"To err is human": differing attitudes to mistakes in EK and Turkish accidents

[SPA83]

You accept that an aircraft can be allowed to fly not being repaired just to save time or money or other crazy reasons, so you must accept too that pilots sometimes may not be in the loop, just along for the ride.

If you think pilots should always be professional, you must also consider that an aircraft must always be in good condition with no defect

[757_Driver]

Worse than this, if I have understood correctly, is that there's no agreement on what to do about the problem. PJ2 says of the TK crash:

Further, there is no understanding in this accident that is useful in the addressing of organizational, systemic or human factors accidents.

So what to do? Are you saying these kinds of risks cannot be reduced?

No. They can and are reduced, but generally only where the operator, regulator and Accident investigation bereau understand what their role is in a true safety system. As soon as "national pride", "Face Saving", "Cultural influences".... call it what you will, comes into the picture then this system fails, and the ultimate manifestation of that failure is often a human factor accident. Many of these accidents are the ones that, after the event, we all say "what the hell....."
We are in real danger in our society (under the current regime) of going down the same route - where face saving, the good of the 'party' and political correctness become more important than telling the truth and having a sensible, safe regulatory regime. After all just look how the Banks regulator bent to political will and refused to question anything - a situation that is hugely relevent to our industry and should be a real wake up call to the importance of the political independence of national aviation regulators.
Fortunately it appears the CAA and AAIB are generally impervious to the insidious influence of westminster, however it is easy to see how these things can happen, and we can all see where it ends up.
To be honest i'm not sure what the answer is, but I see the errors made by the turkish crew, not just as a human error (granted - one that is almost impossible to understand) , but as the symptom of an entire systemic failure.
The fact that this is not the first in this particular airline and authorites case, and indeed is another statistic in quite a shocking record leads one even further to that conclusion.

[PanPanYourself]

I was embarrassed when a year ago today Turkey banned access to youtube.
I was dismayed when they banned richarddawkins.net based on the ravings of an indicted criminal lunatic.
I was in disbelief when they banned several blogging sites owned by google, and free website hosting domains such as geocities.
I was in shock when they made themselves out to be Islamic fundamentalists at the recent NATO submit over a complaint about free speech in Denmark.

But in recent memory I haven't been truly ashamed of my country as much as I was in the days following the THY crash at Schipol. The entire Turkish media made the pilots out to be heroes based on the most ridiculous "evidence" just hours after the accident. The government and THY lied shamelessly and the overwhelmingly ignorant Turkish public lapped it all up.

So yes, when I hear of an accident involving Turks my immediate reaction is to wonder which idiot caused it and how, whereas I give civilized Westerners the benefit of the doubt.

[BOAC]

Originally Posted by 757_driver

turkish crash, the type 3 error has no error traps on an operational basis

- while I would prefer not to 'categorise' here, I would pick you up there briefly (hopefully without diverting the thread too much from the original question!) - there IS indeed an error trap, namely the stable approach criteria. They were patently WRONG at 1000' and again at 500'.. One would HOPE that a TC would be watching this like a hawk, if not actually monitoring airspeed per se, as this should be a major training point for all pilots. What we need to try and understand is why these pointers were missed. Indeed, does this airline have these criteria?

[icarus sun]

I suggest that all pilots in highly automated aircraft should hand fly at least one approach per month. With the AP/AT system disengaged. This should be legal requirement. I know that on long haul operations this may not be easy, then do it in the sim. For the MEL EK near miss i suggest the companies publish on the flight plans the estimated power settings and speeds for the planned weight. These to be cross checked/updated by the actual weights from the computer.

[4Greens]

It may be relevant to suggest that FOQA or Flight Data Analysis becomes a standard tool in all airlines. It is such a good device for Fleet Monitoring and hence improved safety that it should be on your ticket if the airline has it. Those that do are the ones to fly with.

[757_Driver]

I see your point BOAC. I was thinking of the wider issue - off the top of my head, all the 'type 3' traps (i prefer not to categorise either - but it makes it more readable on this thread if we do!) rely on a human factor and if the entire system is broken, then what value those checks and criteria?

Anyway before we get too much into the self flaggelation lets not forget that for all its faults the aviation safety model is still amongst the best in the world. I don't know off the top of my head the number of fatalities annualy in the world but its in the medium / high hundreds if that.
I could have an accident tomorrow (3000 deaths on the road in the UK every year) go to a hospital and get sliced open by a doctor who went to a med school decades ago, possibly anywhere in the world, and has minimal requirements to do any form of ongoing training, education or checking, has no restriction on his duty hours and is largely unnacountable for his errors. over 10,000 'avoidable' deaths occur in uk hospitals every year under this safety model.
So lets no throw the baby out with the bath water and recognise that, largely, we have a great and durable safety regime. Yes we can tinker and improve it, but lets not ruin it in the process.

[FrequentSLF]

Have a question that is long time I want to ask.

