AF447 wreckage found

[deSitter]

It is absurd to posit that automation and crew disconnection have increased safety. The modern turbofan engine, materials engineering advances, and real-time weather forecasting are primarily responsible for airline safety improvements. Having a crew that can't fly an airplane is obviously not helping safety. I know 225 people who would have agreed.

[bubbers44]

I agree, automation, if it had continued working, was the only way these guys could have survived. Their flying skills killed everybody because they didn't have any basic skills.

[deSitter]

Oh I agree there! Automation is wonderful when there are no pilots aboard!

[DozyWannabe]

Originally Posted by deSitter

It is absurd to posit that automation and crew disconnection have increased safety.

But it is disingenuous to blame automation as the primary factor in what you term "crew disconnection". The real problem lies in the attitude of the airlines and their senior management in the endless drive to increase profits (and by extension bonuses and shareholder dividends) by cutting costs wherever possible - of which abusing automation (which despite the hyperbole has always been intended as a pilot *aid*, not a pilot replacement) is merely one facet.

[captdaddy]

Can somebody please tell me what the pilots were seeing in front of them and not what the FDR and CVR tell us was going on ?....without KNOWING (not guessing or assuming) what was being DISPLAYED to the cockpit I think it is erroneous and even irresponsible to make a judgement about how inept a crew was...a blacked out sim with blank and contradictory screens is alot different than gauges that are pegged out either at zero or max....that's the problem I have with all video/LCD displays where some main computer decides to just not send info to various screens because you are outside the "box" of a particular regime of flight...but again the point is NOBODY has been able to establish WHAT was being displayed to the COCKPIT.

[deSitter]

That's a patently absurd argument. Minus the gewgaws of automation, what training would a pilot receive? Surely how to combat a stall - or even simply to understand the risk! I get the strong feeling that these modern crew lack basic knowledge of flight dynamics, as well as the simple kinesthetic understanding of being in charge of a large and fast machine. The entire issue of the sidestick is absolutely revealing - the body is completely disconnected from the machine - you don't even fly it, you point it here and there and wait for error reports. It's a video game in the sky.

Let's not forget that last year, a Libyan crew flew a perfectly good A330 right into the ground.

The entire philosophy of cheapness coupled to gizmoism is responsible. Everybody's an expert. Everyone gets a trophy. Don't we all feel good? But it's cold at the bottom of the sea.

[Dalex64]

That sounds like an argument for putting cameras in the cockpit.

Would it be so bad if they were only pointed at the instrumentation?

[dlcmdrx]

Oh I agree there! Automation is wonderful when there are no pilots aboard!

Is this really a pilots forum??

Go and check your history, dlcmdrx, especially regarding Perpignan. And Quito, for that matter. When you're better educated on both accidents, come back.

Let me guess, you are one of those that say the only fault of perpignan was of the pilots for not being prepared before hand right??

And as far as quito you are one of those that say that the 20 year experienced captain didnt know how to land a plane right??

Check out Bilbao 2001 and come back when you are not so biased in favor of airbus.

[DozyWannabe]

Originally Posted by deSitter

That's a patently absurd argument. Minus the gewgaws of automation, what training would a pilot receive? Surely how to combat a stall - or even simply to understand the risk! I get the strong feeling that these modern crew lack basic knowledge of flight dynamics, as well as the simple kinesthetic understanding of being in charge of a large and fast machine.

But you don't even get your PPL unless you've done stall and spin training and understand the basics to at least some degree. Hell, I was taught about stalling before they even let me up in a Chippy for my first AEF flight!

The problem is that this training is not being maintained by the airlines.

The entire issue of the sidestick is absolutely revealing - the body is completely disconnected from the machine - you don't even fly it, you point it here and there and wait for error reports. It's a video game in the sky.

There hasn't been a new airliner design with directly-connected controls since the '60s. Everything since then has been variations on mechanical and electronic artificial feel while the hydraulics do all the heavy lifting. As I've argued previously, the yoke can be a crutch - just look at the crews who didn't let them go even after it was clear that they'd lost all hydraulic pressure.

If your company doesn't like handflying on the line, pressure them to make you practice it in the sim, or get yourself into a flying club that will allow you to practice maneouvres in your off-duty time.

Let's not forget that last year, a Libyan crew flew a perfectly good A330 right into the ground.

