Air France 447 - On topic only!

[Evan]

Off topic:

He did not say "reduce pitch". He did say "move the stick slightly forward".

He said:

(I can't count the folks I've gotten in arguments with- including lots of professional pilots and Evan) on this whole subject if whether the pilots might want to consider moving the stick slightly forward if there is a stall warning.)

Again, I don't have an argument with this. If you get a stall warning, you are applying too much back pressure for the situation you are in, and to avoid a stall you need to reduce it until the situation changes. HOWEVER, if you get a stall warning due to holding or climbing with inadequate power (airspeed), you need to reduce pitch, not just pitch rate. (the PF of AF447 reduced pitch rate to near zero after the stall warning, but did not level off or descend as was needed).

While related, the two things are not the same. In particular, you can very well reduce the AoA and increase pitch at the same time if an increase in airspeed is taking place.

Remember, we are talking about what to do at stall warning here. Approach to stall. And the thing to do, particularly in level flight where holding altitude is critical (meaning that you need to mitigate the altitude excursion, not avoid it completely), is to manage pitch and add power. Manage pitch means quickly assess the situation, reduce the AoA with pitch management while adding power in a way that does not induce pitch-up effects. Correct?

I think that's everybody's position. If you manage pitch correctly you won't hear the stall warning, let alone be involved in a full stall. See how much pitch mismanagement it took to the AF pilots to get a stall warning, and then it took another pitch mismanagement to make it a full stall.

Remember, we are talking about what to do at stall warning here. Not how to fly in general. If you can't fly the plane to begin with, you probably won't do the right thing at stall warning either. But yes, my original point is that 'manage pitch correctly' IS my position as well, so I don't know where that was coming from.

[Evan]

Yes, but in combination with controlling angle of attack.

That is the meaning of 'manage pitch correctly'.

[3WE]

Really? where? when?

In a simulator, every six months, on just about every major airline. It's pretty damn intensive, so I'm told. Perhaps you should try it. (And hand flying?!?!?- maybe ask some pilots on discussion forums if their goal is to maximize their use of Otto...I know and believe what they say)

(And for clarity, while it is intensive, most folks admit that recurrent training does not cover every possible scenario that might occur).

And, I did not call you a teenie bopper. However, I did point out the fact that other people have, and it appears that you still don't know why.

[TeeVee]

In a simulator, every six months, on just about every major airline. It's pretty damn intensive, so I'm told. Perhaps you should try it. (And hand flying?!?!?- maybe ask some pilots on discussion forums if their goal is to maximize their use of Otto...I know and believe what they say)

(And for clarity, while it is intensive, most folks admit that recurrent training does not cover every possible scenario that might occur).

And, I did not call you a teenie bopper. However, I did point out the fact that other people have, and it appears that you still don't know why.

ahhhh so hand-flying a simulator. very different from hand-flying a real a/c in the real world. sims may be very good, but they are not a complete substitute for the real thing. so now, since we, as a society, appear to be too afraid to allow pilots to hand-fly real aircraft, we place them in sims, train them for a few hours per month--maybe--and pray they react appropriately when the real shit hits the real fan in the real plane with real lives onboard and the very real probability of all of those lives getting snuffed out if the poor pilots don't perform.

yes, this is always a possibility, but i think the risk is increased greatly by forcing pilots to use ap's 95-98% of the time for whatever reasons. yes, ap's are more economical, more exact, more precise, more, more, more. yet when faced with fairly common circumstances, such as crappy weather, they disconnect and hand things over to a very bored, probably half-asleep crew, who, despite what is required of them, have not quite been following what good ol HAL has been doing for the past few minutes. next thing ya know, "i have no control over the airplane...we have no speeds, we've lost all speeds...am i going up? no you're going up!..." few minutes later, splash and damn near 300 people are gone.

yeah, keep forcing pilots to use the autopilot based on the ass-umption that despite not having to physically do anything while HAL flies, that they will resist human temptation to let their attention stray. no way you say? which flight was it that was out of contact over the CONUS and flew past their destination while HAL happily flew the plane??? yeah, they were talking about their schedule! BS! and even if they were, they were SO distracted they lost all sense of time and place, not to mention that they couldn't possibly have had a clue as to the state of the aircraft.

as for the name calling crap, i'm too old and too secure to give a rat's ass who said it or when. but you lowered yourself to their level just by referring to it.

again, the personal attacks detract from whatever validity your point may have. and since it is now apparent that you too do not have real world experience flying real world commercial aircraft, i will treat your remarks accordingly.