Training (or lack of) is coming up as contributory cause almost every time an accident happens.
I wonder who is the ultimate responsible for the training, the pilot or the airline?
I mean if a pilot feels that is not trained sufficiently can ask for more training or his/her position might be in jeopardy for admitting the lack of training?

Thanks

[jetopa]

I mean if a pilot feels that is not trained sufficiently can ask for more training or his/her position might be in jeopardy for admitting the lack of training?

Well, in an ideal world he/she should do so. But: there might be somebody pointing at you and your 'apparent' lack of self consciousness - or whatever you like to call it. Been there, done that. Unfortunately.

Best thing is to publicly discuss such matters so that we all can profit from that. No need to say: 'look, those stupid idiots punched in the wrong waypoint into their FMS and did not realize that it was in a different country with a mountain in their way...'.

Nobody (only the type 1 guys do) does deliberately fly into desaster. The way the Schiphol accident was handled in Turkey was shameful. But the fact that an incident like the one in Melbourne was allowed happen again is, in my opinion, no good PR for Emirates either.

It all boils down to acknowledging that we all are human and that even LH, BA, AF or THY 'heroes' can f... up pretty badly. No need to cover up.

[The Real Slim Shady]

In the final analysis we self regulate: that is why we have 2 pilots, not a man and a boy, but PF and PM.

When the PF screws up and PM doesn't correct the error the self regulation fails and problems occur.

Sound training programmes, adherence to SOPs and an open blame free reporting culture all contribute to the safety management system but ultimately the atmosphere in the office remains the weak link.

[ZuluKilo66]

Congratulations on one of the most thoughtful opening posts I've seen on Pprune. Clear analysis and questions provoking clear, thoughtful responses. I look forward to reading more.

[Tailspin Turtle]

Another Explanation (and Oversimplification)

According to The Right Stuff by Tom Wolfe (an excellent read, okay movie), the TK pilots were presented with a problem, didn't cope with it, and crashed and got killed - they had the wrong stuff. The EK pilots were presented with a problem, handled it, and got back on the ground with nobody hurt - they had the right stuff. According to Wolfe's understanding and description of the test pilot psyche (and that of the race car driver, mountain climber, etc.), he (or she) copes with the risk of fatal failure by "knowing" that they have the right stuff and will not crash, whereas someone who crashes and is killed did not.

[Centaurus]

So what to do? Are you saying these kinds of risks cannot be reduced?

Some operators have a long history of dangerous practices which can largely be blamed on ethnic culture centred around an unshakeable belief of "my Deity will always protect me". A numb acceptance of what will be, will be.

You can debate the subject ad nauseum but what is certain is that cultural mores should have have no place on the flight deck. Where it exists then sooner or later there will be dangerous and undisclosed close shaves where to these pilots their personal Deity does his job and saves the day - or he doesn't step in and people die. Forget James Reason and all the other books of that genre. Interesting stuff to the interested, but gobbly-dook to a culture driven pilot.

[lomapaseo]

You can debate the subject ad nauseum but what is certain is that cultural mores should have have no place on the flight deck. Where it exists then sooner or later there will be dangerous and undisclosed close shaves where to these pilots their personal Deity does his job and saves the day - or he doesn't step in and people die. Forget James Reason and all the other books of that genre. Interesting stuff to the interested, but gobbly-dook to a culture driven pilot.

One may live in a culture, but in the cockpit you live by the rules, not by a culture. That's the whole idea of training before you fly as a two man (or more) cockpit.

This is a good thread

I'm hoping that we can get beyond the culture of a country and their tabloid news pandering to an ingorant public and concentrate more on what goes on or should go on in the living environment of the cockpit.

I like to think of the error chain as:

Skill based

Knowledge based

Rule based


My narrow view at this time is that maybe we should evaluate the rules-based part of human error and ask why were they broke?

If you think that it was culture based then why was it not trained out?

Who here has the facts of what their training consisted of vs all the surviving pilots?

[Blacksheep]

I don't believe that cultural influences are as significant as some may make out. There are plenty of examples of aircraft being allowed to fly themselves into the ground while the crew busied themselves about a set of baffling indications.

What were the cultural aspects of the crew (a 3 crew aircraft too) who allowed their aircraft to fly itself into the Everglades while they messed about trying to fix a landing gear green light? Then there was the B757 that crashed into the sea because the crew were baffled by a pitot-static problem? Another crew, another place and the captain flew the aircraft to a successful landing using the right power settings for aircraft configuration and each flap setting.

A captain barrel-rolled his 747 into the sea while trying to follow a failed ADI indication at night.

I could go on, there are dozens of similar examples, but it is a fact that when distracted by indications that they have never encountered before, human beings become fixated on the anomaly and abandon attempts to monitor other events. It is a Human Factors or basic psychology issue based upon the many millenia of evolution of our species and it cannot simply be written off as a training shortfall, a character flaw or as a 'cultural thing'.

[screwballburling]

We will never eliminate human error, Never, in fact man is fast becoming the "weak link" in the chain.

I am a great believer in keeping things simple. Things now are getting complicated as we're getting away from the basics. Complicate things and more people screw up.

