
Pileits 0
Umm you mean relax the wire TO the controls :)
Weell, I'm not AB typed but I would think that even on that system that if you released everything, it would go back to normal. Now, whether that would have done the trick with the AP off and them in ALTERNATE LAW, who knows?????????????????????????????????
Another Air France Airbus experienced the same situation, ironically at the same location and recovered. Here is how that cockpit crew handled it:
Great post, Elizabeth; Thanks !
Personally, a difference in crews, their experience, and cockpit chain of command in light of this other.
I've been following the discussion carefully on the PPRuNe Forums for almost two years now, so I'm familiar with all of the info that's been published so far, and the hundreds of different opinions/theories about what happened that night. This isn't aimed at Wayne; I simply wanted my post to appear at the newest position here, in order to clear up a few confusions.

First of all, when the A/P disconnected, the PNF announced "alternate law," and the PF (Bonin) announced, "I have the controls." So, they knew what flight law they were in. It is NOT at all clear that Bonin knew that he had almost no flight protections from HAL. Training on this issue will be a major recommendation.

Second, it was stated above that the went from DIRECT Law to Alternate. Nope. From NORMAL Law (flight envelope fully protected) to Alternate.

As to the non-linked side-sticks: what's said above about the computer summing the two inputs is correct but leaves out the fact that each pilot has a red override button on the stick which allows complete takeover of control from the NF pilot. Apparently this was not used at times that "control" (ahem) was passed back and forth.

In reply to Wayne's statement, "#3. I would have thought that since the Captain gave the left seat to the more experienced pilot, that would have put him in charge(PIC) but apparently it didn't happen. Maybe it was assumed when it should have been stated."
The Captain clearly (on the CVR) put the junior (right-seat) pilot in charge. In hindsight, of course, we all wish he hadn't, because there's some evidence that the PNF had a better idea of the situation but did not use proper CRM to insist that the plane was being mis-handled. (!!!)

Confusion was no doubt multiplied by the fact that the stall warning did NOT "sound the whole time," as Jack said. In fact, it shut off several times due to impossibly high AoA, rendering the sensors useless, but resumed as nose-down input was tentatively tried, because if the aircraft had been HELD nose-down for a longer period, the stall could have probably been recovered. This is because the pitot tubes by then had cleared and reported valid speeds, which were way too slow, hence a stall warning. Sorry if this isn't real clear; more coffee is required.

And finally, sure, both A and B airplanes are FBW, but the amount of so-called "protections" is totally different. The design philosophy of the software is different, and IMHO, Boeing's system is far less confusing and allows much more control when there's a problem. I could go on.
Normal should have been my terminology, but I think I was going on what the article said. They may have known when they went to ALTERNATE LAW, but in the ensuing confusion they must have forgotten, and although it would explain why the jr. was PIC, I never saw where the Capain give that order. Override or not, there should be something done on those joysticks and I am in agreement Boeing has the better system.
In the particular Alternate Law the plane was in, ALT2, there is no AoA protection, but roll protection is left in place. The PF apparently spent most of his effort trying to correct roll, while continuing the fatal nose-up input. (AoA is what brought the plane down.) From this, it's fairly clear that he had no idea what protections were/weren't in place in this Flight Law. Training.

Also, I'm not the only one who feels strongly that AB has too many damned Flight Laws, adding to possible confusion.
JD345 0
Step 1: Define the problem
Step 2: Devise the plan
Step 3: Execute the plan

This is basic problem solving... AF447 crashed because they started on Step 3.
On an Airbus you can't stall the plane in NORMAL law. In DIRECT law you certainly can. Any Airbus pilot that is paying attention should be able to tell when the plane is in a reduced flight control configuration (i.e. not in NORMAL law).

This accident stems from the training philosophy of the last decade or two where people who fly airliners are "system managers" and are no longer airplane pilots.
2 words..."my airplane"
One woulld think that Airbus would change the controls so that the left seat controls match the right seat controls. That at least would let both pilots know what the other was doing with them.
There has been MUCH discussion (on PPRuNe) about whether or not the inability of the PNF to see the PF's side-stick was a contributing factor. Pictures were posted showing that he/she cannot see it. AB defenders feel that the PNF "should have known what was going on" by watching the instruments. It isn't even entirely clear just WHAT the pilots saw on their instruments. Too bad we don't have video recording of what went on in the cockpit. This has been successfully fought by pilots, but I think they should stand aside and let this happen, because of cases like this one.
So tell me did any of thosr posters on that PPRune web site have the guts to talk about the fact that to enable the flight plan to be "kosher" they have to enter BREST as the destination and not Paris CDG? So in effect the aircraft has no spare fuel to fly around the weather because already they need another 150km to reach Paris. You know what AF is a joke. The authorities are like a bunch of scared rabbits caught in a bright light. They already KNEW there was an issue with the piotot tubes, they already KNEW the flight plans were compromised and Airbus already KNEW there was an issue with controlling the AB products.
Nope, never heard about the destination issue. Very interesting.
One would think that Airbus would make a change to their control design to have the left and right pilot controls follow eachother. It is relatively easy to do using servo connected controls. At least that would make both pilots aware of what is happening with them.
These guys simply forgot how to fly - due to over-reliance on automation. Forget the instruments, forget the warning horns, Pitch + Power = Performance.
Stephen, you are right on.Mr. Murphy will rear his head at the most inopportune moment at times, and regardless of whether you have a C-150 or a 747 under you, you had better know how to fly the plane.
Perhaps they put too much into the cliche "Think ahead of your airplane."?
I think you lose all pitot indication on an aircraft this size, you need some better pitot tubes. Also, the flight directors must have been showing a lot of blue. How hard can it be to figure out you need to put the nose down? Just sayin.
Well, as it says, for whatever reason, there was total confusion in the cockpit and regardless of what they shoulda, coulda done, even with the pitots coming back, the totally lost situational awareness and it was not really regained and recognized until too late.
And they (Air France) already knew that. There had previously already been 37 unreliable airspeed events with these substandard pitots (all successfully recovered), and they were being replaced at a much-too-relaxed rate! AF is gonna have to pay bigtime for this horrible accident. After the crash, the pitot replacement program was stepped-up radically.
Just what do you mean with the phrase "just saying?"
Stephen, that is a great and valid observation. 'Bus might be well ahead of Boeing in trying to substitute a complex but unthinking and unintutive machine for the incredibly more intelligent and innovative Human, but they're both on the same path. How many times has the Check Dude in the Sim told you to "use the automation- it will keep you out of trouble and you won't bust the check?" Very true in the cookie cutter environment of the Sim Check- you have to do A- B- C without error to get through it. You know how to do it- very little learning taking place- just don't screw up!
From the moment I flew my first Airbus in 1990 I thought there was a serious design flaw. (for those not type-rated on AB; unless the priority button is pushed, the sticks apply an algebraic summation of both inputs.)

