Turkish airliner crashes at Schiphol
lederhosen, whilst not disagreeing with your view (#2628), I would question the implied context and assumptions.
How is knowledge of ‘the basic stuff’ acquired?
How do pilots know that the thrust lever is at idle – ‘hands on’ experience providing a geometric feel for position and motion? How often do pilots look at the engine instruments during an approach, if so what is normal, and in what circumstances?
The point is that these aspects (of skill) are acquired over time and require focussed effort. Scan patterns are developed, how and when to share tasks understood, and what the approach norms are for a range of circumstances – experience – pre packaged solutions retrieved from memory. The issue is how and when the solutions are placed in memory and if they can be retrieved.
Two of the crew were undergoing training, most likely without all of these skills or knowledge. Thus they would have to allocate considerable attention to the learning / understanding processes. They didn’t know what was ‘normal’; they had to deduce this from first principles, from basic training. There was little or no time, or spare attention to see, identify, and understand ‘minor’ abnormalities, probably not able to relate a RA malfunction (even if seen) to the consequences of autothrust mode change. Imagine the likely thought patterns; providing ‘retard’ was seen (annunciation or thrust lever motion) – “why is that occurring now”, “have I made a mistake, what, where, why … ”.
The captain could suffer similar limitations of attention, and although having the required experience, knowledge and skills, his task and workload in supervising both crew and aircraft reduced his attentional capacity. His first thoughts – assuming ‘a problem’ was seen, “what’s this guy done now; that shouldn’t happen, why, what does it mean …”. Thus all of the crew quite understandably could be ‘maxed out’ – behind the aircraft.
This is not excusing the crew, not that from a human factors view that any excuse is required, but to place them and ourselves in the context of the accident and not ‘as now’ with hindsight.
We should recall our memories of the first training flight to a major airport, the first real operational use of the autos, IFR to autoland, perhaps struggling to understand, wanting to learn, not wishing to suffer an error, particularly when observed by another trainee.
A go around at a sensible point – I agree. The evidence suggests that this point was identified late on and the manoeuvre commenced by instinctively pushing the thrust levers forward (late but not necessarily fatally so). Except in this instance the RA/auto system malfunction again retarded the thrust levers probably because the TOGA mode was not selected - an error or oversight in conditions of extreme pressure and stress.
No blame, just understanding, and the need both individually and collectively to learn from this – from the unfortunate experiences of others.
How is knowledge of ‘the basic stuff’ acquired?
How do pilots know that the thrust lever is at idle – ‘hands on’ experience providing a geometric feel for position and motion? How often do pilots look at the engine instruments during an approach, if so what is normal, and in what circumstances?
The point is that these aspects (of skill) are acquired over time and require focussed effort. Scan patterns are developed, how and when to share tasks understood, and what the approach norms are for a range of circumstances – experience – pre packaged solutions retrieved from memory. The issue is how and when the solutions are placed in memory and if they can be retrieved.
Two of the crew were undergoing training, most likely without all of these skills or knowledge. Thus they would have to allocate considerable attention to the learning / understanding processes. They didn’t know what was ‘normal’; they had to deduce this from first principles, from basic training. There was little or no time, or spare attention to see, identify, and understand ‘minor’ abnormalities, probably not able to relate a RA malfunction (even if seen) to the consequences of autothrust mode change. Imagine the likely thought patterns; providing ‘retard’ was seen (annunciation or thrust lever motion) – “why is that occurring now”, “have I made a mistake, what, where, why … ”.
The captain could suffer similar limitations of attention, and although having the required experience, knowledge and skills, his task and workload in supervising both crew and aircraft reduced his attentional capacity. His first thoughts – assuming ‘a problem’ was seen, “what’s this guy done now; that shouldn’t happen, why, what does it mean …”. Thus all of the crew quite understandably could be ‘maxed out’ – behind the aircraft.
This is not excusing the crew, not that from a human factors view that any excuse is required, but to place them and ourselves in the context of the accident and not ‘as now’ with hindsight.
We should recall our memories of the first training flight to a major airport, the first real operational use of the autos, IFR to autoland, perhaps struggling to understand, wanting to learn, not wishing to suffer an error, particularly when observed by another trainee.
A go around at a sensible point – I agree. The evidence suggests that this point was identified late on and the manoeuvre commenced by instinctively pushing the thrust levers forward (late but not necessarily fatally so). Except in this instance the RA/auto system malfunction again retarded the thrust levers probably because the TOGA mode was not selected - an error or oversight in conditions of extreme pressure and stress.
No blame, just understanding, and the need both individually and collectively to learn from this – from the unfortunate experiences of others.
hetlfield;
"Simply fly the plane."
Yup. The thread could have (and should have) begun and ended right there.
safetypee;
Apropos simulation, agree. However, we have seen this important aspect of training come under economic and therefore scheduling pressure as well.