[Black Ram]

The Human Factors working group formed for the investigation is ready to start its work:

http://www.bea.aero/en/enquetes/flight.af.447/info07september2011.en.php

[Deadstick]

The Human Factors working group formed for the investigation is ready to start its work:
http://www.bea.aero/en/enquetes/flig...ber2011.en.php

Hell, I've hand flown an F14 simulator down to the pavement at Miramar. Didn't have such good luck putting it on the boat though. Kina smashed up the back end. Still wouldn't want to try it for reals.

[Fear_of_Flying]

i think the risk is increased greatly by forcing pilots to use ap's 95-98% of the time for whatever reasons.

I don’t think it follows from the fact that pilots may need more meaningful, hands-on experience in their formative years that they need to be hand-flying commercial jetliners around. The risk is not “greatly increased” by use of AP, it is greatly decreased, because the more automated systems are used, the less vulnerability is placed on the human side of the equation. The risk of pilot error is in fact lowered the more time the computers are in control. Your examples are astronomically rare exceptions to the significant increase in safety provided by automated flight systems.

With technological advances, flying a commercial airliner is more economical, reduces pilot workload, and is much safer. Have problems ensued from the evolution towards more automated aircraft? Of course, and some of these have become more apparent in recent years. Aircraft have become so easy to fly that the training standards can be lowered without a terrible net decrease in safety. Pilots also need to have a better understanding of how the various systems operate in order to respond to problems, but seemingly do not. Finally, automation can fail, so there need to be sufficient safeguards.

But to take these problems and to frame them as making commercial air travel less safe is simply not true. Overall safety has improved with automation, and will continue to improve provided those other issues are not neglected entirely.

Today, the spoilers on an aircraft deploy automatically. Decades ago, pilots were responsible for operating them manually. What happened? Sometimes pilots failed to deploy them, and sometimes they inadvertently deployed them at the wrong time (i.e. Toronto, 1970). We could have kept insisting for years and years that pilots be better trained to prevent this from happening again, or we could do exactly what we did: remove that manual operation from their list of responsibilities. Did each successive measure towards automating aircraft make pilots that much less skilled, or did it simply change their responsibilities, and make flying safer?

[MCM]

The problem now is, that air travel is not getting safter. The accident rate plateaued significantly over the mid '00's (although did decrease in 2010. I don't think 2011 is going to be a particularly good year though). Going back a few years, introduction of automation made a dramatic increase in safety. However, we then saw an increase in accidents due to other causes - specifically, human factors.

Since then, we've corrected a lot of HF failings, and the accident rate decreased more. But we've stopped improving.

So what do we do now? The current accident rate, while it might seem fine now, is far too high for continued growth. Predicted flight numbers in Europe by the European commision, multipled by the current accident rate, gives over 100 aircraft accidents per year - thats two per week. Will the travelling public accept that?

[Fear_of_Flying]

So what do we do now?

Someone asked this very question not long ago, and my opinion, which is admittedly irrelevant, was that technology will have to continue to improve and stay ahead of the curve. In my opinion, it is the technical recommendations that come out of the AF447 accident that will be the most important, not the recommendations for pilot training, failings which are more endemic to the industry. Things like alternate law protections, how the flight system responds to UAS, and stall warning function are examples of what might need to be re-assessed.

Evolution typically involves rapid change followed by a plateau, so the fact that we don't see a positive uninterrupted slope in increased safety doesn't alarm me particularly. It simply means we need to take the next step.

Also, I wonder if those statistics truly reflect airline safety, or if they reflect new airlines using older jets and inconsistent standards in parts of the world that hadn't previously seen much air travel. In the US it seems like it's been forever with only Colgan on our hands (knock on wood). Will the predicted growth in air travel in Europe actually correlate with current accident rates? Not sure.

I too have heard that air travel is expected to double in the coming years, though, which is daunting. Presumably accidents may not just increase proportionately to the increased traffic, but may increase at a faster rate due to added pressures on the industry (cranking out more pilots, strain on air traffic control and maintenance requirements, more corners cut to compete in the growing market, etc.). Should that prove to be the case, and to answer your question, no, the traveling public would not accept two accidents a week, I don't think.

But as I said, I don't think we're heading there necessarily.

[Evan]

So what do we do now?