Clearly, we see that the junior pilot in the right seat pulled all the way back and simply remained frozen. A full forward stick force from the left seat would only neutralize, and not overcome the right seat input. (priority button NOT pushed)

There is no sensory feedback and absent any tactile situational awareness, the pilots are in their own cocoon of disinformation.

On a Boeing (fly-by-wire or not), there is no need to look all the way across the cockpit to see what the other pilot is doing. It is right in front of you. There would be a big, fat, honest yoke.... hard back in your belly that would be telling you what the other guy was doing through this whole event. (there is a small indicator that shows stick input, but very counter-instinctive in a stressful situation)

Auto-trim is another sensory deprivation on which the Airbus designers have a lot of answering to do.

Fred, I would not expect the designers to implement changes to systems now. It would be treated as a huge admission of liability.
Exactly so! The final report, due in July, will no doubt blame only the pilots, IMO. France, as part owners of both AB and AF, stand to lose too much. The current reports are incomplete as to the cockpit conversations, and who-knows what-else.
Wow. I didn't even want to read it, knowing the outcome. But, a lesson learned...
JD345 0
It's hard to see the attitude indicator when you're suffering from recto-cranial inversion. It was a non-event until the guy put the plane in what the article says was a 7000 fpm climb, which wouldn't be appropriate in any situation, let alone this one.
My AF Frequent Flyer card landed in a bin
AND......I will NEVER fly on a 'Bus!
That can be very difficult, good luck, I'm boarding A330 on Tuesday
Scott: that is a noble comment but if you are flying any of the legacy carriers, with the exception of AA(for awhile) you may be subject to flying on one as they are in most fleets. You may book a flight showing other equipment but any airline is subject to making an equipment change for various reasons and unless your travel plans are very,very flexible, at some point and time you will wind up on one.LOL
What do you mean by the phrase "just saying?"
I'm just saying- there was more than enough information to FLY THE AIRPLANE.
power and pitch, nose on the horizon, 80% power and it won't fly out of the sky.
That is an eye opening transcript. It makes me think the crew make up for these flight is done on a cost savings basis. I would think it is better to have two captains in the crew and one copilot so there is always a command pilot in the cockpit at all times. Had this been the case maybe AF447 would not have fallen into the sea. At least this is what I come away with after reading the transcript.
based on the fine leadership of Le Captain, I would have just been more cash at the bottom of the ocean
Yes based on this article. Had a full captain been on the flight deck the entire time this may have been a none event. Hind sight is 20/20 so we will never know.
Well, I didn't see it in the article but if the Captain did in fact give Bonin PIC before he left, by qualification, he gave it to the wrong one. That being said, the left seat guy just came off rest and had no idea of what was happening in the cockpit; even when the Captain came back in, he took the jump seat to look everything over, and regardless of hours, the Captain may or may not have known the full background of the PIC. If he was AF qualified, he would have no real reason to question him, especially if it was a pickup crew or one that hadn't been together long.
Things pointed out in this article and some of the remarks made here are the reasons I avoid Airbus. I use to love US Air, they went very where I needed to go. Then they switch to Airbus, I switched to United & Delta. With mergers & United retiring the 737 Hello Amtrak I'm hoping with the United merger we will see more of the Contenetial 737's back on the east coast flights.
Well Delta was 100% Boeing until the NWA merger came along and I think on replacement of those they will stay with Boeing. The future will be interesting. You basically will have to overlook AA. They always ran 2 and sometimes 3 brands just to keep everybody honest
yeah and look at SWA one problem and the whole fleet is down.... SAme goes for Delta in the past.... Boeing or AB doesnt matter as long as the entire fleet isnt just one type....
There is some merit there and as you mentioned with SWA, Boeing has had it's problems too, as all mfg's have had, BUT, unless I am missing one, even where there is a single mfg, most airlines are not relegated to 1 AC type like SWA is, so unless it is a part or something common to all tht mfg's AC, I don't think you'd have a problem.
This again? Really?
Normal law? Alternate law? Direct law? In my opinion, one should never have to second guess how an airplane will react to hand flying control inputs. This entire design philosophy is deeply, deeply flawed.
I totally agree. As an an ATP type rated in any aircraft, you are required to know, in so many words, what the flight envelope is for that Aircraft, and lot's of times in an upset of any kind, you have time to deal with what's happening to you and nothing else. Fly by wire, hydraulic or cable, a control needs to respond per Pilot input. That's what we get paid for. The next step they will for is make the damn thing fly itself. They almost will right now but they ain't figured out how to make them take off and land automatically yet.Even the UAV's have a controller/pilot somewhere.
boughbw 0
@Alistair: Why stop bashing Airbus? The article places the blame squarely on Bonin for failing to recognize the situation, and there is no doubt that Bonin contributed mightily to the crash. But the plane itself represented a major obstacle to successfully completing the flight.
It was Airbus' pitot tubes that froze over - a problem Airbus had identified on the A330. It was Airbus' fault that when the pitot tubes froze over, there was no warning indicating such; nor was any indication given when the author of the article presupposes that they had come unfrozen and back online. How do you reconcile such major variations in speed? If you don't realize the freezing problem a) occurred and b) was solved, what do you make of the readings?
What the author also leaves out is that each of sensor systems disagreed and the computer was determining what readings to show the pilots. Is this also Bonin's fault?
The repeated admonition on each of these threads about AF447 is that the pilots should have first and foremost "flown the plane." My point is simple: on non-Airbus aircraft in the same situations, you have a far greater chance of success in that goal.
Wow. Even a novice flyer in a 152 knows more than these guys did!!
Arthur: I don't think the knowledge was the factor, it just appears they got totally flustered and confused and were unable to apply that knowledge. Although some big iron is different, there ain't anything flying that I know of that won't respond to a nose down for stall recovery. One main thing I noted in the article, was that it said the pitots unthawed and went to working again but I can't find anywhere that any indication was given to the pilots about that, that the ASI was showing good numbers again. If it was, all they should have had to do, I think, would have been to throw the AP back on and the plane would have fixed itself. That's country boy thinking as I am not AB typed. If there was a notification of some type, it got lost in the confusion.
They didn't know they were stalled until it was almost over. There was no horizon-- storms and no moon.
There has been NO, until this accident, high-altitude hand-flying training, at AF at least, and NO high-altitude stall recovery training.
The stall training he did get, for low altitude (wind shear) taught him to noseup (he did) and apply TOGA (he did).

