Air France 447 - On topic only!

[MCM]

We agree then. The PF should stabilize the a/c before doing ANYTHING else. The PNF is there for a reason. Whatever happened to CRM?

Yes, that is the theory.

It takes an incredible amount of discipline as PF to focus on nothing but aircraft flight path while the PNF analyses the problem. Particularly when lights are flashing, sirens are going, and systems are coming and going.
, and the aircraft appears to be performing ok.

It is, unfortunately, human nature.

Continual training and learning are they key, but how do you know how you will react when the adrenaline hits and you get the fright of your life?

We can talk CRM all we want, but please don't pretend it is an easy thing to do when the shit really hits the fan. Must it be done - yes. But don't say its easy.

I would also suspect it is far more likely when there is no turbulence at all - in turbulence the PF would already be paying close attention to the flight path. In smooth air, it would be very easy to have lowered that attention.

I will be VERY interested to see to what extent the BEA think that the "stationary thrust levers" have played in this accident.

[Fear_of_Flying]

Because the press is not what is used to be.

And what did it used to be? These blanket characterizations of "the press" are pointless. The press has changed over time, in some ways for the better, in some ways for the worse, but its essential role is as important or more important than ever.

With respect to this particular incident, the reports have become increasingly more detailed and credible. It is clear that the information is out there, and we are just waiting on the arbitrary whim of the BEA to finally provide the rest of the story. For the most part, all the reports over the last several days make sense when put together as a whole, and for all your protestations, we are now much closer to an answer as to what happened. With the number of people involved in the investigation, it is unreasonable to think no leaks will occur. Moreover, some of the leaks are intentional, some not.

I mean, how can you even compare today's press with that of yesteryear... I receive information. It is from a credible source that I have trusted for years. It does what the news is supposed to do, provide information to the public. By releasing the information in 20 minutes, I have made a profitable decision for my company. By waiting three days, looking for independent or "official" confirmation, it is no longer news.

You are very quick to make demands that there be meaningful change to the airline industry. Well, how do you think such change will come about, without media pounding away at such issues as fatigued pilots, complacent air traffic controllers, or dangerously aging aircraft fleets? Are these generalizations? Sensationalist reporting? Dumbing down of the news? Are they insensitive to the complexity of the issues involved? Of course, all of the above. But if you don't get the public's attention, and define the message in those clear terms, the changes YOU want will never happen.

All media sources are not the same. All reporters are not the same. All issues are not even the same. The CNN coverage of the recent tornadoes has been very good in my opinion. The constant pundit-partisan campaign coverage, however, by all US media, is insufferable.

The "press" reflects modern society. People want instantaneous information. They have 5-second attention spans. They will not read a 30-page report with graphs and charts. Yet I think the information is still out there for those who do want a sophisticated account of the news, and books are out there in abundance as well. The age of information has several layers, and yet you generally go after the outer layer, the daily reports, although you seem perfectly content to comment on them, covering your ass with an all-caps IF all the while. Ever noticed how often the IF is true?

[EconomyClass]

You are very quick to make demands that there be meaningful change to the airline industry. Well, how do you think such change will come about, without media pounding away at such issues as fatigued pilots, complacent air traffic controllers, or dangerously aging aircraft fleets?

So blaming dead pilots is the shortest path to real change? Because all the leaks that the press are joyously printing and quoting all seem pointed at dead people, to the advantage of survivors who sat in their French offices. By the way, considering all the times I've read the word "stall" here, ironic it never got used in reference to either Aribus or Air France. Need some solution for stalls that happen on the ground.

[Spectator]

We agree then. The PF should stabilize the a/c before doing ANYTHING else. The PNF is there for a reason. Whatever happened to CRM?

Perhaps this is why:

Chief Pilot was not in the cockpit at the time of the accident newspapers say.

Yes, I know it is not an excuse, and an "incompetent" crew (no offence implied, there probably is a better word) should never be left alone in charge of the aircraft, but nevertheless there is rumour that the captain himself was not in the flight deck at the time. If this is true it may help explain some of the confusion?

[Evan]

Yes, I know it is not an excuse, and an "incompetent" crew (no offence implied, there probably is a better word) should never be left alone in charge of the aircraft, but nevertheless there is rumour that the captain himself was not in the flight deck at the time. If this is true it may help explain some of the confusion?

You have to distinguish between 'incompetent' and insufficiently trained. The crew on duty was very competent but possibly not properly trained for this scenario. The F/O in the left seat had far more experience on type and far more experience with the route than the captain himself. I feel certain that the captain being off-duty was not the issue here.

