AF447: Human Factors

Kaiarahi · Sat Sep 17, 2011 9:03 am

I tried to post this on the "AF447: Lessons Learned" thread, but the mods have already archived it, so the discussion can't be continued there.

Thanks to canoecarrier for noticing that the BEA has announced the composition of the Human Factors working group mentioned in the third interim report. From the BEA website:

"This working group’s objective is to analyze all aspects connected to the conduct of the flight:

* Crew actions and reactions during the last three phases of the flight described in the third Interim Report, in particular in relation to the stall warning;
* Cockpit ergonomics;
* Man-machine interfaces.

This working group is made up of seven experts:

* Three BEA investigators specializing in human factors;
* A psychiatrist specializing in risk analysis;
* A human factors aviation consultant;
* A type-rated A330 pilot;
* An A330 test pilot."

Work is expected to be completed by the end of December. The BEA note also indicates that the final report will be published in the first half of 2012.

I'm not going to speculate on the Working Group's lines of inquiry, but to me it's interesting that it includes a psychiatrist (médecin-psychiatre) - not a psychologist - specializing in risk analysis.

rcair1 · Sat Sep 17, 2011 4:53 pm

Quoting Kaiarahi (Thread starter):
"This working group’s objective is to analyze all aspects connected to the conduct of the flight:

* Crew actions and reactions during the last three phases of the flight described in the third Interim Report, in particular in relation to the stall warning;
* Cockpit ergonomics;
* Man-machine interfaces.

I think this is a very positive development. We need to understand why 3 experienced pilots did not recognize and recover from the unusual attitude.

JoeCanuck · Sat Sep 17, 2011 7:18 pm

Quoting rcair1 (Reply 1):

Definitely a positive move. This accident could so easily been brushed aside as just another "pilot error" incident, and you hit the nail on the head; the failure to recognize the stall is the real mystery. Finding ways to help pilots maintain spacial awareness clearly, quickly and in a way that gives them confidence that what they are seeing is accurate will be the real trick.

Just another warning light or buzzer or warning won't do the job...especially if the pilots are already overloaded.

That they are drawing from professionals with a variety of expertise for their working group is also encouraging. An effective, reliable solution will probably take some outside of the box thinking.

canoecarrier · Sat Sep 17, 2011 7:33 pm

Quoting JoeCanuck (Reply 2):
That they are drawing from professionals with a variety of expertise for their working group is also encouraging. An effective, reliable solution will probably take some outside of the box thinking.

I mentioned this privately to Kaiarahi earlier. In my opinion BEA has done some very noble work related to increasing aviation safety since this aircraft went missing. Before they'd found the wreckage, at a time when a number of people here on A.net were screaming that FDR/CVR data should be transmitted real time they had already put together a working group to study the feasibility of triggered transmission of flight data.

http://www.bea.aero/en/enquetes/flig...ed.transmission.of.flight.data.pdf

As far back as 2009 they were studying, in the immediate aftermath of this crash, how to localize an FDR/CVR when a plane goes missing outside of radar coverage.

http://www.bea.aero/en/enquetes/flig...ery.working.group.final.report.pdf

And, here we are again, BEA is well ahead of us. I suggest they have been all along and have revealed themselves as a very professional accident investigation organization. We only needed to scroll a little further down the AF 447 page to see the technical documents beyond the press releases.

tdscanuck · Sun Sep 18, 2011 4:26 am

Quoting Kaiarahi (Thread starter):
to me it's interesting that it includes a psychiatrist (médecin-psychiatre) - not a psychologist - specializing in risk analysis.

This may be a translation thing...do psychiatrist/psychologist mean the same thing in France as they do in North America?

Tom.

joelyboy911 · Sun Sep 18, 2011 4:43 am

Quote:
médecin-psychiatre

While I have not encountered this exact term in my French study before, médecin is a doctor, so médecin-psychiatre is a psychiatric doctor in English, whereas the French term for a psychologist (scientist or clinician in the field of psychology) is psychologue.

I don't think psychologist and psychiatrists are ever the same thing (not being from North America, I don't know about there) - as I understand it a psychiatrist is a doctor, and thus able to prescribe medication, but that's going off on a pointless tangent.

At any rate, I'm sure the psychiatrist involved in this study will be sufficiently expert. I'm certainly glad the human factors will be analysed in-depth from this accident.

tguman · Sun Sep 18, 2011 7:29 am

My own understanding of the physciatrist-psychologist differentiation is that a physciatrist is a psychologist that can perscribe medication and is a doctor of medicine not just a Ph. D. that is just my understanding and memory from 10th grade psychology.

