Air France does a 'Memmingen'.
A4, is your difficulty in understanding matched by that of explaining how the problem might be fixed?
There are many aspects to this subject. For example, has the industry identified ‘the problem’ – is it a situation where human input might help with a solution; or it ‘a mess’ – where the human is part of the problem in finding a solution. (A quote from ‘Systems Failure’, a pamphlet on Systems Thinking.)
This subject appears to fall into the category of Wicked Issues, which are associated with complex systems. (Quotes below from Connecting the dots.)
A wicked Issue:-
The site of persistent failures; repeated attempts at ‘solving’ the problem have had little effect.
The source of profound disagreement – over what the problems are, as well as over the improvements to be made.
Unbounded in scope; the issues sprawl outwards and interconnect with many other problems.
Resistant to completion; wicked issues cannot be ‘solved’ once and for all.
Complex, in the technical sense of being in principle unpredictable.
A best we might only:-
Avoid the worst ‘unintended consequences’
Avoid overestimating what can be achieved
Identify ways to be most effective
“If man is not to do more harm than good … he will have to learn that … where essential complexity … prevails, he cannot acquire the full knowledge which would make mastery of the events possible. He will therefore have to use what knowledge he can achieve, not to shape the results as the craftsman shapes his handiwork, but rather to cultivate a growth by providing the appropriate environment, in the manner in which the gardener does this for his plants”. Von Hayek, ‘The pretence of knowledge’.
There are many aspects to this subject. For example, has the industry identified ‘the problem’ – is it a situation where human input might help with a solution; or it ‘a mess’ – where the human is part of the problem in finding a solution. (A quote from ‘Systems Failure’, a pamphlet on Systems Thinking.)
This subject appears to fall into the category of Wicked Issues, which are associated with complex systems. (Quotes below from Connecting the dots.)
A wicked Issue:-
The site of persistent failures; repeated attempts at ‘solving’ the problem have had little effect.
The source of profound disagreement – over what the problems are, as well as over the improvements to be made.
Unbounded in scope; the issues sprawl outwards and interconnect with many other problems.
Resistant to completion; wicked issues cannot be ‘solved’ once and for all.
Complex, in the technical sense of being in principle unpredictable.
A best we might only:-
Avoid the worst ‘unintended consequences’
Avoid overestimating what can be achieved
Identify ways to be most effective
“If man is not to do more harm than good … he will have to learn that … where essential complexity … prevails, he cannot acquire the full knowledge which would make mastery of the events possible. He will therefore have to use what knowledge he can achieve, not to shape the results as the craftsman shapes his handiwork, but rather to cultivate a growth by providing the appropriate environment, in the manner in which the gardener does this for his plants”. Von Hayek, ‘The pretence of knowledge’.
Join Date: Apr 2002
Location: FUBAR
Posts: 3,348
Likes: 0
Received 0 Likes
on
0 Posts
"Mission" fixation / Pride (sorry to French colleagues , but I think that may be the main player here ) / Perceived "loss of face" / lack of knowledge of true capabilities of aircraft/ lack of situational awareness / task saturation / failure to prioritise realism over blind faith. . . . . well, maybe not an exhaustive list, but that is what we are up against.
Normally one of the 2 should be sufficiently aware of how screwed up it is, and normally he should be able to communicate this in a timely fashion to his colleague. . . . as I used to mention in a previous company where "normally" was a buzzword. . . . "normally" is merely an expectation - rather than the reality, past present or future.
Normally one of the 2 should be sufficiently aware of how screwed up it is, and normally he should be able to communicate this in a timely fashion to his colleague. . . . as I used to mention in a previous company where "normally" was a buzzword. . . . "normally" is merely an expectation - rather than the reality, past present or future.
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Alf - that is all terribly grand wording for
"They screwed up and with their experience should have seen the dog's cojones hanging in front of them at 30nm" is it not?.