Simulator sessions are reduced from 4hrs to 3.5hrs and AQP approaches reduce the footprint from 4 sessions per year, (typically taken twice, 1rst day "practise", second day, PPC/IFR) to once every eight months for example. Also, there is a drifting towards an emphasis on documentation rather than on actual outcomes. With LOFT exercises, valuable sim/crew/instructor time is spent waiting for scenarios to unfold in real time. Advocating a return to "new day, new airplane" is not the key either. We have all experienced those sessions which were so heavily scripted with abnormals that learning and practise is compromised, but with new approaches which emphasize thinking (within the AOM) rather than swift iterations of QRH memory items, valuable time is lost in various phases of flight waiting for the next waypoint, etc. It is indeed a compromise.
It is accepted that the purpose of training is to render the unfamiliar, familiar, first to extinguish "surprise" and then to build habits of thinking as per the AOM guidance material, company Ops Manuals and the regs. I agree with your comments on training on the line in regular operations. It is a marginal approach to training which line-operations training captains "make work" because they must but which in truth is very much a compromise in many ways. All training captains have their stories, myself included, which illustrate these points admirably but which, sadly, do not have the necessary impact upon the decision-makers at the airline and within the regulatory authority to bring the issues to a head and cause change. "Kicking tin" remains the primary driver, thus the problems never go away, they just remain dormant until the holes line up once again.
Because historical (1950 - 1970), primary causes of accidents such as altitude awarenes, mid-air collision, navigation error, weather (no/poor radar, windshear), ATC capabilities, mechanical (engine or system) failure, system and instrumentation design (ergonomics) and others have largely been resolved (and the accident rate shows it), the emerging record of accident causation now demonstrates emminent preventability through training and comprehension of human factors. And Human Factors, it must be admitted, extend far beyond the cockpit and is a systematic matter, not a "pilot" matter.
A realistic approach to fares is way overdue. The reason the flying public will aggressively shop for a $10 saving on fares is because the airlines have taught them to do so. Airlines have patronized cheapness by devaluing their product, devaluing the contributions of trained, dedicated employees, magically believing that the principles of flight safety can be done on a quarterly report basis. It cannot.
The astonishing thing about such nickel-and-diming is, that same public will stand in long lines at Starbucks to pay that same ten bucks for two designer coffees.
That is a perception issue, not a marketing issue. Airlines may squeeze fares and wages alike so that investors and CEOs, not safety, come first, but they also squeeze because they don't (and given the expectation handed out during deregulation, can't) take in enough money to pay the bills.
Given that nobody manages by walking around the shop floor anymore, it is a short hop to a shoestring mentality, justified because "nothing happens" and when it does, the full picture of why is never in view, just the picture of the guys or gals at the pointy end.
The present accident record is exceptional notwithstanding the tremendous human suffering which lies behind even one airline fatality let alone the 600+ passengers who lost their lives in 2009. But because new leaders coming into the industry don't seem to know the history of how the industry's safety record got that way, are taking it for granted that safety is somehow magically engineered into the airplanes and "the system" and isn't related to (expensive) programs, (time-consuming) processes and (expensive) specialist resources.
If there is reward to be had in change and positing solutions, I think it may lie along the lines that the thinking and production processes described above must change, and when it does all else will follow because it can be both philosophically legitimated and therefore economically justified.
"Simply fly the plane."
Yup. The thread could have (and should have) begun and ended right there.
safetypee;
Operators shouldn’t schedule training at busy international destinations, but without this how do new pilots learn; simulation is far from perfect in operational scenarios, or gaining the experience after qualification in line operations would unfairly pressure line captains with a ‘training’ task – again a risk to safety.
Simulator sessions are reduced from 4hrs to 3.5hrs and AQP approaches reduce the footprint from 4 sessions per year, (typically taken twice, 1rst day "practise", second day, PPC/IFR) to once every eight months for example. Also, there is a drifting towards an emphasis on documentation rather than on actual outcomes. With LOFT exercises, valuable sim/crew/instructor time is spent waiting for scenarios to unfold in real time. Advocating a return to "new day, new airplane" is not the key either. We have all experienced those sessions which were so heavily scripted with abnormals that learning and practise is compromised, but with new approaches which emphasize thinking (within the AOM) rather than swift iterations of QRH memory items, valuable time is lost in various phases of flight waiting for the next waypoint, etc. It is indeed a compromise.
It is accepted that the purpose of training is to render the unfamiliar, familiar, first to extinguish "surprise" and then to build habits of thinking as per the AOM guidance material, company Ops Manuals and the regs. I agree with your comments on training on the line in regular operations. It is a marginal approach to training which line-operations training captains "make work" because they must but which in truth is very much a compromise in many ways. All training captains have their stories, myself included, which illustrate these points admirably but which, sadly, do not have the necessary impact upon the decision-makers at the airline and within the regulatory authority to bring the issues to a head and cause change. "Kicking tin" remains the primary driver, thus the problems never go away, they just remain dormant until the holes line up once again.