As a pilot, I look forward to this report.

TGUman

Ruscoe · Sun Sep 18, 2011 7:39 am

The inclusion of a psychiatrist indicates to me that they are wanting to exclude mental illness. I think this is very unlikely, and the chances of having three with a mental illness at the same time is remote. If it was just for why normal people do particular things, then a psychologist may have been more appropriate.
Having said that the particular psychiatrist may have a special interest in this area.
Why not have them both?.

Cheers

Ruscoe

Pu752 · Sun Sep 18, 2011 7:51 am

We are talking about human factors in a plane crash investigation, when there are many unknown facts in the biology of the humans studies. No clear respons will ever come out.

Aquila3 · Sun Sep 18, 2011 8:10 am

Quoting canoecarrier (Reply 3):
at a time when a number of people here on A.net were screaming that FDR/CVR data should be transmitted real time they had already put together a working group to study the feasibility of triggered transmission of flight data.

BTW I was one of them, just I was not screaming, but politely put my opinion, that is still that way. "Black Boxes" are a relic of the past and should be replaced/integrated with something more up to date. When at that time I have encountered the fierce opposition of some eminent and very respected members of this fourm I just decided to shut up, keeping my opinion for me. But this is OT.

Then, returning to the human factors, it is not yet clear to me if the AF procedures for PF relieving are still in place as they were in that tragic night or they are now updated. The fact that the less experienced pilot relieves the Captain on a seat were he is not trained for had attracted some critics even from inside the AF environment. What about a complete double crew (two CPs and two FOs or anyway two trained left and two trained right)? I believe some American carriers do like that for such long-haul flights. I understand that this is a big cost issue for AF, but for such long flights it should be feasible. The alternative is to train at least one of the FO for left seat, this woulbe only a matter of anticipating something tha he likely will have to do for the future.

Kaiarahi · Sun Sep 18, 2011 11:16 am

Quoting tdscanuck (Reply 4):
This may be a translation thing...do psychiatrist/psychologist mean the same thing in France as they do in North America?

Yes.

Quoting joelyboy911 (Reply 5):
I don't think psychologist and psychiatrists are ever the same thing

Psychiatry is a medical specialty focused on the diagnosis and treatment of mental disorders and pathologies. Psychology is the science of behavioural and mental processes. Obviously there's cross-over, but psychologists focus on why individuals or groups do things or think in a certain way (behavioural / cognitive science), while psychiatrists focus on mental disorder.

I find the inclusion of a psychiatrist "specializing in risk analysis" interesting. Are they looking at the risk that people with certain behavioural or cognitive *disorders* will act/react in a certain way? Are they looking at the risk that flight crew intake profiling does not disclose certain relevant *disorders*? All speculation, obviously, but the inclusion of a medical specialist rather than a behavioural/cognitive scientist could be considered a little surprising.

Quoting JoeCanuck (Reply 2):

Quoting canoecarrier (Reply 3):

Quoting rcair1 (Reply 1):

canoecarrier · Mon Sep 19, 2011 1:25 am

Quoting Aquila3 (Reply 9):
BTW I was one of them, just I was not screaming, but politely put my opinion, that is still that way. "Black Boxes" are a relic of the past and should be replaced/integrated with something more up to date.

You should read their conclusions then. They looked at everything from real/full time transmittal of FDR/CVR data to what it would take to transmit that data only in an emergency. It's been a few months since I read the report, but I did find it not only comprehensive, but it answered most of the arguments that were being made here on a.net at the time.

Quoting Kaiarahi (Reply 10):
Are they looking at the risk that people with certain behavioural or cognitive *disorders* will act/react in a certain way?

This particular aspect was brought up in an earlier thread. Specifically, that the PF had the type of personality that AF and other carriers were trying to weed out. I suspect a number of human factors will be studied. The time that the handover to the relief pilot is made and how long the 3 crewmembers should be together before the Captain leaves, crew hiring practices based on personality profiles, and short term makeup of crews based on how they would be expected to work together in an emergency are a few that come to mind.

holzmann · Mon Sep 19, 2011 1:58 am

From what little I've read, my impression is that the stall warning horn was somehow counter-intuitive? The horn would sound when exiting the stall and would stop when entering? Or am I totally off my rocker here? It apparently was not a logical warning giving the stress of the circumstances.