Due to which it could be argued (though I'm not arguing) that the Airbus guy who said they didn't recommend handflying on the line had a point!

The entire philosophy of cheapness coupled to gizmoism is responsible. Everybody's an expert. Everyone gets a trophy. Don't we all feel good? But it's cold at the bottom of the sea.

The only widebody that tried to use old-fashioned techniques exclusively was the DC-10, and that led the designers to miss points of failure that they didn't take into account, like the pressurised air volume being so much larger than anything they'd built to date that a significantly large hole in the fuselage could collapse the floor, under which they'd laid the supposedly redundant controls - all of them.

Funnily enough, one of AA's senior captains didn't like what the Douglas guys said about it being impossible for the DC-10 to lose all hydraulics (and consequently all flight controls), and was able to wangle enough simulator time to learn how to control the thing using differential thrust. His name was Bryce McCormick, and the only reason that there weren't significantly more air crash fatalities in 1972 than there actually were was because he happened to be the captain on the flight on which it happened.

Ultimately engineering is there to solve problems, and the FBW advances were designed to solve the problems of reducing weight and thereby extending range and capacity, exposing less of the hydraulic system to risk by making more use of redundant electronic controls and as an added bonus, using obsolete, reliable computer technology to assist pilots with the workload. Airbus's relative newcomer status and lack of legacy models also meant they could get a jump on flight deck commonality across the range.

To hear you talk you'd think that the FBW systems of both Airbus and Boeing were designed with no pilot input at all, when in fact pilots were heavily involved in the specification for both. You're also conflating FBW with the advent of FMS systems, which predated the A320's arrival in service by 16 years, and were enthusiastically adopted by Boeing (in the 757 and 767) and McDonnell-Douglas (in the MD-11) shortly afterwards.

[deSitter]

Don't lecture me on accident reports. I'm well aware of all the things that can and do go wrong. But I always have confidence that the men and women up front will do their best on bad days. I assume there exists a culture of aviation that has its own internal rules, that are not all based on politics, and that in particular, pilots, like concert pianists, would be especially jealous of the honor of their guild, and would not admit members who were not up to snuff.

[jcjeant]

Hi,

To hear you talk you'd think that the FBW systems of both Airbus and Boeing were designed with no pilot input at all, when in fact pilots were heavily involved in the specification for both

It's many urban rumours about Airbus
Pilots were heavily involved in the specification of Airbus is one of them

[britfrog]

do not make the mistake that all pilots , like all doctors, all lawyers etc are good and professional. As I know from first hand experience of being a check pilot for many years there is an awful amount of dross out there.

[Gretchenfrage]

Dozy, you wrote:

Ul

timately engineering is there to solve problems, and the FBW advances were designed to solve the problems of reducing weight and thereby extending range and capacity, exposing less of the hydraulic system to risk by making more use of redundant electronic controls and as an added bonus, using obsolete, reliable computer technology to assist pilots with the workload.

To hear you talk you'd think that the FBW systems of both Airbus and Boeing were designed with no pilot input at all, when in fact pilots were heavily involved in the specification for both.

First, those two statements somewhat contradict each other. At least it demonstrates the heavy unbalance of input, as the commercial pressure certainly would dictate.
Second, I consent that some pilots were involved in design. But which ones?? Most probably management pilots and technical pilots. Now most experienced line dogs will agree, that they are not bad guys, but somewhat estranged to daily operations.
The statements of some manufacturers when launching new technology emphasized way more on how to protect everything from pilot’s mishandling and gaining weight, than giving pilots the right tools to overcome the threats out there, whatever they may be, wherever they originate.

The big issue with Airbus was and still remains the lack of feedback on controls. They let their aircraft use one single channel to communicate with the pilot, the one through the eye, meaning an intellectual, a serial input to the brain. To a small degree they use the audio channel as well, however in this particular case shows how small: The THS movement is not audio connected, you can’t hear it moving, you need to look at it again, with your single serial channel.

Don’t bother mentioning the weight issue, with columns and thrust levers. We don’t need the absolute direct and precisely interlinked feedback from the different systems, what would make the thing heavy. You can buy a rumble joystick and a simple thrust level duplicator for a few bucks in any game shop (not wanting to implement such a cheap solution remains therefore a matter of pride and principle).
A simple movement in the direction of input from the collegue or from the AP/AT is sufficient. We only need the tactile feedback serving the other channel input to our brain. That one works parallel to the intellectual one, thus not impeding it. Additionally we all know that such input is some factors faster than the intellectual one.