Because historical (1950 - 1970), primary causes of accidents such as altitude awarenes, mid-air collision, navigation error, weather (no/poor radar, windshear), ATC capabilities, mechanical (engine or system) failure, system and instrumentation design (ergonomics) and others have largely been resolved (and the accident rate shows it), the emerging record of accident causation now demonstrates emminent preventability through training and comprehension of human factors. And Human Factors, it must be admitted, extend far beyond the cockpit and is a systematic matter, not a "pilot" matter.
A realistic approach to fares is way overdue. The reason the flying public will aggressively shop for a $10 saving on fares is because the airlines have taught them to do so. Airlines have patronized cheapness by devaluing their product, devaluing the contributions of trained, dedicated employees, magically believing that the principles of flight safety can be done on a quarterly report basis. It cannot.
The astonishing thing about such nickel-and-diming is, that same public will stand in long lines at Starbucks to pay that same ten bucks for two designer coffees.
That is a perception issue, not a marketing issue. Airlines may squeeze fares and wages alike so that investors and CEOs, not safety, come first, but they also squeeze because they don't (and given the expectation handed out during deregulation, can't) take in enough money to pay the bills.
Given that nobody manages by walking around the shop floor anymore, it is a short hop to a shoestring mentality, justified because "nothing happens" and when it does, the full picture of why is never in view, just the picture of the guys or gals at the pointy end.
The present accident record is exceptional notwithstanding the tremendous human suffering which lies behind even one airline fatality let alone the 600+ passengers who lost their lives in 2009. But because new leaders coming into the industry don't seem to know the history of how the industry's safety record got that way, are taking it for granted that safety is somehow magically engineered into the airplanes and "the system" and isn't related to (expensive) programs, (time-consuming) processes and (expensive) specialist resources.
If there is reward to be had in change and positing solutions, I think it may lie along the lines that the thinking and production processes described above must change, and when it does all else will follow because it can be both philosophically legitimated and therefore economically justified.
hetfield, we hear you.
But just consider that for any one of the reasons in previous posts the crew were not able to simply fly the plane, or at least not in the manner which you appear to assume.
This ‘inability’ is a combination of human behaviour and the situation in which the crew might find themselves. Change either one or both, then errors and mistakes might be avoided. Humans are difficult to change, and if as in this instance there few or any existing attributes to change, then change may not be an option; thus change the situation.
Arguably the training captain should have foreseen the risks and planned accordingly, so too his management, ATC, the regulators, and at the root of all this the manufacturer who’s design could have avoided the particular circumstances.
Given the choice I would seek a design change as this is a critical contribution at the point of the crew / system interface. Conservative autothrust logic could reject a single failure, and with a dual system disengage the autos, further encouraging the crew to fly the aircraft – to change their plan. In addition, appropriate autothrust logic would mitigate any error in selecting TOGA; compare this with other types.
Yes we should simply fly the aircraft, but flying is not simple; we have to think about it and that requires knowledge, knowhow, and a situation / system which tolerates error.
But just consider that for any one of the reasons in previous posts the crew were not able to simply fly the plane, or at least not in the manner which you appear to assume.
This ‘inability’ is a combination of human behaviour and the situation in which the crew might find themselves. Change either one or both, then errors and mistakes might be avoided. Humans are difficult to change, and if as in this instance there few or any existing attributes to change, then change may not be an option; thus change the situation.
Arguably the training captain should have foreseen the risks and planned accordingly, so too his management, ATC, the regulators, and at the root of all this the manufacturer who’s design could have avoided the particular circumstances.
Given the choice I would seek a design change as this is a critical contribution at the point of the crew / system interface. Conservative autothrust logic could reject a single failure, and with a dual system disengage the autos, further encouraging the crew to fly the aircraft – to change their plan. In addition, appropriate autothrust logic would mitigate any error in selecting TOGA; compare this with other types.
Yes we should simply fly the aircraft, but flying is not simple; we have to think about it and that requires knowledge, knowhow, and a situation / system which tolerates error.
Guest
Posts: n/a
PJ2
The underlying and systemic soft spot in the industry is and has been economics. Deregulation (and its results, heightened competition), has produced an environment that has surprisingly taken quite some time to show itself. I credit an institutional and historic respect for safety, not any recognized obedience to a proper model. Of course fares are too low. The problem is that no one dare raise them, since all carriers are close to the edge with cash and/or credit.
To me, debt should be restricted to capital assets, not payroll. With all the talent in this cockpit, one is taken back by the outcome. Colgan comes to mind, as other rather basic aspects of loss of control gain headlines. I can't say the economic momentum bodes well, as it's rather obvious too many carriers chase the residual core of the client base.