If this is somehow an Airbus "feature" then they should be at fault. Do Boeing horns work the same way? Is Airbus getting a free pass because this is all being debated in European courts? These are my questions.

tguman · Mon Sep 19, 2011 2:30 am

Quoting holzmann (Reply 12):
my impression is that the stall warning horn was somehow counter-intuitive? The horn would sound when exiting the stall and would stop when entering? Or am I totally off my rocker here? It apparently was not a logical warning giving the stress of the circumstances.

Having read the third interim report, my understanding of what was happenning is that because of the initial pitot tube malfunctions, when the airspeed indicator's numbers were below the stall speed the stall horn would stop sounding as the computer believed it was a pitot tube malfunction. As the pilot's recovered from the stall, and the airspeed began to increase, the stall horn would activate as the airspeed would be passing through the stall range.

I do not know how exactly an aerodynamic stall is measured in a large jet like the A330. Do they have a vane or opening on the wing to sense the movement of the centre of pressure as it moves too far forward in the stall?

TGUman

rcair1 · Mon Sep 19, 2011 2:49 am

Quoting tguman (Reply 13):
stall horn would stop sounding as the computer believed it was a pitot tube malfunction.

The stall warning is disabled if the measured airspeed falls below a certain point - otherwise you'd have the stall warning firing as you taxied around and also during the roll.

Quoting tguman (Reply 13):
I do not know how exactly an aerodynamic stall is measured in a large jet like the A330. Do they have a vane or opening on the wing to sense the movement of the centre of pressure as it moves too far forward in the stall?

It - like nearly all commercial airliners - has both pitot tubes which measure airspeed and angle of attack indicators which measure the angle between the relative wind and the wing chord (is it the chord that is the reference?) It's late and I forget

I believe the AB stall warning is actuated by pitot speed only - but I could be wrong - it may be a computed value from both. I don't think A and B are materially different in this regard.

Operation of the stall warning and these devices is covered in detail in other threads.

rwessel · Mon Sep 19, 2011 3:31 am

Quoting rcair1 (Reply 14):
I believe the AB stall warning is actuated by pitot speed only - but I could be wrong - it may be a computed value from both. I don't think A and B are materially different in this regard.

The stall warning is almost wholly dependent on the alpha vanes, and *not* airspeed.

JoeCanuck · Mon Sep 19, 2011 4:18 am

It seems that for some time there was conflicting information coming from the flight instruments. When this happens, pilots are left with the unenviable and sometimes impossible task of having to choose what is true and what is errant information.

What can happen is time is wasted because it has to be a cautious process; a wrong move might make things worse.

Loss of spacial orientation is not uncommon and can effect and, has effected, pilots of any experience or training level...sometimes with tragic consequences.

In spacial orientation loss resulting in a loss of control of the aircraft, in situations where the plane was recovered, was when the plane came out of the clouds or otherwise could see the horizon. Once a pilot can see the ground, knowing your attitude and recovering is virtually automatic.

The current instruments are great navigational tools but sometimes fail at giving pilots the instant information often required to regain spacial orientation.

Giving pilots an easy to use device or technique, as instinctive as seeing the horizon without adding too much to the clutter of bells, buzzers and lights, would in my opinion, go a long way to reducing the dangers of loss of spacial orientation.

I look forward to the findings of the panel.

tdscanuck · Mon Sep 19, 2011 4:43 am

Quoting rcair1 (Reply 14):

It - like nearly all commercial airliners - has both pitot tubes which measure airspeed and angle of attack indicators which measure the angle between the relative wind and the wing chord (is it the chord that is the reference?)

All modern airliners have wing twist so the chord line isn't fixed. The zero is somewhat arbitrary but, for convenience, it's often taken to be the AoA where the wing generates zero lift.

Tom.

Ruscoe · Mon Sep 19, 2011 1:12 pm

Quoting rcair1 (Reply 14):
The stall warning is disabled if the measured airspeed falls below a certain point - otherwise you'd have the stall warning firing as you taxied around and also during the roll

This can easily be handled by landing gear switches, which activate when there is weight on them

Quoting rcair1 (Reply 14):
Operation of the stall warning and these devices is covered in detail in other threads

True, but this is about human factors and in my opinion had the stall warning worked below 60kts, they may well have recovered.