"They screwed up and with their experience should have seen the dog's cojones hanging in front of them at 30nm" is it not?.
It may be grand wording, but I hope an improvement over many posts which infer understanding via blame, and thus it won’t happen to them, but without actually understanding how or why.
So whilst stating the obvious “… should have seen …”, "… why didn’t they…” can be self-gratifying, it does nothing to improve safety.
In order to avoid these types of incidents and similar events resulting in accidents, the industry has to put aside error, blame, culture, etc, in search of 'cause'; instead perhaps look at what goes right, and question why in order to achieve a better understanding so that individuals and operators can experience more of what goes right.
So whilst stating the obvious “… should have seen …”, "… why didn’t they…” can be self-gratifying, it does nothing to improve safety.
In order to avoid these types of incidents and similar events resulting in accidents, the industry has to put aside error, blame, culture, etc, in search of 'cause'; instead perhaps look at what goes right, and question why in order to achieve a better understanding so that individuals and operators can experience more of what goes right.
Ut Sementem Feeceris
Thanks BOAC - pretty much sums up my take 
Events of this magnitude are rare. Alf, I don't know if you're a pilot or not but we're talking about fundamentals here. It's a bit like an accountant trying to add up a ledger, but putting on a blindfold before operating the calculator - you just wouldn't do it because you know the chance of getting the sums correct are zero. A Professional pilot should have the ability to admit they've screwed up. Hopefully they learn from it and don't do it again.
So there is no "problem" to be identified - other than, IMHO, cultural or attitude - be that macho or nationalistic.
Alf - we were both typing at the same time. Your second point regarding blame culture is valid - to a point. Most developed airlines have active monitoring systems in place so that events like this can be investigated and the crew debriefed and retrained as required. However there is definite difference between an approach which was a bit high/fast and only just made the required "stable gates" and one such as this which was nothing short of reckless. Sometimes you've got so say it like it is - this was screwed up on a mammoth scale and the crew should be managed accordingly. Sometimes sanction AND retraining is warranted. The sanction may well go a long way to enhancing safety culture because it sets a precedent - if it's kept quite then no one benefits.
Events of this magnitude are rare. Alf, I don't know if you're a pilot or not but we're talking about fundamentals here. It's a bit like an accountant trying to add up a ledger, but putting on a blindfold before operating the calculator - you just wouldn't do it because you know the chance of getting the sums correct are zero. A Professional pilot should have the ability to admit they've screwed up. Hopefully they learn from it and don't do it again.
So there is no "problem" to be identified - other than, IMHO, cultural or attitude - be that macho or nationalistic.
Alf - we were both typing at the same time. Your second point regarding blame culture is valid - to a point. Most developed airlines have active monitoring systems in place so that events like this can be investigated and the crew debriefed and retrained as required. However there is definite difference between an approach which was a bit high/fast and only just made the required "stable gates" and one such as this which was nothing short of reckless. Sometimes you've got so say it like it is - this was screwed up on a mammoth scale and the crew should be managed accordingly. Sometimes sanction AND retraining is warranted. The sanction may well go a long way to enhancing safety culture because it sets a precedent - if it's kept quite then no one benefits.
Last edited by A4; 21st Jun 2013 at 17:46.
A4, yes ‘a Professional pilot should have the ability to admit they've screwed up’, but more often we don’t do that, even to our self. ‘Hopefully we learn from it’ and strive to avoid similar circumstances.
With the greatest respect to visually impaired accountants who manage their jobs successfully and accurately, I prefer pilots without a blindfold of ‘it won’t happen to me’.
Human performance is variable; this incident can be judged as an (unacceptable) extreme, and although the outcome was safe there was still risk, but at some point that was managed – why, how. The issue is not about the acceptability of risk, but one of human performance where variability is ill determined as too might the level of acceptability – that's in the eyes of the beholder, who draws the line. We all need to draw our own lines, but how can we be sure it is always in the right position.