Success breeds competition. Great success breeds ruinous competition.
I can't think of an easy way out, and in our case, cutting overhead produces some rather pitiful headlines.
bear
The underlying and systemic soft spot in the industry is and has been economics. Deregulation (and its results, heightened competition), has produced an environment that has surprisingly taken quite some time to show itself. I credit an institutional and historic respect for safety, not any recognized obedience to a proper model. Of course fares are too low. The problem is that no one dare raise them, since all carriers are close to the edge with cash and/or credit.
To me, debt should be restricted to capital assets, not payroll. With all the talent in this cockpit, one is taken back by the outcome. Colgan comes to mind, as other rather basic aspects of loss of control gain headlines. I can't say the economic momentum bodes well, as it's rather obvious too many carriers chase the residual core of the client base.
Success breeds competition. Great success breeds ruinous competition.
I can't think of an easy way out, and in our case, cutting overhead produces some rather pitiful headlines.
bear
PEI_3721;
I think your suggestions are excellent, perhaps more along the lines that safetypee was considering - "change the changeable things" in terms of aircraft system design, if I am interpreting his post correctly.
The twp comments I would have and I think you have anticipated it are, such a resilience in single-point system failures (which is partially what redundancy intends) is not easy to achieve so does have its limitations, first in terms of the ultimate reference point, which is resolved perhaps by voting. I am not a systems or computer expert so have to leave that point there but there are others here including Professor Ladkin who has discussed these solutions and their problems, in the past. The second point I might offer is the "swatting at flies" metaphor - which can eventually be successful but is very detailed work... By this I mean, the "RA problem" is fixed, now what about the next single-point item such as the R25 Air-Ground relay in the MD82 which caused the failure of the take-off configuration warning in the Spanair MD82 at Madrid? How does the industry deal with the present pitot tube failure issues where three independant systems failed? There are suggestions of course, and they work, but in such thinking are we not in danger of an extended "monitoring" regress?
We can analyze the notion of redundancy and single-point failure items for a very long time and will continue to find examples in complex systems because that is their nature. The key is resiliency, not brittleness even in failure. The A320 autoflight system has partially handled this particular failure (speed permitted to decrease below Vref) and the accident would likely not have occurred but the design may fail (and has) in other ways perhaps more subtle; mode confusion and the reluctance of crews to disconnect an autoflight system they don't fully understand, when it is mis-behaving (doing something they don't understand, don't like and can't/won't counter) is such an example. The Armavia and Gulfair go-around accidents are two such examples. I have seen others.
I think your suggestions are excellent, perhaps more along the lines that safetypee was considering - "change the changeable things" in terms of aircraft system design, if I am interpreting his post correctly.
The twp comments I would have and I think you have anticipated it are, such a resilience in single-point system failures (which is partially what redundancy intends) is not easy to achieve so does have its limitations, first in terms of the ultimate reference point, which is resolved perhaps by voting. I am not a systems or computer expert so have to leave that point there but there are others here including Professor Ladkin who has discussed these solutions and their problems, in the past. The second point I might offer is the "swatting at flies" metaphor - which can eventually be successful but is very detailed work... By this I mean, the "RA problem" is fixed, now what about the next single-point item such as the R25 Air-Ground relay in the MD82 which caused the failure of the take-off configuration warning in the Spanair MD82 at Madrid? How does the industry deal with the present pitot tube failure issues where three independant systems failed? There are suggestions of course, and they work, but in such thinking are we not in danger of an extended "monitoring" regress?
We can analyze the notion of redundancy and single-point failure items for a very long time and will continue to find examples in complex systems because that is their nature. The key is resiliency, not brittleness even in failure. The A320 autoflight system has partially handled this particular failure (speed permitted to decrease below Vref) and the accident would likely not have occurred but the design may fail (and has) in other ways perhaps more subtle; mode confusion and the reluctance of crews to disconnect an autoflight system they don't fully understand, when it is mis-behaving (doing something they don't understand, don't like and can't/won't counter) is such an example. The Armavia and Gulfair go-around accidents are two such examples. I have seen others.
Come on guys I am with Hetfield on this one! This was demonstrably one of the worst examples of airmanship in recent years. We can philosophise about inept management and imperfect aircraft systems design for ever.
At the end of the day a training captain in what is right now probably europe's fastest growing national carrier (Istanbul THY's base is technically in Europe if not the EU) managed to crash a perfectly flyable aircraft.
PEI 3721 says that two of the crew were being trained. Perhaps I have it wrong, but I understood that there was an experienced safety FO on the jump seat.
Teaching new FOs to recover from a hot and high approach is standard stuff in my airline. However there is a point where you throw it away and go around. Trial and error has indentified this point as 1000 feet in IMC.