Once an AB goes into alternate law, you cannot turn the AP back on at all. :-(
Scott: Question? Is there a way to get it back to NORMAL LAW quickly?
Lots of discussion and truly unfortuant for everyone. I have never flown one of these computer planes, mine had vacuum tubes. I was taught to FLY the PLANE even without airspeed. I have had complete / total pitot/static failure flying over the mountains and used power settings and pitch attitude for control. Granted glass panels must have power but I have to assume that the pitot tube freeze over means they still had electronic pitch control through the ADI and they had more than one of these. Powere and attitude will fly the plane. This looks to me as bad design of the plane for the pitot tube to freeze over, and lack of traning on how to fly on pitch/power coupled with stall recovery. I realize things happen fast and I have had emergencys over the North Atlantic myself that could have resulted in loss of the aircraft. However fly the plane and use what you have is critical.
I agree -- not only is this an idiotic "feature", but if the system then notified the pilots that it was receiving conflicting instructions, Robert would have realized that Bonin was flying into a stall. This one thing could have saved them. But yeah, the real problem is that Bonin clearly panicked and he should have been kicked out of his seat. This was the exact same panic reflex that happened to the US Airways/Colgan crash at Buffalo.
When I was in the Air Force I worked on everything the Air Force had that had four engines or more. I've also done some flying on my own and on some maintenance flights some pilots would allow the mechanic to have some right seat time to feel what the controls were doing.

My specialty was the repair and maintenance of flight controls, a certified rigger if you want. There is not replacement for the feel of the yoke responding to the pilots input. That is what is missing in fly by wire systems. Taking the feel for what the aircraft is doing away from the pilot is as bad as taking the feel of a mechanics tools away from the mechanic.

Without the sensory inputs felt in your hands and arms a big part of the equation is missing I won't go as far as to say the job won't or can't get done but the missing piece slows down the decision making processes. It is easy to ignore a voice, it is almost impossible to ignore a stick shaker which is what would have been shaking had this been an aircraft with standard flight controls. Had there been a stick shaker maybe the pilots would have better realized what was happening rather than trust that the aircraft systems would not allow them to stall the aircraft.
While the joystick is on most of the newer military fighters now, the big iron still has stick/yoke and all the Boeing stuff, whether fly by wire or other, is all stick/yoke and those other features are there. It is not quite the same feel but you can feel the AC respond.
Wayne... the writing is on the wall here. Airbus flight controls are not fit for the purpose. Arrogance and hubris will win over the reality here, as usual. The FAA could regain some credibility here by DEMANDING changes but as usual they will chicken out.
They pulled a Marvin Renslow.
wow! what a read! the confusion re those joysticks and pilot's confusion caused the crash...scary to know when so many lives in the pilot's hands.
Dave Mahoney's comment says it all: wow...
I am no superstar pilot, but I've been in a very similar situation, and walked away from it because I immediately reverted back to the basic understanding that at a specific power setting, and at a specific pitch, the aircraft will behave a specific way. In my case, we (inadvertently, and foolishly) took off into icing conditions with the pitot heats off in a high-performance jet aircraft. The pitot tubes froze over, the airspeed indicator started acting as an altimeter, leading me to increase my pitch, leading to further increases in "airspeed". This went on for a few more seconds until my first indication of a stall - which immediately told me that something was wrong with my instruments. In an instant, I reverted to what I knew: pitch + power = performance. I lowered the nose to what would be a normal pitch, got my bearings, realized the issue, and took corrective action (pitot heat: on).

This airplane was fly by wire. There were wires connecting the interconnected yokes directly to the control surfaces! Call me a fuddy-duddy, but that kind of interaction with the airplane makes us remember that we're flying an airplane, not a system.
very good technical article - excellent findings after two years.
error of two copilots without training on who is in control. the discovery of the real reason behind this will save thousands of lives in future.
It's hard to believe professional pilots will make that mistake, mind blowing.!!!!
Scott: I didn't see any mention of a stick shaker as a stall warning. Doesn't the A330 have a stick shaker?
I didn't see any mention of a stick shaker stall warning device. Doesn't the A330 have a stick shaker?
It don't have a stick to shake
Avi8tor2 0
I'm probably gonna "stir" some of you up with my biased opinions however, after logging flight time for more than 25 years, of which 12 of those years have been as a Professional Pilot, coming from a true background of Wolfgang Langewiesche's "Stick and Rudder" flying all the way up the ladder to Jets/Turbines, who else out there would make a bet with me that the most "inexperienced" co-pilot on this doomed flight, started his career in a program similar to the one recently implemented by the ICAO, the MPL or Multi-Crew Pilot's License, where he probably never experienced real stick and rudder flying?!? Granted, he was a fully qualified and highly trained "Systems Manager" but was he truly a fully qualified, highly trained PILOT?........
You may stir up some of them but not me cause I am just like you in that opinion. Those programs ain't that bad if you know how to fly, BUT, it don't help to have a smart aleck plane that won't do what you tell it to either. Like sombody said above, AB has too damn many flight laws and if you have an upset of some type, you don't generally have the time to figure out which one your in, let alone change it. If I activate a control, I want it to do what it is supposed to do without anything between me and it.Just sayin'
JD345 0
Put traction control that's supposed to help you modulate the throttle in your car when it's slippery... next thing you know, everybody's just pounding on the gas and sliding all over the place, and they don't know to drive when it's slick anymore.

Put ABS that's supposed to help you pump the brakes on your car... next thing you know everybody's just mashing on the brakes no matter what the conditions and they don't know how to brake when it's slick anymore.

Put a system that's supposed to help you fly the plane on an Airbus... next thing you know, a guy grabs the stick and pitches up into a stall for no discernible reason...

Just sayin'...
That was deeply disturbing. Seriously, despite the attempt at soothing words at the end of the article, I find my faith in pilots disturbed.
Just saying... Is Air France cursed. Since the Year 2000 they have had notable crashes-total hull losses; The 2000 Concorde crash, the crash in 2005 at Pearson Airfort YYZ, and this one.
I do have two serious questions:

1. Was the JR pilot made PIC because of the need for currency hours (an ongoing issue for oceanic crews)?

2. When it said the pilot adjusted the radar, was it in a weather-above mode, before he corrected it, and since they thought they were leaving the problem clouds in that mode, could the pilot have then decided that he wanted to get up to that clear spot ASAP? That would make some sense of the mad climb.