I agree that the industry should put aside the search for ‘the problem’; “…accept and accommodate ambiguity over what the problems are, ‘defining the problem’ need not , and, often, should not, be the starting point …”
We should heed the principles of human factors, of systems thinking, and of how safety might be improved - the fundamentals.
We should accept that this type of event can be seen from a different viewpoints, not exclusively of being ‘reckless’ or ‘they should have done better’, but with views attempting to understand how this, the next crew, or ourselves might avoid such circumstances.
How can the industry avoid or overcome those issues identified by cps (#102), and identify which items come together ‘by chance’, or be of significance in particular circumstances.
There are many pilots operating today who can and do accomplish that, but are they able to explain how or why, or provide assurance their performance will always be adequate.
With the greatest respect to visually impaired accountants who manage their jobs successfully and accurately, I prefer pilots without a blindfold of ‘it won’t happen to me’.
Human performance is variable; this incident can be judged as an (unacceptable) extreme, and although the outcome was safe there was still risk, but at some point that was managed – why, how. The issue is not about the acceptability of risk, but one of human performance where variability is ill determined as too might the level of acceptability – that's in the eyes of the beholder, who draws the line. We all need to draw our own lines, but how can we be sure it is always in the right position.
I agree that the industry should put aside the search for ‘the problem’; “…accept and accommodate ambiguity over what the problems are, ‘defining the problem’ need not , and, often, should not, be the starting point …”
We should heed the principles of human factors, of systems thinking, and of how safety might be improved - the fundamentals.
We should accept that this type of event can be seen from a different viewpoints, not exclusively of being ‘reckless’ or ‘they should have done better’, but with views attempting to understand how this, the next crew, or ourselves might avoid such circumstances.
How can the industry avoid or overcome those issues identified by cps (#102), and identify which items come together ‘by chance’, or be of significance in particular circumstances.
There are many pilots operating today who can and do accomplish that, but are they able to explain how or why, or provide assurance their performance will always be adequate.
Join Date: Jan 2008
Location: ...
Posts: 3,753
Likes: 0
Received 0 Likes
on
0 Posts
At FL207, 33nm to touch down, 276 KIAS the captain selected the autopilot into open descent, reduced the thrust to idle, deployed the speed brakes and selected a speed of 300 knots. The aircraft reached a rate of descent of about 5000 feet per minute and descended through FL120 about 20nm before touch down, the captain selected 240 KIAS which also reduced the rate of descent. The aircraft descended through FL100 13.5nm before touchdown, the autopilot was disconnected, the landing gear was selected down, the crew reported runway in sight. Doing 250 KIAS the aircraft intercepted the localizer 8nm before touchdown at 6000 feet, 3400 feet above 3 degrees glideslope, the captain selected the go-around altitude into the flight control panel which caused the flight director to revert to vertical speed (4400 fpm rate of descent). Descending through 3550 feet, 1700 feet above glide, about 5nm from touchdown, flaps still at position 0, vertical speed -4400 fpm, speed brakes and landing gear extended, the first officer (ATPL, 1,700 hours on type) transmitted they were established on the ILS 19. The autothrust was disconnected, the engines were reduced to idle thrust. The aircraft received landing clearance. 8 seconds after the landing clearance the first officer advised they were "a little high" and requested a 360 (full circle) to the right. During these communications the captain re-engaged autothrust and autopilot, the glideslope capture activates, a GPWS warning "Sink Rate!" sounds at 836 feet AGL, 220 KIAS, 2500 fpm rate of descent 2.5nm before the runway threshold. The GPWS calls "Pull Up! Pull Up!" and "Too low! Terrain!"
To start of by accelerating to 300kts and then slow down to 240 kts at a point where you haven't even captured the profile is an enormous beginners mistake. You either speed up and stick with it until your below the profile, loosing the speed at lower altitude in denser air with every bit of drag there is, or you slow down straight away and put the drag out.