This is not even a particularly good example of economics dumbing things down. Turkish requires all their pilots to be graduates and as I understand it can afford to be pretty selective at least in their market. The total hours and flying experience in the cockpit are unlikely to have been the problem.
Much as we may all chafe at the concept of rules and regulations, FOQA systems and safety pilots are there for a good reason, despite what we might think we learnt about airmanship on a dark and windy night in a turboprop.
Nothing like a good debate on Pprune (and this is nothing like it)!
At the end of the day a training captain in what is right now probably europe's fastest growing national carrier (Istanbul THY's base is technically in Europe if not the EU) managed to crash a perfectly flyable aircraft.
PEI 3721 says that two of the crew were being trained. Perhaps I have it wrong, but I understood that there was an experienced safety FO on the jump seat.
Teaching new FOs to recover from a hot and high approach is standard stuff in my airline. However there is a point where you throw it away and go around. Trial and error has indentified this point as 1000 feet in IMC.
This is not even a particularly good example of economics dumbing things down. Turkish requires all their pilots to be graduates and as I understand it can afford to be pretty selective at least in their market. The total hours and flying experience in the cockpit are unlikely to have been the problem.
Much as we may all chafe at the concept of rules and regulations, FOQA systems and safety pilots are there for a good reason, despite what we might think we learnt about airmanship on a dark and windy night in a turboprop.
Nothing like a good debate on Pprune (and this is nothing like it)!
Join Date: Apr 2003
Location: Germany
Age: 76
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
Hmm...
I don't think the Dutch controllers would have given a clearance that meant absolutely that the accident crew would have had to intercept the GS from above, otherwise that should have come up before now.
On the other hand, at six miles the glideslope would be at about one thousand eight hundred feet, assuming a 3° slope and 300 feet per nautical mile, with Schiphol just slightly below sea level.
Does anyone know the exact phraseology of the approach clearance given by Schiphol to the accident crew?
On the other hand, at six miles the glideslope would be at about one thousand eight hundred feet, assuming a 3° slope and 300 feet per nautical mile, with Schiphol just slightly below sea level.
Does anyone know the exact phraseology of the approach clearance given by Schiphol to the accident crew?
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by chuks
Does anyone know the exact phraseology of the approach clearance given by Schiphol to the accident crew?
According to radar data the localizer of the instrument landing system was intercepted at approximately 5,5 NM from the runway threshold.
Guest
Posts: n/a
lederhosen
This accident could well be an example of financial pressure, especially in a growing carrier. Economics has a multitude of ways to stress a firm. This a/c was #1 to land in front of two following (more?). g/a is a bag of snakes, financially, schedule wise, ATC wise, etc. Not saying this crew wasn't considering a way out, I am saying a g/a is not a good 'choice' in today's climate. Flying full loads is wonderful, and on time performance is swell, but when that is a must to do better than break even, the pressure is there. Two things: Eustress. Eu ~ beneficial. Even 'good' stress is still stress. There is an expression in the States. "The Hurrier I go, the Behinder I get." I've taken short cuts many times, sometimes without even knowing it. Also, there are some aspects of five-way CRM that need understanding here, as well.
rgds.
bear
This accident could well be an example of financial pressure, especially in a growing carrier. Economics has a multitude of ways to stress a firm. This a/c was #1 to land in front of two following (more?). g/a is a bag of snakes, financially, schedule wise, ATC wise, etc. Not saying this crew wasn't considering a way out, I am saying a g/a is not a good 'choice' in today's climate. Flying full loads is wonderful, and on time performance is swell, but when that is a must to do better than break even, the pressure is there. Two things: Eustress. Eu ~ beneficial. Even 'good' stress is still stress. There is an expression in the States. "The Hurrier I go, the Behinder I get." I've taken short cuts many times, sometimes without even knowing it. Also, there are some aspects of five-way CRM that need understanding here, as well.
rgds.
bear
Guest
Posts: n/a
Truth is AMS ATC can be very sharp. Sometimes a bit too sharp.
In similar circumstances to the accident flight, I've had "follow the glide-you'll get the localiser in a minute".
Given that the ILS platform alt. at AMS is often 2000', if you are high on the ILS, you are not just high.
You are high, close in. Quite a difference from being high at 9 miles for example.
Much less room to sort it out and much too near to all those lovely GPWS triggers.
Just my 2p worth.
chuks. Oh yes they would.
In similar circumstances to the accident flight, I've had "follow the glide-you'll get the localiser in a minute".
Given that the ILS platform alt. at AMS is often 2000', if you are high on the ILS, you are not just high.
You are high, close in. Quite a difference from being high at 9 miles for example.
Much less room to sort it out and much too near to all those lovely GPWS triggers.
Just my 2p worth.
chuks. Oh yes they would.