It boggles my mind that this has not been sincerely addressed by the regulators or investigators:

WHY DEPRIVE THE HUMAN OF AN INPUT CHANNEL INTO HIS CONTROLLING DEVICE? A CHANNEL THAT IS OLDER AND MORE INSTINCTIVE AND MUCH FASTER THAN THE ONE HE LATER AQUIRED, THE INTELLECTUAL ONE?

As a pilot being placed into the modern cockpit to supervise and program the automation and to intervene when it screws up, I need all the channels and inputs I can get, especially the parallel ones, as my brain starts working more constrainly in stress.

You can point at the not using the unreliable speed checklist, badly using the stall recovery procedure, not realizing the THS position, not knowing that the stall warning goes out below 60kts, being slow in realizing that the AT was off, the lever position not where the power was, having tocheck on ECAM and click up and then down with it, etc. etc. (all single channel eye-brain operations).

But what bugs me more is the switching of stick priority back and forth, no double inputs, as this is not allowed, the swinging of the stick up-down-left-right, the moving of the thrust to TOGA-idle-TOGA and so forth, the shouting “I have no control”.
It reveals a completely lost PF(no feedback on his tactile channel), a PNF that has no clue what the PF is swinging (again no tactile feedback).
Not that this would be the initial reason for the crash, but to me it certainly points to a huge weak spot of the Airbus design.

I know however that I will be cried down by the lobbyist and all others will shrug their shoulders and say “so what, there are so many ABs flying around and so much money involved, nothing is going to change”.

Let us thus wait for the next pilot error

[jcjeant]

Hi,

What about the next step ....
The use of voice recognition software ... and the pilot will not even have to use his hands and feet ... he will just talk to the aircraft via the microphone ..
Even the disabled can finally be pilots

[Jazz Hands]

Let me guess, you are one of those that say the only fault of perpignan was of the pilots for not being prepared before hand right??

Several things went wrong in Perpignan, not the least of which was the deliberate erosion of safety margins which are there to provide the wiggle-room needed to get yourself out of a fix.

If you ignore the basic preparations, and cut corners, and rush procedures, and perform stunts at 3,000ft which are meant to be done at 12,000ft, and wrongly assume - despite being a test flight - that the aeroplane is fully functional, then it's a bit much to start blaming the automation once you've boxed yourself in and left yourself no time, space or height to work it out. And that's assuming you've been adequately trained in the first place.

I don't have an Airbus bias. I've just read too many accident reports where basic airmanship seems to have gone out of the window.

[Owain Glyndwr]

Gretchenfage:

Second, I consent that some pilots were involved in design. But which ones?? Most probably management pilots and technical pilots. Now most experienced line dogs will agree, that they are not bad guys, but somewhat estranged to daily operations.

I know better than to indulge in arguments with pilots, but since I happen to have known the pilot most involved in the development of the AI FBW system (Gordon Corps), I must take up cudgels on his behalf - he died tragically walking to a crash site in the Himalayas.

Contrary to the somewhat snide slur implied by "not bad guys, but not really up to the job" Gordon was an ex ARB, ex CAA, test pilot who had flown many more types than, I suspect, anyone else writing in this forum. This for passing them off in certification, not joy riding. He probably knew more of the good and bad points of As and Bs and quite a few other manufacturer's designs than any line pilot.

So please - check on facts before denigrating the manufacturers pilots!

[Mimpe]

A bit off track, but as much of the discussion has been on the limitations of automation and the pilot response.

People ridicule the upset recovery system used by Cirrus in its user friendly SE GA aircraft.

I bet if the 3 crew faced with the first 15 -18000 ft of altitud loss had a magic (last resort button ) to press that would rectify the upset, and put it back in S and L flight attitude at 85 % odd throttle , they would have been more than relieved.

[TJHarwood]

As this forum is meandering far more than the main AF447 (thread 5) forum on the key issues, could we please not lose total focus? The industry is safer, and automation has played its part. The significance of AF447, and several other LOC incidents in recent years, is assessing where the industry currently is in the limited circumstances where pilots are left to their own devices. Aircraft are stopping pilots making many mistakes which could turn fatal, and I think most veterans would accept that, but can we say that many commercial airline pilots are ready to now suddenly be handed back manual control and ride to the rescue in the way most could once have done......? We are where we are through AF447, and it is not happy reading.