And how much time did the JR. pilot have in light aircraft? I understand the value of experience, but in a practical sense, the JR pilot should have fresher memories of direct piloting. But either way I guess the act of panic made him a non-player.
JD345 0
I rode ( on Monday and suppose I can't complain because it did land, on the ground, intact at the end of the flight. They had "Please turn off electronic devices" instead of a no-smoking sign. I hadn't seen that before.
JD345 0
okay, this was supposed to be replying to Scott Hawthorn halfway up the page, but whatever.
did anyone in the cockpit notice the negative g forces of them falling at 10,000ft/min? I mean with the nose up and feeling the negative g's would tell you that you are falling nose up. From what I understand, planes are similar to cars in a way, once you let go of the steering, it straightens out. And COME ON MAN, why did Bonin keep trying to climb when they already discussed that the plane can't maintain altitude if it climb any higher? Look even I (not a pilot yet)in Fs2004 experience my plane not maintaining altitude when I tried to tell the autopilot to climb higher in my 777-300. I was over one hour into my flight form YVR to PEK and I tried to climb to FL360 I think but the plane got up to it and then promptly sank down and tried again over and over. I then realized that if I fly at FL300 the roller coaster ride stopped and was a smooth level flight. I then notice in the FMC, that it had economic cruise altitudes that I should fly. Lesson learned. Bonin had no excuse, to me that was suicidal. He had training in a real plane and the technicalities and I had none but I still figured that I should stop climbing. If he had left the plane level, it could be that none of this would have happened. Make no mistake, I am not Monday Morning Quarterbacking, I know there is stress involved and what not. I believe that one major purpose for training is for you to default to it even in panic and you can't think straight. Ask martial artists for example. They realy don't have time to think of what their opponent will throw next so there training kicks in automatically. The other copilot seemed too passive in the whole ordeal and continued to let Bonin mess things up. Its dark outside should he not be keeping his head in the cockpit? He is not the PIC, in that case could he have noticed that Bonin was still pulling on the stick? Anyway...
There would not be any negative Gs during the descent. Even at 10,000 FPM it would be a steady state drop. The attitude of the nose wouldn't matter either.
I'm flying with Boeing from now on!
I'm flying with Boeing from now on!!

[This poster has been suspended.]

There is certainly a fault with the method Airbus uses to control the aircraft. If there was no fault the accident would not have happened. You cannot sit on the fence here.
The computers on those planes are continually moving fuel around to achieve the best Lift/drag ratio. That makes it very difficult to hand fly the plane. I've hand flown a 727 on a two hour flight and it is not easy (but quite possible) to maintain altitude to within 50 feet. You can feel the cabin crew pushing their trolleys up and down the aisle. But we had infinitely more stick and rudder flying experience (and training) than those Airbus guys, often hand flying the machine all the way to top of climb and from top of descent to landing.
Yes Ralph I accept all you say but the 727 is a different era, almost stone age compared to the new AF stuff. The AF controls divorce the PIC from control. The issue is not with the housekeeping tasks, this issue is with that stick, who controls it and the way it reacts to two people controling it. It is a bit like putting a condom on your willy and both partners not knowing if the damn thing is in yet.
AB rated pilots - Is it true that AB training is to pull back on the stick in the event of a stall warning (in NORMAL law) due to stall protection (especially when near coffin corner)?

(I've seen some summaries that implied this, but no actual confirmation that this is indeed what's taught)
It is clear that Airbus believes itself to be above the laws of physics and the capability of the human brain. Moreover it has decided that the machine will decide the course of action. There is something very seriously wrong with the Airbus ethos. I hope the FFA takes them to task rather than just sit on the fence.
The manufacturer and the safety agencies need to develop a "mandatory" standardised training system for all airlines who purchase a particular model of aircraft. Also the safety agencies have the final say on how these training systems are setup and run.
Yes Kevin..recent accidents have highlighted the flaws in the simulator many times have they told us "the circumstances could not be replicated in the simulator"? Really this is beyond belief...the AF simulators were obviously designed for ex military "stick" pilots who are increasingly rare to find these days.
I was told that's because these planes have never been flight-tested in high-altitude upset scenarios, so there's no data with which to program the sims. Apparently, such flight-testing was deemed to be "too dangerous." Yikes!

Stop the airplane, I want to get off!
Scott: I had heard that before and don't yet know whether truth or talk. I can't believe there was certification with out it BUT, I guess that depends on the certifying authority.

BTW, question I asked you further back up and I think you missed. Once in ALTERNATE LAW, how big a deal is it to go back to NORMAL LAW, and there have been several comments here about just letting everthing go and the AC taking back hold of itself???????? truth or fiction?????
I did try to reply earlier....someplace up there. My dim memory after reading thousands of posts, but I'm not positive about this.....once the airplane leaves Normal Law, it cannot be reset back until it lands!
Surely, there are AB pilots here who could verify this.
10-4, Tks. Wayne
jbermo 0
Like it or not, automation will not go away. The likes of AF 447 and Colgan have practically guaranteed that full automation will come. No pilot, no pilot error, no pilot training, no pilot duty limits, no pilot layover expense, no pilot labor issues.

The proven efficiency's of today's military UAV's are the modern equivalent of the cammels nose under the tent!
WOW... Reading all these comments and opinions reinforces the old addage when humans interact/rely/program any computer. "Garbage In = Garbage Out" How unfortunate for the souls on board!!
Besides being a pilot, I handle IT management as well and all I can say is AMEN!!!!!!!!!!!!
Old news... see this detailed report from July - at
Michael: you are pretty much correct. There was a little more of human feel here what with the pilot conversations and all but not really not anything new as to what happened. This article has generated 121 comments so far and the others have had double that. This gave a little more insight to the total confusion that existed, loss of situational awareness and drilled down on the complexty of AB systems, but the bottom line, it stll crashed and all are dead, and the courts, insurance companies and lawyers now have it all, so in that respect, it will stay a blame game, rather than full truth coming out. This just helps lay it on the Pilot's and if accurate, probably should be. Big Question is what brought them to that state of panic/confusion.
jc1742 0
A good reason that technology will never replace good old fashioned hand flying. I hope all airlines take heed to the lessons to be learned from this tragic event.
I will have to disagree, A/C crashes occurred much more often a long time ago. With automation, not that it should substitute a pilot, but has done a lot of work for pilots, making it easier, and safer to fly. Look at the numbers, how many airline crashes have there been in the past few years?
Granted, as automation has come on it has made it safer and much easier for the older generation of pilots that have saw it advance, BUT, those guys had the stick and rudder experience and knew how to FLY THE PLANE. A lot of the newer pilots are without that basic experience and that is where the problems that do emerge get talked to death in these columns. I have said many times before that all we talk about are what makes the news like this one. The media nor anyone else picks up on the hundreds of upset notes in a pilot's trip log that are buried in Airline's files somewhere, where the pilot does their job, recovers and goes on.
Phenomenal article. Was interesting to see how close the NOVA documentary came to determining the cause(s).

While we can sit around and discuss human vs. machine error, design flaws, training, psychology as the main cause, accidents like these happen due to a series of events that linked together (and not corrected) lead to tragedy.