Ruscoe

tdscanuck · Mon Sep 19, 2011 1:53 pm

Quoting Ruscoe (Reply 18):
Quoting rcair1 (Reply 14):
The stall warning is disabled if the measured airspeed falls below a certain point - otherwise you'd have the stall warning firing as you taxied around and also during the roll

This can easily be handled by landing gear switches, which activate when there is weight on them

That doesn't cover the other case, where certain dynamic maneuvers can put your airspeed lower than that without stalling the wing. The point is that you can't use AoA sensors as an input to a stall warning system when you're in flight regimes where the AoA sensors don't work.

Quoting Ruscoe (Reply 18):
True, but this is about human factors and in my opinion had the stall warning worked below 60kts, they may well have recovered.

They ignored/disbelieved the initial stall warning (which was correct) for *50 seconds*...at that point, it's pretty obvious the flight crew has discarded the stall warning horn as a valid source of information. Then proceeded to ignore/disbelieve the altimeter, the attitude indicator, and the VSI for the rest of the flight. Why do you think having one more warning in the mix would make any difference?

Tom.

rcair1 · Mon Sep 19, 2011 3:50 pm

Quoting Ruscoe (Reply 18):
True, but this is about human factors and in my opinion had the stall warning worked below 60kts, they may well have recovered.

Possibly - as I mentioned in earlier posts a long time ago - the fact that the stall warning _came on_ when they started to recover had to be confusing. Certainly in the calmness of our computer discussions - it is obvious they should have 'known' that this could happen - but frankly, I think that is overstating what a pilot will retain. These are very complex systems and system behaviour in the realm they were in are not well understood, because they are, well, in a flight realm they are not supposed to be in.

Quoting tdscanuck (Reply 19):
has discarded the stall warning horn as a valid source of information. Then proceeded to ignore/disbelieve the altimeter, the attitude indicator, and the VSI for the rest of the flight. Why do you think having one more warning in the mix would make any difference?

Agreed. More warnings are not the answer. I don't know what it is - but that is not it.

tguman · Mon Sep 19, 2011 5:42 pm

Quoting rcair1 (Reply 20):
- the fact that the stall warning _came on_ when they started to recover had to be confusing.

Absolutely agree with you there. If the action you are doing is causing a stall warning, you undo that action.

JoeCanuck · Mon Sep 19, 2011 5:53 pm

Quoting tguman (Reply 21):

Absolutely agree with you there. If the action you are doing is causing a stall warning, you undo that action.

One of the problems was the pilots had conflicting information...so what do you believe, the stall warning or information which may be showing something other than a stall?

Pilots are taught right out of the box to rely on their instruments. In training, instruments are generally completely disabled or covered to simulate failure...not active giving inaccurate information.

Having conflicting instruments is significantly more difficult to deal with than instruments which have simply gone dark or otherwise shut down.

At the very least, training may have to be changed to include more practice on how to discern accurate information from faulty.

LTC8K6 · Mon Sep 19, 2011 6:17 pm

Quoting tguman (Reply 21):

Absolutely agree with you there. If the action you are doing is causing a stall warning, you undo that action.

Only to someone unfamiliar with how the stall warning works, which should not be the case here.

bonusonus · Mon Sep 19, 2011 10:31 pm

Quoting LTC8K6 (Reply 23):
Only to someone unfamiliar with how the stall warning works, which should not be the case here

But the FDR showed that whenever the stall horn sounded, the PF reacted to it, usuallywith a nose-up input, which made the warning go away.

tdscanuck · Tue Sep 20, 2011 4:28 am

Quoting bonusonus (Reply 24):
Quoting LTC8K6 (Reply 23):
Only to someone unfamiliar with how the stall warning works, which should not be the case here

But the FDR showed that whenever the stall horn sounded, the PF reacted to it, usuallywith a nose-up input, which made the warning go away.

The PF appears to have developed a serious case of confirmation bias at some time during the event...his actions are only consistent with a firm belief that he wasn't in a stall despite the fact that he was.

Tom.

canoecarrier · Tue Sep 20, 2011 4:53 am

Quoting tdscanuck (Reply 25):
The PF appears to have developed a serious case of confirmation bias at some time during the event...his actions are only consistent with a firm belief that he wasn't in a stall despite the fact that he was.

Well not only that, but he also consistently took command as evidenced by the FDR noting that the priority button was hit by the PNF, followed by a verbal command that the PNF wanted command and only seconds later the PF retook priority of the sidestick.

I'd suspect this is something else the psychiatrist will be looking at as well. It's been brought up a number of times that had just one of this crew been a different person with a different command style or personality there is a better than average chance that this event wouldn't have ended up in the plane hitting the water. The CVR transcript shows a crew that just weren't working together and there wasn't a clear leader.