Join Date: Dec 2008
Location: DXB
Posts: 513
Likes: 0
Received 0 Likes
on
0 Posts
I don't think the Dutch controllers would have given a clearance that meant absolutely that the accident crew would have had to intercept the GS from above, otherwise that should have come up before now.
The trainee pilot was sent short, high and fast on the ILS. Identifying this and rejecting it within a high pressure, high traffic environment isn't an easy thing. The captain should have made a point in going around at 2000' since nothing cleared him to go lower. It's obviously not so simple since a majority of crews just go for it.
chucks, you might rightly think you are capable to safely perform this kind of interception, but accepting it is just comforting ATC in pushing the limits further. Making a point on the use of acronyms in a forum is one thing, but this would be more useful in the air. Intercepting a GS from above beyond the FAP should be a no go for all of us
Join Date: Apr 2003
Location: Germany
Age: 76
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
Time travellers alert!
Why do I detect such a high Vz or -W in some of these posts?
If this accident crew had been given an intercept that put them at their lowest cleared altitude still above the glideslope then that would certainly be a factor in this accident but I guess that is something we will need to find out about from the yet-to-be-published report.
There's a difference between a "tight turn-on" when you have no chance of getting hold of that slippery GS signal before all you know is that it's somewhere beneath you after you have reached your minimum approach altitude, pretty hopeless in both practical and legal terms, yes, and one where you have to step lively to catch it before it gets away while simultaneously configuring your aircraft. The first one is dangerous, stupid and illegal, the second is doable but sub-optimal and I think most of us know the difference between the two.
Whatever nasty tricks the beancounters and ATC play on us, the Captain is still the final authority as to the safe conduct of the flight with the FO required to speak up and advise him, so that sometimes we are paid to just do a go-around and deal with the consequences of that rather than try to salvage a situation that has become unsafe.
It might be that while your operation has an SOP that states you must be stabilised by one thousand feet AGL you have a reasonable idea that is just eyewash in the Ops Manual; they really want you to just get to the gate on time and never mind how. Two things argue against doing that:
1. If you get caught breaking the rules those people you were trying to "help" will throw you to the wolves without a second thought.
2. You are going to be the first people to the scene of the accident!
If this accident crew had been given an intercept that put them at their lowest cleared altitude still above the glideslope then that would certainly be a factor in this accident but I guess that is something we will need to find out about from the yet-to-be-published report.
There's a difference between a "tight turn-on" when you have no chance of getting hold of that slippery GS signal before all you know is that it's somewhere beneath you after you have reached your minimum approach altitude, pretty hopeless in both practical and legal terms, yes, and one where you have to step lively to catch it before it gets away while simultaneously configuring your aircraft. The first one is dangerous, stupid and illegal, the second is doable but sub-optimal and I think most of us know the difference between the two.
Whatever nasty tricks the beancounters and ATC play on us, the Captain is still the final authority as to the safe conduct of the flight with the FO required to speak up and advise him, so that sometimes we are paid to just do a go-around and deal with the consequences of that rather than try to salvage a situation that has become unsafe.
It might be that while your operation has an SOP that states you must be stabilised by one thousand feet AGL you have a reasonable idea that is just eyewash in the Ops Manual; they really want you to just get to the gate on time and never mind how. Two things argue against doing that:
1. If you get caught breaking the rules those people you were trying to "help" will throw you to the wolves without a second thought.
2. You are going to be the first people to the scene of the accident!
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by chuks
If this accident crew had been given an intercept that put them at their lowest cleared altitude still above the glideslope then that would certainly be a factor in this accident but I guess that is something we will need to find out about from the yet-to-be-published report.
All factors here are relevant to this accident. I am of the 'old school' - the Captain is ultimately responsible. However, there are many here of the 'why because' school who argue in a larger arena.
In my opinion vectoring at AMS needs 'tweaking'.
Join Date: Dec 2008
Location: DXB
Posts: 513
Likes: 0
Received 0 Likes
on
0 Posts
According to the report:
hetfield, do you seriously think that this crew was the only one to accept approaching under these conditions? there must be a lot of GA in AMS!
- The standard procedure for runway 18R prescribes that the aircraft is lined up at least 8 NM of the runway threshold at an altitude of 2000 feet.
- Air traffic control is allowed to offer a line up between 5 and 8 NM of the threshold and instructed the crew in such way that the aircraft lined up at approximately 6 NM at an altitude of 2000 feet.
- Whether the line up is always performed at 2000 feet, regardless of the distance to the threshold being 5 of 8 NM, is still under investigation.
- According to radar data the localizer of the instrument landing system was intercepted at approximately 5,5 NM from the runway threshold.
- The standard procedure is that the glide path is approached from below.
- Because of the shorter line up for the runway at 2000 feet the glide path had to be approached from above.
hetfield, do you seriously think that this crew was the only one to accept approaching under these conditions? there must be a lot of GA in AMS!