While it would be great to say that all of the above causes in the crash of AF447 will be addressed and corrected, the bottom line is that commercial aviation is a business. Any "fixes" will be the ones that impact profits and ROI the least, postponing but not preventing a future accident from occurring.
Blaming Airbus for design shortcomings in this incident is bizarre. No aircraft design is perfect. All aircraft have their quirks and vices. One can argue at length on the pro's and cons of independent control sticks or "over automation" or whatever, but the bottom line is that the aircraft is designed like that and it is up to the pilots to handle it competently. If an engine fails in flight and the crew react incorrectly by, say shutting down the wrong engine, then there is no point in blaming the engines.
In this case the co-pilot (Pilot Flying) reacted completely wrongly when the autopilot disconnected (something that the automatics are quite correctly designed to when airspeed input fails), by pulling the stick back and holding it back for the entire scenario. He seemed oblivious to the fact that the flight controls were in alternate law (which was clearly displayed on the instrument panel, - which a competent pilot would have looked for and taken note of!) and that a sustained steep climb from 37000' on any airliner would result in a stall, - no matter what thrust capability the aircraft had. (In fact the TOGA thrust he had set, made matters far worse because of the high thrust on the underslung engines, coupled with the very low aerodynamic controllability at low speed and high altitude would have made it very difficult to get the nose forward even if he had tried) Also, both First Officers seemed unaware that the aircraft had stalled (despite the fact that the aircraft warning system announced very loudly and very clearly 75 times that it was stalled!!!). Furthermore, I am unaware of any commentators mentioning the fact that BOTH control sticks are fitted with an override button that would instantaneously disconnect the opposite side stick inputs if pushed. If the pilot on the left had used that he would have had FULL control to attempt to do what it appears he had a vague idea he should have been doing i.e. PUSH THE NOSE DOWN AND GET FLYING SPEED. The appalling truth is that if both pilots had simply taken their hands off the controls and closed the thrust, the aircraft would have in all likelihood recovered on its own!! There are other contributing factors like the radar being set incorrectly (which started off the whole chain of events), which is a common error on the Airbuses but again for which a competent pilot would be particularly watchful, and the uncertain chain of command, but I think my point is made.
No-one seems to want to say it, but I'm afraid it is true, that this is a case of abominable incompetence, nonexistent airmanship and atrocious training - very little to do with the aircraft. Air France should be hanging its head in shame.
Hilton you cannot be serious...I agree that there are some human errors here but hell, if the right hand seat cannot see or feel the main control input how can you claim there is no flaw. Have I missed something here?
Hi Roland. In a scary situation like the one we are talking about, it is not necessary to look to see if the opposite pilot is stirring porridge with the stick or not - simply press the override button and you have full control.
My main point is that most aircraft have areas were they could be improved upon, but competent crew are aware of those and are especially careful to avoid being taken down by them. This crew appears to have been clueless and seem to have committed the most basic of errors and now the aircraft is being blamed.
I disagree. Since the young feller was left as PIC, there are big cultural obstacles, especially if one is French, to taking over without having a concrete reason. Without solid evidence of PF's inappropriate flight inputs, Robert deferred to the less experienced pilot. For more info on this cultural problem, see Malcolm Gladwell's superb book, "Outliers." <>
Of course, in hindsight, that was a tragic attitude to have, nevertheless, it was a factor. Numerous previous air tragedies have been linked to this dynamic.
I posted similar comments months ago and there were quite a few noses outta joint here, full of conspiracy theories about weather, aircraft systems, AF management, etc. In the end, they took a flawed but flyable aircraft from FL350 to the bottom of the ocean. In military flying, the guy that screws up pays the price. Sadly here passengers also paid the price.
That is just unbelievable. This article really makes this accident seem so avoidable had the crew just had the slightest bit of communication.
1. Airbus and Boeing both make great airplanes.
2. Hindsight is 20/20.
3. I wasn't in the there.
I can remember from my USAF days that there used to be a psychiatric exam,FSM 35 or something like that, tailored to the clearance or job a person was going into, that was to specifically try and determine a person's reaction to adverse conditions or high stress levels. I don't think this is done by any airline. It, or something like it may have prevented this type situation by catchg a reaction beforehand. As some said, hindsight is 20-20 and we weren't there. Every aircraft has their quirks but you know what you have when you get in it, regardless of the brand. It does seem that AB has some different things that a lot of us would be hard pressed to get used to, BUT, that doesn't change the fact that for whatever reason there was total panic and confusion in the cockpit. All we will hear from now on is what AB, the insurance companies and lawyers want us to hear. Fact is, plane crashed and all are dead, and all we can do is wonder why. They now have it hung squarely on the pilots and it will be a battle between those 3 mentioned for the least liability.
too bad the cabin wasn't full of Lawyers
JD345 0
This is on the pilots, based on what I've read there's no doubt of that. They lost control of the situation and that's what brought the plane down.

It's a little more than simply an airplane having quirks, though... it would certainly appear to myself and others that the Airbus philosophy can actually serve as a substitute for basic airmanship. That's a failure of training, of course, but the Airbus systems make it possible to train pilots in such a way, so that when a situation like this arises, they're well-trained in flying the Airbus computer but have very little practical experience flying an actual A332.

This incident is on the pilots and their failure to think clearly and use all of the information at their disposal (for example, STALL STALL STALL STALL STALL STALL STALL STALL STALL STALL). The point on Airbus is that in this specific incident, in, say, a competing 767, the guy in the right seat can't pull back on the yoke while the guy in the left seat pushes the yoke forward. The controls of a 767 rely on the competence of the pilot instead of the competence of the computer to prevent a stall. The controls of a 767 don't operate under different rules in different situations.

These pilots probably would have eventually found a way to crash a 767 or anything else that day, too, and again, in the 6 or 7 minutes that this incident lasted, at any point up until the last minute or so any of the three ATPs sitting in the cockpit could have easily gotten out of the situation. But it appears that the lack of awareness from the pilots couldn't have lasted as long as it did in another aircraft... if the Airbus systems didn't contribute directly to the crash, I think they did at least fail to mitigate the problem where they could have... particularly with the dueling sidesticks, something physically impossible on most other airplanes.
Jack: I had a reply do that to me the other day too, so don't feel bad. I can agree with your comments on the AB system, yet also agree that it's over complicated. Being typed in a 767 among others, I also agree that these guys could have crashed anything, but I do believe, though not intentional, that the AB system led to most of their confusion and panic. At the contols of a Boeing, you have no doubt what you are in. That being said, if you are flying an AB, you are supposed to know all that crap about it and be able to handle it.
JD345 0
It seems if you post the comment without logging in first, it does save what you typed when it prompts you to log in, but it doesn't post it under the guy you were intending to reply to. I guess that's what I get for expecting computer software to do my job for me :)
JD345 0
what? This is supposed to be replying to Hilton Davidson a few posts above. Two reply fails, sorry.
Unbelievable ..... A plane so idiot proof dented by a frog idiot!
lets hope it has info
Wow... just... wow...
Look- how hard can it be.

Design an airplane with the Pilot as an integral part of the Flight Management System and train him like we all trained for the past 50- 60 years. The Human Brain is light years more complex and capable than the best the Computer Engineers can come up with.

I soloed in 1959 and, retiring in 2008, have over 30,000 hours. With each generation of Pilots I have noticed more and more reliance on the Computer and less and less on Basic Airmanship. As in this case, when the checklist doesn't cover it, put your hands over your face and die.