[britfrog]

pi=s poor piloting

sadly there is a similarity between the 2 accidents, ie the pilot over riding the system, only a raving loony would practice flight at the bottom of the envelope with no altitude or hope of recovery if the sh-t hit the fan, the perpignan accident was very poor piloting nothing else and so was af447 there is no inherent problem with the a/c but sadly there are a few nuts behind the column that need a bit of tightening. the modern regime that insists that the autopilot should fly the a/c has some merit as the a/c can fly happily by itself however the side affect of this is that pilots do not any longer get enough hands on experience and they forget the old adage first aviate then navigate. what kind of a pilot holds an a/c at up to 43% nose up for up to 75 seconds and still keeps the joystick fully aft? and he hasnt realised after this time what is going on? this is beyond reason it is sheer madness. The only sane conclusion, however unlikely, is that he wanted to commit suicide.

[DozyWannabe]

Originally Posted by Gretchenfrage

Second, I consent that some pilots were involved in design. But which ones?? Most probably management pilots and technical pilots. Now most experienced line dogs will agree, that they are not bad guys, but somewhat estranged to daily operations.

As I understood it the group from which requirements were drawn involved a fairly broad cross-section of pilots.

The big issue with Airbus was and still remains the lack of feedback on controls. They let their aircraft use one single channel to communicate with the pilot, the one through the eye, meaning an intellectual, a serial input to the brain. To a small degree they use the audio channel as well, however in this particular case shows how small: The THS movement is not audio connected, you can’t hear it moving, you need to look at it again, with your single serial channel.

I'd venture to say that *your* big issue with Airbus is the lack of tactile feedback. You're not alone, Heino Caesar of Lufthansa felt the same way - however a lot of pilots have no issue whatsoever with it.

You can buy a rumble joystick and a simple thrust level duplicator for a few bucks in any game shop (not wanting to implement such a cheap solution remains therefore a matter of pride and principle).

If you really believe it's that simple, give it a try, and good luck getting that gear certified. "Pride and principle?" - rubbish. In order to pass certification the devices have to be proved to be able to resist failure to a probability ratio of 10^-9 per hour. My CH simming gear went back twice - the first time failing after 6 months, and that was kept in a dry, sealed box, not constantly exposed to the ever-changing temperature and humidity of a line aircraft - though somewhat ironically, the cheapo rumble stick that I used when flying the A320 sim (sadly, I don't have the free time anymore) never gave me a day's trouble.

We only need the tactile feedback serving the other channel input to our brain.

Who do you mean by "we"? If you mean pilots, then it's clear that a significant number consider it a "nice to have" rather than a necessity in day-to-day operations, otherwise Airbus would never have been able to challenge Boeing in the airliner stakes.

If you mean human beings in general, I'd say that it becomes a matter of training and what we become used to. I maintain that tactile feedback is useful in the initial training scenario in order to accustom the pilot with the inputs required to maneouvre the aircraft, and additionally to understand the factors of how air resistance affects the control surfaces at various attitudes and speeds - in line flying, when you're supposed to have all that stuff (as well as stall recognition and recovery) down, I (along with others) don't see it as so much of a necessity.

WHY DEPRIVE THE HUMAN OF AN INPUT CHANNEL INTO HIS CONTROLLING DEVICE? A CHANNEL THAT IS OLDER AND MORE INSTINCTIVE AND MUCH FASTER THAN THE ONE HE LATER AQUIRED, THE INTELLECTUAL ONE?

No need to shout. As I said earlier there are several things to take into account. The A320 project went onto the drawing board in 1982 and was just about ready to enter service by 1988. This meant that in those 6 years (which is by no means a short time to develop an airliner) every new system had to go through a long certification process which, coupled with the engineering complexity maxim, meant that those systems had to be kept as simple as possible.