When you are this close and this high the only option that could lead to possible success is to slow down and get the drag out. With FULL and gear at 160-170 kts the airbus can give quite an impressive rate of descent at lower speed which buys time and thus improves the angle of descent. And at least it's fully configured. Still don't think it would work but in my opinion it's the only strategy that could have been attempted.
He then seems to manage to get the engines running above idle in v/s with speed brakes and gear out, at every step he is adding to his problems.
Anybody can make a mistake and get high, this guy doesn't know how to operate his aircraft though and is too arrogant to ask for some 360's or extra vectors.
The weather being good has little to do with it, weather doesn't change the laws of physic nor distance to the threshold. In fact when I can see the runway rushing towards me while I look way to high it's extra confirmation that something is going wrong! It's not a motivator to push on.
Last edited by 737Jock; 22nd Jun 2013 at 08:05.
Alf,
And here is the rub, somebody has got a bloody big alarm bell and rung it in the ears of the Air France community, and they are still not getting it! 
‘Hopefully we learn from it’
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Not just 'a bell' either, RWU - more like a Westminster Abbey session - are they going to admit the problems?
For all the reams or words that psychologists will write about this and other recent events, there comes a time when one has to say let's not give all these events grand names and analysis - they are 'cock-ups', plain and simple, and I have yet to hear of any course one can introduce to stop those, and in many ways by introducing the 'psychology' approach one is almost admitting the problem is pretty wide-spread, instead of a few mavericks.. Cut out the rot is the best way. 'Hang 'em high' is sometimes the best solution.
I suppose looking on the bright side a pilotless aircraft would have flown an orbit..............
For all the reams or words that psychologists will write about this and other recent events, there comes a time when one has to say let's not give all these events grand names and analysis - they are 'cock-ups', plain and simple, and I have yet to hear of any course one can introduce to stop those, and in many ways by introducing the 'psychology' approach one is almost admitting the problem is pretty wide-spread, instead of a few mavericks.. Cut out the rot is the best way. 'Hang 'em high' is sometimes the best solution.
I suppose looking on the bright side a pilotless aircraft would have flown an orbit..............
Join Date: Jun 2000
Location: last time I looked I was still here.
Posts: 4,507
Likes: 0
Received 0 Likes
on
0 Posts
This, and the Thessaloniki incident, are all taking about the captain did this to screw up he kept digging etc. In TSL the operator might have a cockpit gradient and culture problem. This has been mentioned for years and the MCC CRM courses in western airlines are supposed to have levelled one and eradicated the other. Yet this is AF; a major ancient player.
In both cases a perfectly sensible (hopefully), educated, qualified, intelligent individual = F/O sat and watched while some other clown tried to kill him, yet did not a lot to help his survival. That leads to 2 questions. How did the clown get to be in LHS in the first place and then dig this hole? And why did the F/O stare death in the face so meekly?
In both cases a perfectly sensible (hopefully), educated, qualified, intelligent individual = F/O sat and watched while some other clown tried to kill him, yet did not a lot to help his survival. That leads to 2 questions. How did the clown get to be in LHS in the first place and then dig this hole? And why did the F/O stare death in the face so meekly?
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Maybe there is a huge 'gradient' in AF? Look at the farting around in 447 where no-one seemed to be prepared to say boo to l'oie until le chef arrived.
Join Date: Nov 2009
Location: flying by night
Posts: 500
Likes: 0
Received 0 Likes
on
0 Posts
Thanks Ian W for pointing me to " Multiple resources and performance prediction" (Wickens, 2002) http://sunburst.usd.edu/~schieber/ps...ckens2002b.pdf
some heavy reading...jumping to the conclusion:
"Basic psychology and neurophysiology must identify the characteristics of human information processing that make such endeavours more, or less, successful. The analysis and prediction of human productivity and safety in high workload environments requires models to predict such di€fferences."