[Gabriel]

Yes, I know it is not an excuse, and an "incompetent" crew (no offence implied, there probably is a better word) should never be left alone in charge of the aircraft, but nevertheless there is rumour that the captain himself was not in the flight deck at the time. If this is true it may help explain some of the confusion?

The A330 has a required flight crew of two. Not only that, but any one of these two has to be able to fly alone should the other one become impaired.

In long flights, depending of the length, 3 or 4 flight crew members may be required. In the longer flight, with 4 crew members, you have two crew teams, one on duty and one resting (in turns). On long but not so long flights, you have 3 flight crewmwmbers, two of them on duty and one of them resting (again in turns). You simply can't expect one person to be on duty for more than 8 hours, unless you want to have an accident with fatigue as a contributing factor.

The captain was taking his scheduled rest when the accident sequence started. There were other two fully qualified pilots at the controls at that time. Who told you that the captain have done any better than the other two?

[TeeVee]

Yes, that is the theory.

It takes an incredible amount of discipline as PF to focus on nothing but aircraft flight path while the PNF analyses the problem. Particularly when lights are flashing, sirens are going, and systems are coming and going.
, and the aircraft appears to be performing ok.

It is, unfortunately, human nature.

Continual training and learning are they key, but how do you know how you will react when the adrenaline hits and you get the fright of your life?

We can talk CRM all we want, but please don't pretend it is an easy thing to do when the shit really hits the fan. Must it be done - yes. But don't say its easy.

I would also suspect it is far more likely when there is no turbulence at all - in turbulence the PF would already be paying close attention to the flight path. In smooth air, it would be very easy to have lowered that attention.

I will be VERY interested to see to what extent the BEA think that the "stationary thrust levers" have played in this accident.

MCM, as a real world pilot with REAL WORLD experience, you know this to be true. Evan on the other hand lives in Mr. Rogers' world of make-believe, where everyone and everything is perfect and not subject to errors in perception, judgment, response, etc. Where everyone is trained in every conceivable and inconceivable scenario and responds perfectly to it. Where every machine is 100% perfect and guaranteed to work in all circumstances and settings. of course you, as well as others here, know that arm-chairing these types of events is the easiest thing in the world to do. sitting in the cockpit of what is or what appears to be a failing aircraft with practically every alarm, chime, claxon and buzzer going off simultaneously, knowing that you too are going to die when you plunge into the middle of the atlantic ocean couldn't possibly affect your judgment or ability to fly...

If in the end the cause proves to be pilot error, well, the pilots paid the ultimate price, did they not? 200+ people also paid, unfortunate to say the very least, but part of the risk involved in flying. i suspect, the real causeS will be multiple, yet the final report will lay blame on the pilots so that the world does not lose all faith in Airbus products.** this will be so even if they find fault with the aircraft systems. pure, unadulterated CYA and protect the almighty bottom line.

Although I think he is still ignoring me, Evan, perhaps you need to take your head out of the books for a while and consider this: for the very same reason I'm pretty sure you have not scored a 100% on every test you have taken in your life and will not score 100% on the tests you take in he future, the AF447 pilots and the people that trained them didn't score a 100% this time around. it is unfortunate, but it is life, and life is not fair. it is in fact, at times, quite cruel.

** p.s. from a legal perspective, air france will not be off the hook if pilot error is determined to be the cause. the legal principle of "respondeat superior" superior will make AF as the employer liable for the pilots' actions/inactions.

[Evan]

We can talk CRM all we want, but please don't pretend it is an easy thing to do when the shit really hits the fan. Must it be done - yes. But don't say its easy.

I'm not saying it is easy. I don't believe I ever said it was easy, and for this reason being a commercial pilot isn't for everyone. It should be reserved for an elite, highly trained few who can be expected to do what it very difficult under extreme pressure. Unfortunately, this doesn't seem to be the universal standard.

Situations like this will never be easy but they can be made easier. If Air France had properly trained the crew for UAS recovery at cruise altitude, to the point where they had an instinct for it, and Airbus had provided the correct procedural guidelines to follow, managing this would have been a lot less difficult. Can we agree on that?

I know we already agree on this: Even if it isn't easy, must it be done? Yes.

[Gabriel]

I agree with MCM, you can't possibly train a pilot to react (including recognize first) to every abnormal or emergency situation as if it was second nature. That's why pilots don't need to memorize most of the abnormal and emergency procedures, and there are QRHs and checklist instead.

I agree with Evan that memory items must be learnt (not only to recite them, but also to apply the correctly) to second nature level. But as MCM said, whether even a properly trained pilot will be up to the task in the heat of the battle, it's impossible to foresee. We'll all agree that proper understanding and training will increase the chances a lot. As also will proper and clear procedures and a simpler and friendlier interface.