Guest
Posts: n/a
ATC were definately 'packing' that day.
I knew there was a problem as we were catching up the no.2 too fast. (plainly he was catching the Turk too fast.)
We (no.3) were given 160 kt, but I could see on the TCAS it was still too fast especially given our distance out with two ahead. There just wasn't the room.
So we slowed to less than 160 to try and keep back from the Transavia.
I knew there was going to be trouble and sensed that someone would be going around. I was trying to make sure it wasn't us! They seemed to have packed too tight.
I was tickled when the Transavia was sent round and thought our trouble was over as there would now be a nice gap.
Not so tickled when I learned why.
Our intercept was working nicely, btw. Maybe they'd tried squeezing the Turk in ahead of us? Hard to tell.
I knew there was a problem as we were catching up the no.2 too fast. (plainly he was catching the Turk too fast.)
We (no.3) were given 160 kt, but I could see on the TCAS it was still too fast especially given our distance out with two ahead. There just wasn't the room.
So we slowed to less than 160 to try and keep back from the Transavia.
I knew there was going to be trouble and sensed that someone would be going around. I was trying to make sure it wasn't us! They seemed to have packed too tight.
I was tickled when the Transavia was sent round and thought our trouble was over as there would now be a nice gap.
Not so tickled when I learned why.
Our intercept was working nicely, btw. Maybe they'd tried squeezing the Turk in ahead of us? Hard to tell.
Join Date: May 2007
Location: Western Pacific
Posts: 721
Likes: 0
Received 0 Likes
on
0 Posts
Regarding the ATC issue, I think that it is just one of the threats that needed to be managed on that flight, along with weather, being in a training environment & dozens of others. A causal factor maybe, but the real problem. The managing the threats issue is the same on every flight. For me, the real crux of the matter is the basic operation of the aircraft.
There are two predominant schools of thought in that area. One is that the pilot is a necessary evil, a liability if you like, who invariably makes incorrect decisions & can get easily overloaded in high workload situations. In this scenario, technology is the final line of defense. The goal of this camp is to restrict the input of the fallible human to the minimum & to eventually get the pilot out of the cockpit (literally) as soon as the flying public will accept it. As it is the last line of defense, the failure of technology is unacceptable. Hence we get talk of improving the technology to prevent failure, along with multiple redundancy in order to make everything fail safe.
The opposing view is that the pilot is the last line of defense, where technology is used as an aid, along with rigorous training to ensure the highest level of safety & efficiency. In this scenario the failure of technology should be more often than not an inconvenience, rather than a major problem.
While the people in the first camp genuinely believe that their paradigm is the safest option, the higher initial cost of technology is sold to the hard-nosed, profit driven airline executive on future cost savings, rather than safety. Initial & recurrent pilot training is easier & less time consuming, so is therefore cheaper. With easier & cheaper training you can increase your pool of potential candidates & therefore can drive down pilot terms & conditions due to excess supply. Another cost saving. However, in the financial basket case that is commercial aviation, the enthusiasm for cost saving has driven a reliance on & use of technology way past what it was
designed for & is currently capable of.
Pilots these days are taught 100% reliance on the autoflight system & 100% adherence to the SOP’s. This is the modern airline’s training department answer to reducing costs, as demanded by management. They can turn out pilots with less simulator training & just a handful of line training sectors – and they do! This is fine when all is going well, but this accident is just another example of what can happen when something in the system fails.
Some believe that the problem can be solved by adding more complexity & redundancy, but I believe that the answer is to once again teach pilots to fly, not just manage systems. That means not just reading the FMA, but actively checking you have actually got what it says you have. It means constantly monitoring attitude, altitude, airspeed & power. It means guarding the control column & thrust levers when configuring on approach (& until flaps up on departure). Moving thrust levers are designed for tactile feedback of what the power is doing, but you are getting very little feedback if you hand isn’t on them! It means not regarding the autoflight system as another pilot & completely turning the aircraft over to it when you engage it. Single engine in the 737 sim is a good way of identifying these pilots. With no rudder control or autothrottle when the autopilot is engaged in single engine ops, you can soon see who hands complete control to the automatics. The autopilot goes on & very soon after there are speed control issues, or the aircraft starts turning against the heading bug. Almost always the autopilot is considered to have failed & is taken back out, when all that would have sufficed is to add or subtract some power, or apply the correct amount of rudder.
In my opinion a safe landing could have been accomplished at Schiphol. It may have been an unstable approach, but a safe landing was achievable. All that was required was that when the aircraft finally intercepted the G/S and with the speed approaching the bugged speed, for someone to recognize that the thrust levers weren’t coming up as they should & to manually push them up tp achieve the reference N1. But this would have required a human to be ‘flying’ the aircraft & not handing that responsibility over to the autopilot that was engaged. It would have required a hand on the thrust levers & an awareness of the airspeed. That simple & that complex.