In summary, scientists are starting to understand more about "cognitive overload", and there are some interesting theories, yet - more research is needed.
some heavy reading...jumping to the conclusion:
"Basic psychology and neurophysiology must identify the characteristics of human information processing that make such endeavours more, or less, successful. The analysis and prediction of human productivity and safety in high workload environments requires models to predict such di€fferences."
In summary, scientists are starting to understand more about "cognitive overload", and there are some interesting theories, yet - more research is needed.
BOAC, “… one is almost admitting the problem is pretty wide-spread, instead of a few mavericks …”
Yes, ‘it’ is widespread. It is called human behaviour and, even after selection, training, and education, it varies across the complete spectrum of operations.
You wish to draw a line to delineate the norm from the maverick; by all means do so, but please specify where the line is to be drawn for every situation which a pilot might encounter, to enable us to avoid these situations.
I choose not to draw a line, instead seek those factors which contribute to variable human behaviour, and then how these could be improved to achieve a more consistent and hopefully less variable human performance.
RAT 5, Assuming that the PF misjudged the situation (variable human performance), then for other than probability, what would indicate that the PM understood the situation any better than PF, and thus could have chosen to intervene.
Both pilots were in the same situation, thus their perception and understanding could have been similar, as we expect in everyday normal operations. But without evidence we cannot tell – probabilities or otherwise.
To some extent cross monitoring depends on detecting a difference in understandings, but if the mind-sets are similar then there may be no disparity to detect – again as in normal ‘safe’ operations.
There is opportunity for self-monitoring as a defence – “how’s my thinking”, “what does this mean”, but this too can suffer from ‘dual’ failures, and there may be other disparities between pilots due to training and experience.
An analogy is like suffering spatial disorientation, where both pilots can be similarly affected; except in this instance it might have involved cognitive disorientation – overload, which demands intentional resource. Whereas with the less attention-demanding spatial disorientation, we might be able to recall the mental drill of ignore our senses and believe the instruments.
Yes, ‘it’ is widespread. It is called human behaviour and, even after selection, training, and education, it varies across the complete spectrum of operations.
You wish to draw a line to delineate the norm from the maverick; by all means do so, but please specify where the line is to be drawn for every situation which a pilot might encounter, to enable us to avoid these situations.
I choose not to draw a line, instead seek those factors which contribute to variable human behaviour, and then how these could be improved to achieve a more consistent and hopefully less variable human performance.
RAT 5, Assuming that the PF misjudged the situation (variable human performance), then for other than probability, what would indicate that the PM understood the situation any better than PF, and thus could have chosen to intervene.
Both pilots were in the same situation, thus their perception and understanding could have been similar, as we expect in everyday normal operations. But without evidence we cannot tell – probabilities or otherwise.
To some extent cross monitoring depends on detecting a difference in understandings, but if the mind-sets are similar then there may be no disparity to detect – again as in normal ‘safe’ operations.
There is opportunity for self-monitoring as a defence – “how’s my thinking”, “what does this mean”, but this too can suffer from ‘dual’ failures, and there may be other disparities between pilots due to training and experience.
An analogy is like suffering spatial disorientation, where both pilots can be similarly affected; except in this instance it might have involved cognitive disorientation – overload, which demands intentional resource. Whereas with the less attention-demanding spatial disorientation, we might be able to recall the mental drill of ignore our senses and believe the instruments.
Last edited by alf5071h; 22nd Jun 2013 at 22:08. Reason: typo
I thought most airlines had gates to meet on the approach. If you don't meet the gate then you do not proceed to the next phase and must do something to achieve that gate before you can proceed. The first gate would be 250kt at 9000ft at 30 nm.Airbus publish this in the fcom but in this case the flight was so far away from it that it beggars belief that they continued. If you are trying that hard to make an approach work then it probably will not work and you should only persevere if the aircraft is on fire.
I have noticed a certain misplaced arrogance from observing AF operations over the radio over the years and I would not fly with them by choice.What is more disturbing is that their head of safety who oversaw a numer of hull losses is now a big wheel in EASA.