Regulate, regulate, regulate. What we are seeing now in EVERY industry is a corporate morality based entirely on profit, and there can no longer be any doubt that corporations, in deference to their shareholder value, will NOT regulate themselves. There is no longer any compunction or sense of social responsibility in the corporate ethos.

If we had strong pilot training standards, including human factors training, that were strictly enforced, we would see a reduction in fatalities and hull-loss incidents. Just think of how many avoidable tragedies have occurred lately due to pilot error. Then think about how many are the fault of the equipment. Then do the very simple math.

That is my case for more and stronger regulation. The airlines are rolling in profit so there can no longer be an argument that such regulation would destroy the industry. ANd prices will not rise significantly because market competition must provide a price the consumer will agree to.

I've yet to hear a viable case against it.

[Fear_of_Flying]

Regulate, regulate, regulate. What we are seeing now in EVERY industry is a corporate morality based entirely on profit, and there can no longer be any doubt that corporations, in deference to their shareholder value, will NOT regulate themselves. There is no longer any compunction or sense of social responsibility in the corporate ethos.

If we had strong pilot training standards, including human factors training, that were strictly enforced, we would see a reduction in fatalities and hull-loss incidents. Just think of how many avoidable tragedies have occurred lately due to pilot error. Then think about how many are the fault of the equipment. Then do the very simple math.

That is my case for more and stronger regulation. The airlines are rolling in profit so there can no longer be an argument that such regulation would destroy the industry. ANd prices will not rise significantly because market competition must provide a price the consumer will agree to.

I've yet to hear a viable case against it.

Well, an accident can be very detrimental to the profit of an airline. Accidents can even spell its demise in some cases. Therefore, it is not in the financial interests of airlines to have accidents. I think airlines do what they can to survive - it's been a tough industry to make much headway in the last decade and beyond - and so do they extend themselves to the greatest degree possible to provide safety? Probably not, but I do think they meet current safety requirements, and do make an effort within financial constraints. Some safety requirements may be inadequate, but that's not the fault of corporate greed.

Also, I don't like it when people confuse "more regulation" for "better regulation". That's rather akin to throwing money at a problem - if it doesn't end up in the right places, nothing changes. If a regulation is ineffective, the thing to be done is ask why it's not working, and make it smarter, not keep it and add three more equally ineffective regulations so you can say you've addressed the problem.

A case in point is pilot training. I'm sure better regulations could, to a degree, increase piloting skills. What I'm far less convinced of is that the kind of airmanship we would like to see - which is only called upon one in every zillion flights - is going to be achieved using the model of flight school that is currently in place. That is one of the root problems that a few more regulations about sim time and memory items is not going to change.

Following your logic through to the end, though, we would be far better off with government owned and operated airlines. Then there is no evil profit driving the industry, just a sincere desire to make air travel the safest and most enjoyable experience it can possibly be.

[TeeVee]

I don’t think it follows from the fact that pilots may need more meaningful, hands-on experience in their formative years that they need to be hand-flying commercial jetliners around. The risk is not “greatly increased” by use of AP, it is greatly decreased, because the more automated systems are used, the less vulnerability is placed on the human side of the equation. The risk of pilot error is in fact lowered the more time the computers are in control. Your examples are astronomically rare exceptions to the significant increase in safety provided by automated flight systems.

With technological advances, flying a commercial airliner is more economical, reduces pilot workload, and is much safer. Have problems ensued from the evolution towards more automated aircraft? Of course, and some of these have become more apparent in recent years. Aircraft have become so easy to fly that the training standards can be lowered without a terrible net decrease in safety. Pilots also need to have a better understanding of how the various systems operate in order to respond to problems, but seemingly do not. Finally, automation can fail, so there need to be sufficient safeguards.

But to take these problems and to frame them as making commercial air travel less safe is simply not true. Overall safety has improved with automation, and will continue to improve provided those other issues are not neglected entirely.