Airline Training is a complete farce. The Feds say you will do A,B,C as quickly as possible and move on. No curve balls- there isn't time. We are trained to handle predictable emergencies using predictable checklists. Nothing else can possibly happen because it isn't authorized. Sullenberger doesn't exist.

Can't you see a bunch of old guys up in the clouds, dressed in leather or old uniforms, shaking their heads..... watching our Profession become a Trade?
Well, the "old guy", left the cockpit to take a nap during a crucial part of flight when the plane was entering an active area of thunderstorms. Really the first question that should be asked is why the plane did not plan a deviation, and put themselves in that situation. Why wasn't the Captain there? A combination of things led to the accident, I think its more complex than simply stating that younger pilots don't have the feel for the plane. But there seems to be no simple answer to this. Weather, I can't imagine the turbulence, look at the satellite, wind shear, low visibility, stress and overall little experience and miscommunication.
They had no fuel to deviate. The flight plan was cooked.
I had some super competent F/O's- probably some better than me- but I wouldn't have been in the back in the ITCZ.

Not at all a question of young vs old. Just that the younger Pilots have been trained to believe that the Computer is infallible, that smart snappy button pressing is competent Piloting, that if there isn't covered by a Checklist it can't happen. This is just one more confirmation that Murphy isn't even out of bed yet. I "know my procedures" and "can recite my checklists".....

Oh, well- my opinion is clear. Just remember- if nothing goes really wrong the 'Bus is as safe as your bed.
When I learned to fly the first thing the CFI made VERY clear to me was, "when I say MY AIRCRAFT you let go of everything and get your feet of the rudder controls, How is it that nothing like that is ever said, MY AIRCRAFT buddy, let go of everything, stabilize and then turn the auto back on.
I think it will take an AB type rated pilot to answer, but somebody here sad that you couldn't go back to NORMAL LAW until after landing. The AP came off, which took it into ALTERNATE LAW. If the law can't be reset, I don't know if the AP can or not.
Blaming Airbus design shortcomings for this incident is pointless. No aircraft design is perfect. All have their foibles, quirks and vices. One can argue at length on the pro's and cons of independent control sticks or "over automation" or whatever, but the bottom line is that an aircraft is as is and it is up to the pilots to handle it competently. If an engine fails in flight and the crew react incorrectly by, say shutting down the wrong engine, then there is no point in blaming the engines!
In this case the co-pilot (Pilot Flying) reacted wrongly when the autopilot disconnected (something that the automatics are quite correctly designed to do when airspeed input becomes unreliable), by pulling the stick back and holding it back for the entire scenario. He seemed oblivious to the fact that the flight controls were in alternate law (which was clearly displayed on the instrument panel, - and which a competent pilot would have looked for and taken note), and that a sustained steep climb from 37000' in any airliner (in unprotected control law) would result in a stall, - no matter what thrust capability the aircraft had. (In fact the TOGA thrust he had set, made matters far worse because of the high thrust on the underslung engines, coupled with the very low aerodynamic controllability at low speed and high altitude would have made it very difficult to get the nose forward even if he had tried) Also, both First Officers seemed unaware that the aircraft had indeed stalled (despite the fact that the aircraft warning system announced very loudly and very clearly 75 times that was the case). Furthermore, I am unaware of any commentators mentioning that BOTH control sticks are fitted with an override button that would instantaneously disconnect the opposite side stick inputs if pressed. If the pilot on the left had used that he would have had FULL control to attempt to do what it appears he had a vague idea he should have been doing i.e. PUSH THE NOSE DOWN AND GET FLYING SPEED. The shocking truth is that if both pilots had simply taken their hands off the controls and closed the thrust, the aircraft would, in all likelihood have recovered on its own!! There are other contributing factors like the radar being set incorrectly (which started off the whole chain of events), which is a common error on the Airbuses but again for which a competent pilot would be particularly watchful, and the uncertain chain of command and who was flying, but I think my point is made.
No-one seems to want to say it, but I'm afraid it is true, that this is a case of incompetence, poor airmanship and very bad training. All very little to do with the aircraft. Air France needs a thorough overhaul of its training procedures.
Finally it needs to be said that nothing in the above should be taken to mean that I believe that I don't make mistakes when I fly. In fact I make them all the time. That is part of the reason we have multi-crew aircraft, CRM, standard operating procedures and recurrent training. However. if (God forbid) I ever do make a major untrapped pilot error, then I fully accept that that is where corrective action is needed, not with blaming the aircraft.
I have no connection with Airbus Industry at all. Other than the fact that roughly 50% of my 23,000 hours flying experience was accumulated on Airbus aircraft (and roughly the other 50% on Boeing).
Oops. Sorry guys. It seems I have posted my comments twice.
I have learned a lot from these great discussions. My Dad was a USAF pilot & flight instructor. We in family are all lifelong pilots; but, of light aircraft.

When I go for my annual recurrency training, my elderly instructor always starts with basic flying and aircraft control, even occassionally covering the entire instrument panel. We do slow flight, all sorts of stalls, controlled banked turns, single engine, etc; BEFORE any instrument training; despite thousands of hours logged.

Don't airline pilots do the same things ?

Did I understand correctly that the PIC of an AirBus cannot over-ride the computers and take manual control of the aircraft ? If so, this scares the hell outta me !