But, most of us already have accepted that their fates were sealed not long after the PF pulled up after AP disconnect, gained altitude, lost airspeed and protections and by the time the Captain showed back up it would have taken special training to recover. They were losing altitude around 10K feet/min shortly afterwords, recovery from the situation they were in would have taken some time and extreme maneuvers none of them were trained in or prepared for.

Back on topic, there's a reason BEA is setting up this working group. If you haven't, read the CVR transcript. It can be done better.

N14AZ · Tue Sep 20, 2011 7:56 am

Quoting canoecarrier (Reply 26):
Well not only that, but he also consistently took command as evidenced by the FDR noting that the priority button was hit by the PNF, followed by a verbal command that the PNF wanted command and only seconds later the PF retook priority of the sidestick.

I'd suspect this is something else the psychiatrist will be looking at as well. It's been brought up a number of times that had just one of this crew been a different person with a different command style or personality there is a better than average chance that this event wouldn't have ended up in the plane hitting the water. The CVR transcript shows a crew that just weren't working together and there wasn't a clear leader.

That's exactly what I thought when I saw this thread. After reading all the reports and the CVR transcripts one has the feeling that - speaking as a layman - that something in "the chemistry between the crew members" was wrong.

It's hard to believe that the "chemistry between the pilots" can be a contributing factor for the crash of a plane, that had a perfect safety record before but there must be something about it.

tdscanuck · Tue Sep 20, 2011 1:45 pm

Quoting N14AZ (Reply 27):
It's hard to believe that the "chemistry between the pilots" can be a contributing factor for the crash of a plane, that had a perfect safety record before but there must be something about it.

At the time AF447 went down it was the third A330 major incident (second with fatalities).

It's still a fantastically safe aircraft but it's not accurate to say it had a perfect safety record.

Tom.

N14AZ · Tue Sep 20, 2011 2:09 pm

Quoting tdscanuck (Reply 28):
At the time AF447 went down it was the third A330 major incident (second with fatalities).

Are you refering to the accident with the A 330 prototype in Toulouse? I cannot think of any other major incident / accident. So what was the third incident?

liquidair · Tue Sep 20, 2011 2:50 pm

Had someone told me about this accident, and i hadn't read the previous 22 threads on this subject, I would've sworn blind the PF acted intentionally.

Of course, we can be certain he didn't. But the fact is his actions are, for now, inexplicable. And I'm not so sure that psychology alone can explain why he reacted so strongly to the PNF's attempt to take control.

I really hope this team and their study will shed light on the subject. As a passenger, it's a sobering thought that this can happen.

[Edited 2011-09-20 08:04:15]

Viscount724 · Tue Sep 20, 2011 9:01 pm

Quoting N14AZ (Reply 29):
Quoting tdscanuck (Reply 28):
At the time AF447 went down it was the third A330 major incident (second with fatalities).

Are you refering to the accident with the A 330 prototype in Toulouse? I cannot think of any other major incident / accident. So what was the third incident?

At the time AF447 went down it was the second fatal A330 accident, including the early test crash. The third was a year after AF447 -- the 8U (Afriqiyah) A332 landing accident at TIP in May 2010 with 103 fatalities (one survivor).

tdscanuck · Wed Sep 21, 2011 2:49 am

Quoting N14AZ (Reply 29):
Quoting tdscanuck (Reply 28):
At the time AF447 went down it was the third A330 major incident (second with fatalities).

Are you refering to the accident with the A 330 prototype in Toulouse? I cannot think of any other major incident / accident. So what was the third incident?

The dead-stick into the Azores...major incident, no fatalities.

Tom.

Kaiarahi · Wed Sep 21, 2011 2:53 am

Quoting liquidair (Reply 30):
I would've sworn blind the PF acted intentionally.

Of course, we can be certain he didn't.

I understand what you're trying to say, but:

1. It was intentional (i.e. not involuntary).

2. We don't know what he was trying to (intentionally) achieve - e.g. deliberately bring the plane down, responding to confirmation bias (see post 25), convinced that he knew better than the PNF or the instruments (I've had trainees who thought this), etc.