I have noticed a certain misplaced arrogance from observing AF operations over the radio over the years and I would not fly with them by choice.What is more disturbing is that their head of safety who oversaw a numer of hull losses is now a big wheel in EASA.
Last edited by tubby linton; 22nd Jun 2013 at 19:52.
Ut Sementem Feeceris
With all due respect Alf, you have a way of using some "big" words trying to almost "justify" this crews actions. As I said a couple of posts back, this was an "off the scale event" which needs, to put it bluntly, a severe reality check/correction for the individuals involved.
You ask "where the line is drawn"? Well, if you've programmed the FMGC correctly, put the descent winds in and the constraints you require/want in the descent, the aircraft will draw them for you! And it TELLS YOU it "ain't gonna work"......
BASIC training when you come onto jets is, typically, 3 x your height + 10nm to slow down for a "normalish" profile - these guys didn't even get close and no amount of what / where / why / how / fluffy can start to "soften" what they did.
The more I've thought about it the more "angry" I feel about it. It could be argued that these guys are the equivalent of the market traders who "bet the farm" in the hope it would come good but didn't and lose £x hundred million. The difference is that these characters had the potential to take 150+ soles to meet their maker.
It can be analysed to death - they screwed up and didn't sort it early - bottom line.
You talk of "dual failure" of cognitive overload. Overload, typically, only occurs in the latter stages. If you're 20,000' at 30d there is NO overload at that point. Any pilot with any NORMAL, EXPECTED "cognitive recognition" would say AT THIS POINT, "we need more track miles" - stress, risk, error, high workload - REMOVED.
Sorry to sound blunt - this was a totally unnecessary event. Macho BS/pride - call it what you like - no amount of psychobabble will ever justify this to me.
You ask "where the line is drawn"? Well, if you've programmed the FMGC correctly, put the descent winds in and the constraints you require/want in the descent, the aircraft will draw them for you! And it TELLS YOU it "ain't gonna work"......
BASIC training when you come onto jets is, typically, 3 x your height + 10nm to slow down for a "normalish" profile - these guys didn't even get close and no amount of what / where / why / how / fluffy can start to "soften" what they did.
The more I've thought about it the more "angry" I feel about it. It could be argued that these guys are the equivalent of the market traders who "bet the farm" in the hope it would come good but didn't and lose £x hundred million. The difference is that these characters had the potential to take 150+ soles to meet their maker.
It can be analysed to death - they screwed up and didn't sort it early - bottom line.
You talk of "dual failure" of cognitive overload. Overload, typically, only occurs in the latter stages. If you're 20,000' at 30d there is NO overload at that point. Any pilot with any NORMAL, EXPECTED "cognitive recognition" would say AT THIS POINT, "we need more track miles" - stress, risk, error, high workload - REMOVED.
Sorry to sound blunt - this was a totally unnecessary event. Macho BS/pride - call it what you like - no amount of psychobabble will ever justify this to me.
Join Date: Jan 2008
Location: Denver, CO
Posts: 265
Likes: 0
Received 0 Likes
on
0 Posts
The more I've thought about it the more "angry" I feel about it. It could be argued that these guys are the equivalent of the market traders who "bet the farm" in the hope it would come good but didn't and lose £x hundred million. The difference is that these characters had the potential to take 150+ soles to meet their maker.
The "punters" sitting in A thru whatever ARE NOT paying for someone to "hope" it will all come out right.
Teh simple
who financed the airframe's purchase didn't to so with the intent that someone piloting the multi-million $/L aircraft would reach a point - in routine operation - "hoping" it would all work out fine.And lastly, AF certainly didn't (and prolly doesn't routinely) put people in the LH seat who can only hope a routine approach and landing will come out okay.
First order of business - heads on spikes.
Next, sort out the culture which caused this to happen, and make it public so everyone can learn.