Well, Rick, I am boeing and RJ typed in multiple AC's but not AB. I have heard that same thing about the AB system. I don't believe it is a wholesale thing, just partly in one of their flight laws. If it is so, it scares the hell out of me too, but some AB type rated pilot is going to have to get on here and explain it. To date, and I may have missed one, but in every story about 447 since it went down, there have only been one or 2 that have admitted to flying a Bus and they have said nothing about the Laws. Now, I don't know if they are ashamed to admit to fling a Bus or if it just too complicated to explain. If that is it, then it definitely needs changing. You don't have that worry nor all them laws to fool with on a Boeing, (well the computer is working when the AP is
I ask becaues I find myself going airlines to Europe more frequently in recent years and have always preferred AirFrance, despite my previous comments.
Last two trips have been on US carriers and both used AirBusses.
Humans have had to learn the same lessons over and over; looks like this is one of those deals.
Regardless of the brand you are flying, the Pilot(s) up front are supposed to be type rated on that AC and by that they are supposed to know all the features of that AC in front of them. In that light, if you fly as a PAX, it shouldn't make any difference on what AC you are on.
Wayne, I don't to be in ANY aircraft of which the PIC does not have full control when he or she needs it.
Wayne may I ask you a professional question. If you were captain on a long flight with less than 40 minute fuel reserve would you log a false destination. This AF flight was logged as X to Y where Y was 150km short of the actual destination Z.?
That just now seems to have surfaced and I don't really know what to make of it. That seems strange as this was a scheduled flight. I never had to put up with that. Though flying big iron, it was corporate so I was responsible for my own fueling and unless it was a short hop and I didn't want to land real heavy, most time I had full fuel. I have only seen this partial, computer generated fueling since I retired and started doing commercial fill in with AA and Delta. Captain always has the right to override that if he sees weather up or something but most don't.
Thank you that is the reply I was hoping for. This particular route with AF and this machine had the destination as BREST. This was the way it had to be to keep within the rules. 98% of the time they made it to Paris CDG but there was little left to account for all the variables. The claim I am making is this: The crews on this schedule knew they had little option for deviations. Lisbon and S Maria were possible defaults as was Brest and this route was very profitable for AF hence the pressure from management. I do not enjoy this exposit of what went down but it would be unfair to continue this discussion without this fact.
That is a very common and safe procedure, Proposed Redispatch. As ya'll know, the 10% in time fuel pad for overwater flights goes back decades and is ultraconservtive in todays' environment. All you do is Dispatch to an Intermediate point along the Route for which you can meet the fuel requirement. At a designated point prior to your Dispatch Destination, if you and the Dispatcher agree that you can now meet the fuel requirements to proceed to your actual destination, you do so. If not, you land at the Dispatch Destination and buy more fuel. Done it hundreds of times with a 100% success rate. As they say, it ain't Rocket Surgery.
Herb..Yes I see. But I cannot see how that crew would not divert around that storm and I cannot get the idea out of my mind the reason they did not divert was because of fuel. They would need to land at Lisbon.
I understand the concern, Roland, but the PRD doesn't cut fuel so short that a minor enroute weather diversion would cause a problem. What I- and I think most International Captains- did is look at the weather, traffic situation, distance from a good alternate (not always the one filed),etc., at the destination and decide how much fuel I wanted to have at Top of Descent. That became my planning number and to heck with what the FAA required, which was always less. Pushing to FAA minimums can- and has- caused fuel starvation flameouts on landing roll.

Sometimes, when weather closes a field for an extended time, it is more prudent to land short of the destination and wait it out on the ground rather than getting into the holding furball. When the weather gets better, hop on over- fat on newly loaded fuel.
Finally understood.We will hopefully learn from others.
@Jack DeMarre: You really nailed it with those automotive-based analogies. Without having practiced the basics, no operator of any machine can be expected to recover adequately -- or even timely recognize the problem -- when the machine's controls go haywire.
Just to add a little more spice to the question of why Airbus aircraft behave as they do, here's the still unresolved case of the computer taking over the flight controls:

On 7 October 2008, an Airbus A330-303 aircraft, registered VH-QPA and operated as Qantas flight 72, departed Singapore on a scheduled passenger transport service to Perth, Western Australia. While the aircraft was in cruise at 37,000 ft, one of the aircraft's three air data inertial reference units (ADIRUs) started outputting intermittent, incorrect values (spikes) on all flight parameters to other aircraft systems. Two minutes later, in response to spikes in angle of attack (AOA) data, the aircraft's flight control primary computers (FCPCs) commanded the aircraft to pitch down. At least 110 of the 303 passengers and nine of the 12 crew members were injured; 12 of the occupants were seriously injured and another 39 received hospital medical treatment.

The Australian Transport Safety Bureau (ATSB) released detailed interim factual reports into this occurrence on 6 March 2009 and 18 November 2009.

On 26 June 2011, the draft final ATSB investigation report into the occurrence was provided to directly involved parties for comment. In accordance with the requirements of the International Civil Aviation Organization's Annex 13 to the Convention on International Civil Aviation, these parties have been allowed 60 days (until 26 August 2011) to review the draft report and provide comments. All submissions will be considered before the report is finalised for public release.

As of 16 September 2011, the ATSB had received comments from all of the directly involved parties. The review of those comments is ongoing, involving additional liaison with a number of international stakeholders. As part of that liaison, the aircraft manufacturer requested until 21 October to provide additional information.

It is anticipated that the final report will be released to the public in early November 2011.
Thanks for that information Geoff. The ATSB reference number is AO-2008-070. The ATSB is, in my view, the best authority for hunting down the facts free of any commercial or political interests. I see that the ATSB is awaiting data from Airbus and the expected conclusion date is running over.
Once again, I despair at the standards of training of Air France flight crews. It would appear they do not know how to prevent a stall, (as also shown with the flight engineer shutting down an engine on the ill-fated Concorde, without obtaining the Captain's permission). How many more of their aircraft will fall out of the sky before they remedy their faults? Back to basics with the training please!!
This is absolutely shocking! I'm surprised how did they even managed to get their licences in the first place!
Also if the alarms were going off, and they had didn't have a clue about what was going on, then why not go and wake up the Captain??!!! I think that's the least they could've done...
This is absolutely shocking! I mean how did even managed to get hold of their licences in the first place?!
Also if the alarms were going off and they didn't have a clue about what was going on, then the least they could've done is woke the Captain up!
Not as simple as that Harshad. Both instruments to measure forward velocity were out of service. The little control sticks were one factor and both pilots were giving inputs to them. Boeing has handle bar yokes on both sides so both pilots have visual and tactile oversight as to what is going on. In my view the flight plan adopted by the crew was faulty, they should have made a wide detour around the storm. I agree that many of us are shocked at the consistent abuse of the conrols, but these big airliners are not as simple as our first trainers were, and Airbus were already concerned that pilots should "revisit basic flying skills". I agree though that there is something seriously wrong with the AF CRM and culture.
Good points on the linked yokes vs side stick controllers. I think most of us are nostalgic for our stick- in- the- right- hand- throttle- in- the- left days, but I have to admit that, for grownup flying, the yoke is far more practical.

You mention "both instruments to measure forward velocity".... I assume the 'Bus provides some kind of FMS generated groundspeed readout, so they actually had several. As an Astute Aviator on this thread mentioned, 80%, wings level, nose about 2-3 (whatever it is on the 'Bus)nose up. Keep the GS between 400- 450 kts where they were, use what you are SURE is good info, and you can play WTF until you run out of gas.
Hiya Herb..I know for a fact AF had some scares with faulty pitot tubes (two) and AF were giving them the brush off. This was ongoing when that particular plane crashed. As usual AB were dithering as AF became increasingly annoyed. When both pitot tubes froze up I guess the FMS disconnected the autopilot, prolly gave off a few audibles, but I am not sure. This route is not for the inexperienced or faint hearted.
correction: Airbus were not attending to the faults AF had found with the pitot tubes...
The South Atlantic definitely is not the place for the Newbie.... only place I've ever flown where the Charts were marked "No FIR" (GIG- JNB back in my Pan Am days). We gave HF reports to Comapany in Rio halfway and then to Windhoek the rest of the way. Don't know why.... guess they could launch the dugouts if we ditched. Just fly Quadrantals and, with all that sky and almost no airplanes, hope for the best!
Total number of documented UASes ascribed to those pitots, 37 (all carriers). The rest of them didn't crash.
"Cognitive overload" is increasingly a factor in airline accidents and incidents. It certainly added to the pilot mistakes in AF447, but often happens closer to the ground. As Roland Dent observed the ATSB is a highly professional organisation, and Australian aviation is generally among the safest in the world. Nevertheless the two intersect more frequently than the public recognises. In the last few days two ATSB reports on incidents where flight deck confusion due to incorrect configuration for landing could have had serious outcomes emphasise the importance of CRM. In the Jetstar incident in Melbourne the FO/PF had only 1966 total hours and only 300 on the A320. Jetstar (the low cost subsidiary of Qantas) trains its new pilots via a cadet scheme approaching the MPL system - which raises that issue. In conjunction with poor CRM at times of cognitive overload......? Here are the links to the ATSB reports: (Jetstar A320 Melbourne) (Qantas Boeing 767 Newcastle)
Excellent point about Cognitive Overload, Geoffrey- let's define it.....