One of the things going through my mind is that this crew displayed significant signs of disfunctionality. For example, why did the captain hand off command to an F/O who had only been type-rated for 5 months and had 800 hours on type, rather than an F/O type-rated for 8+ years with 4,500 hours on type? Why did the captain not direct who was P/F when he arrived back in the cockpit? What was the history of relationships (professional and other) between these crew members - did the captain and the F/O PNF have a history (recent or not) of conflict? Did the PF and PNF have a history of conflict? What was the PF trying to achieve (gorilla in the room, but was there anything in his background/life which would point to a deliberate attempt to terminate the flight - an awful question, but it has happened before)?

liquidair · Wed Sep 21, 2011 2:37 pm

You make an interesting point.

The PF's actions were intentional, yes- but we've all assumed the outcome was accidental.

I personally think it unlikely the PF wanted to create any harm, but I wonder if the Human Factors team is looking to the possibility, however remote, that perhaps there was an element of sabotage.

It is an awful question, you're very right- but I agree it does have to be asked.

What are the prevalent opinions of people on here?

JoeCanuck · Wed Sep 21, 2011 3:55 pm

It's almost as if the PF's actions ended up being directed to shutting off the warnings as opposed to breaking the stall. In other words, if he pulled back and the warning went away, perhaps he considered that problem solved.

That a pull back shut off the stall alarm would have added to the confusion since this action is counter intuitive to what pilots are taught virtually from their first flight.

Quoting tdscanuck (Reply 25):

The PF appears to have developed a serious case of confirmation bias at some time during the event...his actions are only consistent with a firm belief that he wasn't in a stall despite the fact that he was.

This is getting to, what I believe, is the crux of the matter. I believe that ultimately, contradicting instrument readings and contra intuitive reactions to control inputs created enough confusion that the pilots ran out of time trying to trouble shoot what was happening.

They just couldn't be sure of their spacial orientation and they couldn't trust the instruments to tell them. That's pretty close to a no win situation...and it was for them.

canoecarrier · Wed Sep 21, 2011 4:22 pm

Quoting Kaiarahi (Reply 33):
What was the history of relationships (professional and other) between these crew members - did the captain and the F/O PNF have a history (recent or not) of conflict? Did the PF and PNF have a history of conflict?

I've wondered this as well. Most of us can relate to having a coworker that they don't like or trust. Sometimes to the extent that we don't really listen to their opinions. If these men had flown together in the past, I'd expect BEA to interview the Captain or flight attendants that were on the crew with them. Did they spend time together at the crew lounge, drink in the bar together, or sit as far apart from each other in the crew shuttle to the airport?

Quoting Kaiarahi (Reply 33):
What was the PF trying to achieve (gorilla in the room, but was there anything in his background/life which would point to a deliberate attempt to terminate the flight - an awful question, but it has happened before)?

Too many factors point to this not being intentional. If he'd put tape over the pitot tubes when he did his walk around in RIO then I'd buy into that. Waiting around for hundreds of hours of flight time for an infrequent icing event and loss of airspeed to crash a plane seems highly unlikely. But, I'm sure they'll be thorough and at least ask the question before ruling it out.

liquidair · Wed Sep 21, 2011 4:27 pm

Not all actions are premeditated.

But on the whole, I agree with you canoecarrier

tdscanuck · Thu Sep 22, 2011 2:06 am

Quoting JoeCanuck (Reply 35):

It's almost as if the PF's actions ended up being directed to shutting off the warnings as opposed to breaking the stall.

That's exactly what he was doing...he didn't know he was in a stall therefore there is no reason for him to try to break it. If he took an action that triggered a stall warning, then took another action that made the warning go away, AND he thought he wasn't stalled before taking the action, then it all makes sense at the time. That's what confirmation bias is...you accept data that matches your mental model of what's going on and ignore data that doesn't.

Quoting JoeCanuck (Reply 35):
They just couldn't be sure of their spacial orientation and they couldn't trust the instruments to tell them.

Actually, they could (their attitude system worked fine the whole time and, once the initial icing passed, airspeed/altimeter were working fine too). They just didn't know that.

Tom.

JoeCanuck · Thu Sep 22, 2011 2:23 am

Quoting tdscanuck (Reply 38):
Actually, they could (their attitude system worked fine the whole time and, once the initial icing passed, airspeed/altimeter were working fine too). They just didn't know that.

...and once you get contradictory information from your instruments, deciding what is reliable and what isn't has the potential to overwhelm the most capable pilot.

In my mind, it would be better if contradictory reading instruments just shut off altogether than give erroneous information. If push comes to shove, an attitude indicator and power readings are enough to fly the plane...if not navigate...just as long as you are sure you can count on the information they are providing.