I can't for a second think that simple "pilot error" can be to blame. That'd be like, after a CFIT accident, blaming the mountain. Systems design, hardware failures, procedures, interfaces, safety culture, personnel selection, training and monitoring, all play a role when there is a pilot error.

And even if you make all those just perfect, you just can't get totally rid of pilot error. You'll always will have an excellent pilot, with very good training, flying a very good plane in very good condition, having his "oh, shit" moment and the worst possible time. That's one of the inherent risks of flying. I don't think for a second that this was the case here:

- Bad pitot tubes.
- Tricky thrust lock feature, that requires you to take the thrust levers out of the climb detent to set climb thrust!, and not explained in the unreliable airspeed procedure.
- Tricky abnormal alternate law that will hold the altitude despite a diminishing speed (by increasing pitch, even with the AP off and no pull), with no slow speed or stall protection.
- Confusing cascade of failures, with a lot of warning flags, lights and sounds playing simultaneously and a dozen of EICAS messages to address, without a single highly highlighted indication of what's up.
- Not well trained pilots, at least regarding this emergency, with a history industry-wide of pilots ALMOST NEVER following the prescribed unreliable speed procedure.

Room for improvement anyone?

[Evan]

I can't for a second think that simple "pilot error" can be to blame. That'd be like, after a CFIT accident, blaming the mountain. Systems design, hardware failures, procedures, interfaces, safety culture, personnel selection, training and monitoring, all play a role when there is a pilot error.

If the WSJ article is accurate, this is almost certainly what it will come down to. With proper training, I don't think the A330 systems are inherently dangerous, but I think Airbus will shoulder some blame for providing flawed procedures that fail to take them fully into account.

And even if you make all those just perfect, you just can't get totally rid of pilot error. You'll always will have an excellent pilot, with very good training, flying a very good plane in very good condition, having his "oh, shit" moment and the worst possible time.

Which is one reason there are two of them (or three in this case).

[Gabriel]

Which is one reason there are two of them (or three in this case).

And even if you make all those just perfect, you just can't get totally rid of pilot error. You'll always will have an excellent pilot, with very good training, flying a very good plane in very good condition, having his "oh, shit" moment and the worst possible time, with the second pilot either not detecting the error (maybe because he is too busy with the troubleshooting part), not daring to contradict the other (authority gradient), or failing at convincing the other that he is in error.

Fixed.

[Evan]

And even if you make all those just perfect, you just can't get totally rid of pilot error. You'll always will have an excellent pilot, with very good training, flying a very good plane in very good condition, having his "oh, shit" moment and the worst possible time, with the second pilot either not detecting the error (maybe because he is too busy with the troubleshooting part), not daring to contradict the other (authority gradient), or failing at convincing the other that he is in error.

Which gets me back to:

The PNF is there for a reason. Whatever happened to CRM?

You will never completely eradicate pilot error, or human nature, but you can do A WHOLE LOT MORE than what is being done now to minimize it.

Despite repeated incidents of UAS previous to this crash, in which the danger became clearly apparent, apparent enough that it begat a conversation between AIrbus and certain operators and resulted in a revised Thales BA pitot design and an AIrbus recommendation to implement them... despite all that awareness of the danger... apparently (according to the WSJ article) UAS procedures and CRM were not covered in crew training. I find that to be incredible.

[Gabriel]

Despite repeated incidents of UAS previous to this crash, in which the danger became clearly apparent, apparent enough that it begat a conversation between AIrbus and certain operators and resulted in a revised Thales BA pitot design and an AIrbus recommendation to implement them... despite all that awareness of the danger... apparently (according to the WSJ article) UAS procedures and CRM were not covered in crew training. I find that to be incredible.

Incredible, yes.
Surprising, not, given what the BEA found that in 12 (I think) other UAS events, and in none (I think) of them the UAS procedure was followed, INCLUDING one AFTER the AF crash, where the UAS was already a strong suspect of having been a necessary link in the chain of events.

[3WE]

Room for improvement anyone?

Flight control computers that pretty much always generate traditional responses to changes in airspeed and attitude and even perhaps phugoid behavior.

A spring on the control stick to make it want to stay neutral and let the pilots know when they are theoretically forcing a control input away from a fat, dumb and happy trim setting.

[EconomyClass]

Air France underwent an extensive review of its safety operations following the 2005 crash landing of an A340 in Toronto. The report was highly critical of its safety and pilot training standards. Executive Vice President of Operations Gilbert Robetto says 90% of the recommendations have been implemented.

http://www.aviationweek.com/aw/gener...aw081009p1.xml

Anybody ever specify what the 10% was that hadn't been implemented?