I'd say that Cognitive Overload occurs when you are presented with more stimuli than you are trained to handle.

After Airline Retirement I flew Falcon 2000's- a really wonderful FRENCH Airplane- for some years. Flew a lot with a retired AAL guy- between us we had about 50,000 hours. I won't say that we were never surprised, but it never suprised us to be surprised. You have to expect these supposedly inantimate Airplanes to sink their incisors into your butt at the most inappropiate times in the most inopportune ways.

When I soloed 52 years ago I suffered Cognitive Overload on final trying to simultaneously hold airspeed, glide path, and track in a 150- and then I had to flare!

Fast forward to my first airline job- Plumber on a 727 after I had been an AC and IP on C-130's in USAF. It was five months after reporting before I flew my first "Solo" Line Trip- five months of Ground School, Simulator (when we got one)and Aircraft Training. After that it was a week every six months and- one day a month- we came in for Ground School on one system in the Morning and an Oral on another in the afternoon. And this was on the 72'- a pretty straightforward Airplane.

Yep, all that training was expensive. Then- not only AB but Boeing- put out the Sales Pitch that they could build an airplane that might cost a lot, but it could be safely flown by complete idiots with minimal training. The Bean Counters were awed. The Pitch was based on the premise that the manufacturer could predict everything that could possibly go wrong and provide a laminated checklist to get the Crew through it. No need to spend expensive training time going over valves and relays- if Light "A" comes on flip switch "B.

So here we are. Devilishly Complex machines, maybe three weeks to get the Rating but "don't worry- the airplane will take care of you."

And when it doesn't, you have 447.

The Training Departments? Most of the Training Captains want to be Fleet Captains and the Fleet Captains want to be VP Ops- how much money can they save the Company?
polska55 0
Air France did not get the plane's pitot's fixed/replaced as Airbus recommended, correct?
Airbus could not recommend because they were not sure themselves. Not an AF fault. Airbus were slow in responding and AF was annoyed.
spudtu 0
Not a pilot, but have studied aviation since I was a teenager. I'd be interested to know what some of you *real* pilots think of my take on the AB systems and this crash.

Obviously, the young pilot in the right-hand seat (PIC?) made an inexplicable and grievous error by pulling the stick back almost immediately after the AP disengaged. And then nailed it by never bringing it back to straight and level, but keeping back pressure on the stick except for one brief moment near the end.

So right there you could say the crash was due to human error, period. As others have said, absent this error these guys would have muddled through okay. Maybe I'm missing something.

But in more traditional AC systems, various posters have also noted that there are two or three features that would have probably caught this poor rookie's error.

Most notably, that in (almost?) every old-timey AC, including modern Boeings, the left and right yokes are linked and move simultaneously, which in this case would have allowed the left-hand pilot to notice what a crazy thing the other guy was doing with the elevators.

So in moving to a "smarter" flight deck system, AB actually eliminated at least one cross-check on human error. Almost as if the designers of the AB made the very glib assumption that we don't have to worry as much about human error because the computer is so good.

This is real disturbing, especially with the eventual trend towards pilotless flight decks. Even if those computer programmers are themselves pilots, they can't possibly think up every scenario. With a real human next to the throttles, at least there's a fair shot that the plane will make it home, even from that unanticipated scenario.
Spud, Pilot Error is a very convenient way for everyone to explain accidents. That way nobody gets sued.

I guess one could contend that, in any accident that could have been prevented by any action the Crew could have taken, Pilot Error was the cause. That's almost all of them.

These guys obviously weren't trained for this type of scenario. As I mentioned before, I never flew the 'Bus but even on the glass cockpit Boeings the training emphasis was definiely on How to Serve the FMS rather than how to fly the airplane. I think this might be a holdover from the early days of the computerized airplanes when the guys checking out on them knew how to fly but not how to use the new systems.

These F/O's were hardly Rookies. They were in their 30's with 3000+ hours. We have guys in their early twenties with under 500 hours flying Supersonic Fighters in dozens of countries. I was an Aircraft Commander flying C-130's all over the world at 26 with 1200 hours. The Training is what matters. As they say, some guys have 10,000 hours and others have the same 1000 hours ten times.

I completely agree with you about the idiocy of trying to design the Pilot out of the Control Loop. Not only does the Pilot have access to the information the FMS uses- pressures,temperatures,RPM's, etc.- but he can hear, feel, smell. The Magic Box should provide the Pilot with information and the Pilot should be trained on how to interpret and use this info. If the Pilot has to overstress the airplane to keep from hitting something, he needs to be able to do it. If the airplane is so bent that it has to be scrapped, the Junk Dealer is happy- but not as happy as the Crew and Pax.
spudtu 0
Herb, thanks for your perspective on that. Probably calling the F/O flying
that 447 a "rookie" was careless of me. And I see your point that "pilot error" would cover a whole lot of crashes, even ones where equipment failed catastrophically.

Your comment earlier about the factories' initial Pitch to Eliminate the Training (and by consequence, The Thinking) was jarring. But have they even eliminated making a mantra of pushing the yoke forward when the airspeed drops drastically? I mean drilling it in so that it's reflexive, unless low altitude prohibits? Watered down that axiom to the point where the instinct to TOGA was stronger? Or was the classic response to a stall never drilled into young pilots that intensely?

I'd have to guess the answer is no, that pilots are still learning the instinct, and not just the theory, to nose-down in a stall, otherwise there would be a lot more 447s in the past decade or so. But it does make it harder to understand the F/O's response on the 447.

How did the airlines sell a gutting of training programs to the FAA? Seems like another ball dropped, but then that shouldn't surprise, right? Considering how long it seems to take the agency to put NTSB reccommendations into practice in other cases.
skorki99 0
Scrap all Airbuses with the side stick and Buy BOEING with a control column !!!
Somehow, thought of similar causes but did not want to name it.. However,Shocked!.. just Shocked!......


アカウントをお持ちではありませんか? 今すぐ登録(無料)!機能やフライトアラート、その他様々な設定をカスタマイズできます!