Sometimes no information is better than wrong information.

[Edited 2011-09-21 19:30:21]

canoecarrier · Thu Sep 22, 2011 3:31 am

Quoting tdscanuck (Reply 38):

Actually, they could (their attitude system worked fine the whole time and, once the initial icing passed, airspeed/altimeter were working fine too). They just didn't know that.

I seem to remember Pihero saying they had a bright moon at their eight o'clock that night, so they were close to a daylight flight. Not to rehash the many, many discussions of the past, but they'd already passed the weather I thought when this event occurred. They may have had some horizon indication visually. It's possible they may have regained the horizon before they crashed.

That said, if they did it was never mentioned in the CVR transcript released by BEA. Thanks again Tom for your input. Always appreciated.

rcair1 · Thu Sep 22, 2011 3:53 pm

Quoting tdscanuck (Reply 38):
If he took an action that triggered a stall warning, then took another action that made the warning go away, AND he thought he wasn't stalled before taking the action, then it all makes sense at the time. That's what confirmation bias is...you accept data that matches your mental model of what's going on and ignore data that doesn't.

Quoting JoeCanuck (Reply 39):
deciding what is reliable and what isn't has the potential to overwhelm the most capable pilot.

Quoting JoeCanuck (Reply 39):
just shut off altogether than give erroneous information.

In the steam gauges days - I used to carry little shower soap sticker holders that could be placed over the failed instruments to you would remember they were failed in the scan.
In this case, however, I'm not sure there was enough information/time to make that determination.

Much has been said about how the airspeeds became valid - but that is based on data the pilots did not have.

gregarious119 · Thu Sep 22, 2011 5:20 pm

Maybe I'm being too elementary here, please take it easy as I only dream of being in the left seat some day.

Is it too simple to affix a clear ball with a BB or pellet inside to the dash, somewhere visible in the cockpit. It seems like all these guys had was electronic, computerized equipment that may or may not have been failing. In light of their stall condition, would something as simple as a little ball been able to give them an indication that the plane was completely disoriented?

rwessel · Thu Sep 22, 2011 7:41 pm

Quoting gregarious119 (Reply 42):
Maybe I'm being too elementary here, please take it easy as I only dream of being in the left seat some day.

Is it too simple to affix a clear ball with a BB or pellet inside to the dash, somewhere visible in the cockpit. It seems like all these guys had was electronic, computerized equipment that may or may not have been failing. In light of their stall condition, would something as simple as a little ball been able to give them an indication that the plane was completely disoriented?

The ball on the traditional turn-and-bank indicator is just that, although only in one dimension. See:

http://en.wikipedia.org/wiki/Turn_and_bank_indicator

The "ball" is the lower part of the instrument - it's usually a curved tube with a, *ahem*, ball in it.

The problem is that a ball (either a traditional one or a two-dimensional one like you suggested) tells you nothing about your actual attitude, just your acceleration(s) in the horizontal plane. For example, in a properly executed turn, the ball stays in the exact center, even though the aircraft can be banked quite steeply.

In any event, AF447 was in a fairly normal attitude the whole way down. The stall was the combination of the (relatively) normal deck angle, plus a very high rate of descent creating an extreme angle of attack.

Plus the attitude indicator, which requires no external inputs (just the internal gyroscopic stuff), worked perfectly the whole time, and given that the PF was able to keep the wings level (despite the aircraft wanting to drop a wing periodically during the stall), shows that he was trusting it.

Pihero · Thu Sep 22, 2011 8:07 pm

Quoting Kaiarahi (Reply 10):
Psychiatry is a medical specialty focused on the diagnosis and treatment of mental disorders and pathologies. Psychology is the science of behavioural and mental processes. Obviously there's cross-over, but psychologists focus on why individuals or groups do things or think in a certain way (behavioural / cognitive science), while psychiatrists focus on mental disorder.

Yes, and that's where the group's study will concentrate on.

Quoting Kaiarahi (Reply 10):
I find the inclusion of a psychiatrist "specializing in risk analysis" interesting. Are they looking at the risk that people with certain behavioural or cognitive *disorders* will act/react in a certain way? Are they looking at the risk that flight crew intake profiling does not disclose certain relevant *disorders*? All speculation, obviously, but the inclusion of a medical specialist rather than a behavioural/cognitive scientist could be considered a little surprising.

That point towards more of an individual parameter thna group dynamics for the psychiatrist; the rest of the group will be free to study interactiuons and undercurrents.