Here is an E.G., At flying school we were taught to use full power for t/o. If you didn't get full power you aborted! Now we are asked to use reduced power for t/o!! Can you imagine taking off in say a trainer at 2200 rpm instead of say 2500 -2600 rpm? of course not, yet we are doing it in airline operations FFS!! Are the savings worth it?

There is a recipy for disaster right there, as it is one more weak link in the chain.

This my 2 cents worth from a someone who started in this business 40 years ago this year. Maybe I have it all wrong and should give it up. Look at the airbus accident in the South of France late last year. All so unnecessary, in my view.

[biscuit74]

I think Blacksheep's description is excellent.

The TK issue was apparently very a much a Human Factors, classic distraction, case. Described at length by James Reason and others. The cultural aspect may be that if others in the crew noticed something wrong, perhaps they felt inhibited about mentioning it in time. In which case the holes in th Swiss Cheese start to line up. Training activity + distraction + cultural inhibition. Any one of those could happen anywhere. We'd like to believe they should not all line up at once.

I think that is why this for many pilots seems to hover between Gibon2's Type 2 & Type3. For myself I view it as a Type 3, because loss of situational awareness in what was not an apparently high stress environment should not happen to a well organised & competent pilot. This is part of what we teach people as basic airmanship. But I am an (now ex-) instructor not an airline pilot.

The EK event sounds to have been an oversight and a failure, presumably, to follow procedures. Since I understand it is quite common for departure routines to be interrupted - not unusual, in most professions the 'ideal' and the actual worlds are quite different - an error like this is easier to visualise and I suspect that is why rather more folk have the 'there but for the grace.....' feeling about that one, hence leaning towards Type 2.

Interesting discussion. I don't see any way to totally avoid these things, because of the way human beings operate.
Maintaining awareness of the pitfalls and snags in any activity does help. However we all get complacent, in all activities. Every so often, at the very least someone gets a nasty surprise. If we are lucky, we learn from that 'incident' and the general awareness threshold are back up higher for a while. That seems to be a consistent pattern in most pursuits, not just aviation. In theory, if we could graph the awareness curve, intervention at the right moment can help improve safety. The services used to do this with extra supervision at the various high risk flying hours points.

[White Knight]

Screwballfellah - a Cessna 152 is grossly underpowered even at full bore.. A 345 (which I fly) is fantastically overpowered at full thrust (108,000 lbs). The two also fall under totally different performance and certification categories so the comparison is ridiculous

I suggest that after 40 years you've still got some learning to do with regards to Performance A aeroplanes............................

On a related matter - if you go and use TOGA on every take-off you'll soon find that your engine failure rates shoot up!!!

[Herc708]

More raw data flying

Icarus Sun has a good idea but I would go further and remove the FD as well (FD is too easy) - pure raw data at least once a month

Unfortunatley the hysteria being created about level busts etc is having the side effect of reducing the overall skill level of the pilot population. In a complex SID or STAR only the very brave fly it raw data on a regular basis

Ideally, each pilot should fly a raw data SID or STAR once a month and allowance this should be included in SOPs. Extra briefing etc but the advantages outweigh the disadvantages. This avoids the embarrassment at the Sim when re-takes are the norm because everyone has forgotten how to fly raw data

[bjornhall]

My simplistic understanding of rule based behavior goes something like this: You are faced with a situation, you analyze it (more or less subconsciously) to identify it with some category for which there is a rule, you recall or choose that rule, and you apply that rule. The process is not rational in the way it appears in writing, but it nevertheless follows a chain or a cyle of "identify - choose - apply".

IMHO, if that is the way one manages speed control on final approach, then there is something seriously wrong. My view is that it should be a skill based, not a rule based behavior. Admittedly an amateur pilot's view, but am I really that far off the mark?

Skill based behavior on the other hand is an automatic, ongoing process. Importantly, it is far harder to get distracted from a skill based activity than it is from a rule based activity. Distraction means the "identify" step never occurs, and then the "apply" action is never taken; furthermore, one does not realize that something important was missed, until the indications become overwhelming (e.g., stick shaker).

Transforming an ability from a rule based to a skill based behavior is the role of training; maintaining it in the skill based category to prevent a relapse into rule based or even knowledge based is the role of recurrent training.

I like to think there are few professional pilots who make Type 3 errors because they fit in the "wrong stuff" group; i.e., that they could not perform to an acceptable standard no matter how well they are trained. I think they make such errors because they are insufficiently trained, and that with sufficient training they could perform as required.

My tentative conclusion then is that the mere existence of Type 3 errors indicate training problems.

Type 2 errors, on the other hand, are precisely the type of problems where what can now be referred to as "classical human factors issues" and the traditional solutions to such problems can be applied (read: J. Reason, Swiss Cheese, System Failures and so on).

EDITED to add: On the other hand, "someone should always mind the shop" seems to have its correct place in the rule based category. It would be interesting to understand if that step failed? Did nobody realize that nobody was minding the shop? Or did the person who was doing so get distracted and failed to perform it properly?