There are two predominant schools of thought in that area. One is that the pilot is a necessary evil, a liability if you like, who invariably makes incorrect decisions & can get easily overloaded in high workload situations. In this scenario, technology is the final line of defense. The goal of this camp is to restrict the input of the fallible human to the minimum & to eventually get the pilot out of the cockpit (literally) as soon as the flying public will accept it. As it is the last line of defense, the failure of technology is unacceptable. Hence we get talk of improving the technology to prevent failure, along with multiple redundancy in order to make everything fail safe.
The opposing view is that the pilot is the last line of defense, where technology is used as an aid, along with rigorous training to ensure the highest level of safety & efficiency. In this scenario the failure of technology should be more often than not an inconvenience, rather than a major problem.
While the people in the first camp genuinely believe that their paradigm is the safest option, the higher initial cost of technology is sold to the hard-nosed, profit driven airline executive on future cost savings, rather than safety. Initial & recurrent pilot training is easier & less time consuming, so is therefore cheaper. With easier & cheaper training you can increase your pool of potential candidates & therefore can drive down pilot terms & conditions due to excess supply. Another cost saving. However, in the financial basket case that is commercial aviation, the enthusiasm for cost saving has driven a reliance on & use of technology way past what it was
designed for & is currently capable of.
Pilots these days are taught 100% reliance on the autoflight system & 100% adherence to the SOP’s. This is the modern airline’s training department answer to reducing costs, as demanded by management. They can turn out pilots with less simulator training & just a handful of line training sectors – and they do! This is fine when all is going well, but this accident is just another example of what can happen when something in the system fails.
Some believe that the problem can be solved by adding more complexity & redundancy, but I believe that the answer is to once again teach pilots to fly, not just manage systems. That means not just reading the FMA, but actively checking you have actually got what it says you have. It means constantly monitoring attitude, altitude, airspeed & power. It means guarding the control column & thrust levers when configuring on approach (& until flaps up on departure). Moving thrust levers are designed for tactile feedback of what the power is doing, but you are getting very little feedback if you hand isn’t on them! It means not regarding the autoflight system as another pilot & completely turning the aircraft over to it when you engage it. Single engine in the 737 sim is a good way of identifying these pilots. With no rudder control or autothrottle when the autopilot is engaged in single engine ops, you can soon see who hands complete control to the automatics. The autopilot goes on & very soon after there are speed control issues, or the aircraft starts turning against the heading bug. Almost always the autopilot is considered to have failed & is taken back out, when all that would have sufficed is to add or subtract some power, or apply the correct amount of rudder.
In my opinion a safe landing could have been accomplished at Schiphol. It may have been an unstable approach, but a safe landing was achievable. All that was required was that when the aircraft finally intercepted the G/S and with the speed approaching the bugged speed, for someone to recognize that the thrust levers weren’t coming up as they should & to manually push them up tp achieve the reference N1. But this would have required a human to be ‘flying’ the aircraft & not handing that responsibility over to the autopilot that was engaged. It would have required a hand on the thrust levers & an awareness of the airspeed. That simple & that complex.
Last edited by Oakape; 3rd Feb 2010 at 11:27.
Guest
Posts: n/a
Oakape
You put that 'two camps' issue quite well there.
I'm firmly in the second group. I use the autos but they are there to help me, not the other way around.
If I'd been in that Turk 737, I'd have armed the approach and then disconnected AP and AT, If it looked like I could catch the GS in time with a bit of a pitch forward, all well and good. Otherwise it's a GA.
Trouble with all this automation is that folk forget, or are too scared, to hand fly.
The increase in 'loss of control' accidents is the inevitable result.
So offset these accidents against the savings you outlined.
Can't be much of a bargain, really.
You put that 'two camps' issue quite well there.
I'm firmly in the second group. I use the autos but they are there to help me, not the other way around.
If I'd been in that Turk 737, I'd have armed the approach and then disconnected AP and AT, If it looked like I could catch the GS in time with a bit of a pitch forward, all well and good. Otherwise it's a GA.
Trouble with all this automation is that folk forget, or are too scared, to hand fly.
The increase in 'loss of control' accidents is the inevitable result.
So offset these accidents against the savings you outlined.
Can't be much of a bargain, really.
Trouble with all this automation is that folk forget, or are too scared, to hand fly.
New first officers straight from type rating training in the simulator, will have experienced very little raw data hand flying since the majority of training is using automatics - even for simple flying, like a circuit. Loss of confidence sets in and soon the pilot convinces himself that twiddling the automatics and pressing buttons is easier than hand flying.
It probably is - but that is not the point. Someone who lacks confidence in his own ability to hand fly accurately day or night, VMC or IMC, is not the type you want up front on the inevitable dark and stormy night and the autopilot suddenly decides to go on strike.