Today, the spoilers on an aircraft deploy automatically. Decades ago, pilots were responsible for operating them manually. What happened? Sometimes pilots failed to deploy them, and sometimes they inadvertently deployed them at the wrong time (i.e. Toronto, 1970). We could have kept insisting for years and years that pilots be better trained to prevent this from happening again, or we could do exactly what we did: remove that manual operation from their list of responsibilities. Did each successive measure towards automating aircraft make pilots that much less skilled, or did it simply change their responsibilities, and make flying safer?

i think you've missed the point entirely. this is not about HAL vs. the humans. it's about the inescapable truth that until computers and avionics have evolved to 3 or 4 generations from now, there is always going to be the need for a human in the cockpit. and the less experience that human has in flying the real deal, the more likely she/he is to screw up on a cataclysmic scale and kill her/himself and everyone else onboard.

there is no denying that automation is here to stay in aviation. however, you are dead wrong for thinking or advocating that it should be used at all times possible and rely on the "training" and sim time pilots have for those times when they have to take the controls. that is precisely what causes a lot of these accidents--pilots that freak out in the face of multiple system failure, or perhaps more precisely, multiple system shutdown due to a single item failing, e.g., pitot tubes.

[TeeVee]

Regulate, regulate, regulate. What we are seeing now in EVERY industry is a corporate morality based entirely on profit, and there can no longer be any doubt that corporations, in deference to their shareholder value, will NOT regulate themselves. There is no longer any compunction or sense of social responsibility in the corporate ethos.

If we had strong pilot training standards, including human factors training, that were strictly enforced, we would see a reduction in fatalities and hull-loss incidents. Just think of how many avoidable tragedies have occurred lately due to pilot error. Then think about how many are the fault of the equipment. Then do the very simple math.

That is my case for more and stronger regulation. The airlines are rolling in profit so there can no longer be an argument that such regulation would destroy the industry. ANd prices will not rise significantly because market competition must provide a price the consumer will agree to.

I've yet to hear a viable case against it.

often when folks argues for more regulation they fail to realize the real costs. it's not about simply enacting more rules and regs. it's about investigative expenses, prosecutorial expenses, salaries, retirement, benefits etc, etc.

now, when you talk about criminal consequences, you sometimes get responses like Evan's when it was announced that certain entities were being prosecuted in, what was it, the Concorde aftermath? his response was something along the lines of "there is nothing to be gained from criminal prosecution" or something like that.

Why? who knows? clearly fines are worthless. the possibility of being sued and having to pay tens of millions of dollars (another thing Evan despises and rails against--lawyers) obviously does nothing.

so yeah, the only thing that MIGHT work is criminal prosecution. i read that the poor slob in peru that failed to remove the tape from the 757 that later crashed due to inoperable pitots, was prosecuted criminally and there were those that said it was stupid to jail a simple maintenance worker who "made a mistake." well i say, if your mistake was so huge that it cost tens of people their lives, you go to jail. then, maybe, your cohorts will be more careful and do their jobs.

i think we have enough regulations, although some might need to be tweaked a bit. what we need is a bigger hammer and the authority to drop said hammer on a few choice nuts.

[Fear_of_Flying]

i think you've missed the point entirely. this is not about HAL vs. the humans. it's about the inescapable truth that until computers and avionics have evolved to 3 or 4 generations from now, there is always going to be the need for a human in the cockpit. and the less experience that human has in flying the real deal, the more likely she/he is to screw up on a cataclysmic scale and kill her/himself and everyone else onboard.

there is no denying that automation is here to stay in aviation. however, you are dead wrong for thinking or advocating that it should be used at all times possible and rely on the "training" and sim time pilots have for those times when they have to take the controls. that is precisely what causes a lot of these accidents--pilots that freak out in the face of multiple system failure, or perhaps more precisely, multiple system shutdown due to a single item failing, e.g., pitot tubes.

I actually didn't miss the point of what you were trying to say, I disagree with it, as well as I disagree with how you characterize what I am saying.

Picking up on what Evan said previously, you simply don't fly a modern jetliner by hand flying it around the countryside, and any suggestion that we should start doing so is absurd.

In my opinion, pilots need to learn their trade in a way that gives them real experience and meaningful flying hours so they can deal with an emergency when called upon. But once they're flying a passenger jet, it's time to let the automation do its job.

[TeeVee]

I actually didn't miss the point of what you were trying to say, I disagree with it, as well as I disagree with how you characterize what I am saying.

Picking up on what Evan said previously, you simply don't fly a modern jetliner by hand flying it around the countryside, and any suggestion that we should start doing so is absurd.

In my opinion, pilots need to learn their trade in a way that gives them real experience and meaningful flying hours so they can deal with an emergency when called upon. But once they're flying a passenger jet, it's time to let the automation do its job.

and therein lies your problem...it's called a contradiction. stop for a moment and thing about the text i highlighted. then come back and tell me it makes sense.