Quoting tdscanuck (Reply 19):
it's pretty obvious the flight crew has discarded the stall warning horn as a valid source of information. Then proceeded to ignore/disbelieve the altimeter, the attitude indicator, and the VSI for the rest of the flight. Why do you think having one more warning in the mix would make any difference?

A very good point which kills the *add another instrument-and-warning* crowd.

Quoting tdscanuck (Reply 25):

The PF appears to have developed a serious case of confirmation bias at some time during the event...his actions are only consistent with a firm belief that he wasn't in a stall despite the fact that he was.

Let's see the oft spoken *contradicting instruments* :
-The ADI :
1/- pitch up --> IVSI and altimeter values increase, ASI decreases. When run out of energy, IVSI tends to zero, then winds down, leading the altimeter.
2/- pitch down --> ASI increases, IVSI and altimeter decrease.
- The Altimeter will vary at all times with the IVSI indication.
- The ASI : one should look at it as representative of a great partl of the flight total energy 1/2 m.v squared.
- The IVSI normally follows the ADI lead : negative pitch --> IVSI increases.
From the above, the discrepancies came from the incompatible pair : ADI showing 12° (+) of pitch and IVSI at minus 6000 ft/min ( confirmed by the unwinding altimeter).
So, we are left with two pairs of mutually validating instruments :
A/- ADI + ASI : high value of nose up pitch and airspeed zero, indicating a *normal* null value of kinetic energy
B/- Altimeter and IVSI showing a fast descent.
The question is : how fart from diagnosing a stall were they ?

Quoting canoecarrier (Reply 26):
suspect this is something else the psychiatrist will be looking at as well. It's been brought up a number of times that had just one of this crew been a different person with a different command style or personality there is a better than average chance that this event wouldn't have ended up in the plane hitting the water. The CVR transcript shows a crew that just weren't working together and there wasn't a clear leader.

To be perfectly honest that crew was dysfunctional even more so when we consider the loss of a pilot due to the airline : union requirement for left-hand seat controls operation.
That situation hasn't been addressed yet and I hope it won't require some major re-thinking of long-haul SOPS.

tdscanuck · Fri Sep 23, 2011 1:27 pm

Quoting gregarious119 (Reply 42):
Is it too simple to affix a clear ball with a BB or pellet inside to the dash, somewhere visible in the cockpit.

Yes. That would not have provided any information the crew didn't already have and would not have shown them they were in a stall.

Quoting gregarious119 (Reply 42):
It seems like all these guys had was electronic, computerized equipment that may or may not have been failing.

None of the electronics or computers were failing. The air data probes got momentarily screwed up (and correctly announced that fact to the crew) and appeared to have cleared even before the aircraft stalled, let along all the way through the rest of the flight. Electronic instruments also have a massively better reliability record than steam gauges and tend to have better failure modes (they announce that they're dead, rather than just presenting bad data).

Quoting gregarious119 (Reply 42):
In light of their stall condition, would something as simple as a little ball been able to give them an indication that the plane was completely disoriented?

Nope.

Tom.

AirlineCritic · Sat Sep 24, 2011 6:12 am

Quoting liquidair (Reply 34):
The PF's actions were intentional, yes- but we've all assumed the outcome was accidental.

It was accidental. I don't know if it was confirmation bias, inexperience, bad way of communicating among the three, but it was not done on purpose.

Think about it. You'd have to be incredible clever and made of pure evil to be able to act like he did on purpose and not make it obvious to the other two. In theory, it would be possible, but even then he would have had to rely on the other two pilots behaving just like they did on that night. I'd rule this out.

WingedMigrator · Sat Sep 24, 2011 6:33 am

Quoting tdscanuck (Reply 38):
That's what confirmation bias is...you accept data that matches your mental model of what's going on and ignore data that doesn't.

The seat of the pants might also have played the usual unwelcome role. Who could have predicted, least of all the pilots, that an airliner could "fly" at an AOA of 45 degrees and remain mostly controllable? The attitude rates and accelerations apparently never got too crazy, all way down.

Kaiarahi · Sat Sep 24, 2011 11:01 am

Quoting AirlineCritic (Reply 46):
It was accidental. I don't know if it was confirmation bias, inexperience, bad way of communicating among the three, but it was not done on purpose.

You're missing the point. As far as we know, the PF's actions were intentional (i.e. not involuntary), but we don't know what his intention(s) were.

Who is online

Popular Searches On Airliners.net