Combine this with the fact that the hardware specified had to be proven and therefore a few years old at the time. I'd hazard a guess that the hardware was specified around 1985, which meant the venerable (in computer terms) 80186 (introduced 1982) and 68000 (introduced 1979) were pressed into service. They were quite advanced for their time, but even by the standards of the late '80s and early '90s they were fairly low down the pecking order performance-wise. When Boeing caught on to how the airlines were responding to the new FBW offerings and put the 777 on the drawing board, they had more advanced proven technology to work with, and so the complexity of a force-feedback system was less of a challenge. Added to which the FBW concept in general had been proven by Airbus, so Boeing probably had an easier job getting their basic control logic systems certified in the first place.

As a pilot being placed into the modern cockpit to supervise and program the automation and to intervene when it screws up, I need all the channels and inputs I can get, especially the parallel ones, as my brain starts working more constrainly in stress.

What this tells me is that you have an innate distrust of technology, and that colours your perception of the systems - which is quite ironic given that the modern Airbus and Boeing systems are designed to degrade quite gracefully - they've convinced the certification authorities and the regulators that they won't automatically control you into an unstable attitude and then hand you back the controls, yet you seem to be fully convinced that they will.

Now I'm not saying they won't or can't, but I believe the chances of it happening are suitably remote, and I hope that given a competent pilot, the aircraft is capable of being recovered in the highly unlikely event that they ever do.

You can point at the not using the unreliable speed checklist, badly using the stall recovery procedure, not realizing the THS position, not knowing that the stall warning goes out below 60kts, being slow in realizing that the AT was off, the lever position not where the power was, having tocheck on ECAM and click up and then down with it, etc. etc. (all single channel eye-brain operations).

That's a fair amount of conjecture in there, and a couple of outright misapprehensions. They didn't perform the stall recovery procedure *at all*, the PNF clearly called out that A/P and A/THR were off, followed by announcing Alternate Law - and none of us know whether the PNF was ever "head down" reading the ECAM messages, let alone scrolling through them - he certainly seemed to be aware enough of how the PF was controlling the aircraft to tell him several times that he was overcontrolling in pitch and roll.

But what bugs me more is the switching of stick priority back and forth, no double inputs, as this is not allowed, the swinging of the stick up-down-left-right, the moving of the thrust to TOGA-idle-TOGA and so forth, the shouting “I have no control”.

He had no control because he'd stalled the aircraft! Look, regarding the priority switch, the Airbus design (due to the factors I mentioned earlier) was designed to operate in an environment where the crew follow procedure, which is that one pilot should be PF and the other PNF at any given time in 99% of operational situations, and as such is an aid to enforcing CRM and flight deck discipline. Double inputs *are* allowed by the system, but they are summed, meaning that in an emergency situation, the pilots can theoretically command twice normal pitch-and-roll rate in an emergency situation if they co-ordinate properly, and that a pilot can counteract the inputs of an incapacitated pilot in the other seat if the situation is recognised. Compare that to the old yoke system whereby whoever was the strongest decided the direction of the aircraft, or the more modern yoke in the 767 when opposite inputs cause the elevators to move in opposite directions (as EgyptAir 990 appeared to prove).

It reveals a completely lost PF(no feedback on his tactile channel), a PNF that has no clue what the PF is swinging (again no tactile feedback).
Not that this would be the initial reason for the crash, but to me it certainly points to a huge weak spot of the Airbus design.

But that's just it - you're hammering the known aspects of this case into the frame of your particular dislike of the Airbus design. There's no evidence that the PNF was unaware of the inputs being made because he told the PF to ease off several times, and there's no evidence that tactile feedback would have told the PF that he was overcontrolling, and it looks like he simply had no appreciation of how little input is actually required to get positive response from the aircraft at altitude.

I know however that I will be cried down by the lobbyist and all others will shrug their shoulders and say “so what, there are so many ABs flying around and so much money involved, nothing is going to change”.

It's so easy to blame lobbyists and beancounters, because it's not well-known by the public that pilots were involved in the Airbus FBW design, that the march towards pilotless airliners was a fabrication of the press (as was the junk about the Habsheim A320 "thinking it was going to land") and that the politics being played by pilot's unions in the wake of the A320's introduction were largely a response to the cr*p being spouted by the press, and not because they'd flown the A320 and found it wanting.

Let us thus wait for the next pilot error

This wasn't just pilot error, this is a systemic problem affecting the airlines and the industry as a whole. PPLs start their ATPL training knowing how to recognise and recover from a stall - that this knowledge is not periodically revised and enforced is a sad indictment of the real issues that cost-cutting and poor corporate morale produce.