Air France 447 - On topic only!

[Fear_of_Flying]

Just to clarify my point. I think the conversation went something like:

Evan expressed concern over the lack of CRM training for two relief copilots.

3WE responded by mentioning that even in the absence of formal CRM training, good airmanship, communication and common sense should have allowed the pilots to work something out.

My main point was simply that a crisis of that nature is not the ideal time to hope that level heads will always prevail. Added to that is the fact that the sidesticks don't, as I understand it, give effective visual or tactile feedback in the way that the Boeing control column might, so it wasn't immediately obvious to the other pilot what the PF was doing.

I would also add that while verbal communication is important, it is once again relying on humans, and if they are the same people who are making deer-in-the-headlight responses in their inputs, why would I hope they're going to pull it altogether in their verbal communications?

Did the pilots screw up? Of course, but as Evan says, understanding that doesn't get one very far. Was Airbus to blame? No, but that isn't to say there aren't aspects of the Airbus systems that proved less than advantageous in this situation.

I apologize I don't have the perceptive powers needed to make brilliant black-and-white proclamations like "The pilots f***ed up, what more do you need to know?" I'll leave that to the inscrutable judgment of our esteemed commercial jetliner pilots on this board.

[Evan]

Evan expressed concern over the lack of CRM training for two relief copilots.

Just to clarify my point, I'm not expressing concern about that. I'm expressing shock and outrage.

It is so far beyond my comprehension that Air France or the BEA would have such inadequate training standards in place at this point in history. There are so many lessons in that history. Airbus went to enormous lengths to create a safer aircraft, but what good is it if the people operating it sit on the hands and whistle?

To your point about the visual reference of the control column, I don't think the other pilots would use this even in a Boeing scenario. They would be looking at the attitude indications on the instruments, as they should be doing. AF447 was responding to pitch-up commands. The instruments would have shown a pitch-up attitude. That alone would have told them what the PF was doing.

[Fear_of_Flying]

To your point about the visual reference of the control column, I don't think the other pilots would use this even in a Boeing scenario. They would be looking at the attitude indications on the instruments, as they should be doing. AF447 was responding to pitch-up commands. The instruments would have shown a pitch-up attitude. That alone would have told them what the PF was doing.

Boeing pilots would be looking at the attitude indications on the instruments? Not if they were these pilots. Don't get me wrong. I'm sure these guys would have crashed in a Boeing, too. However, there were a couple of points in the sequence when they were both trying to control the aircraft, or when one pilot took control back from the other, that it might have been more apparent what was going on in a Boeing.

It does seem for the most part that the two pilots at least understood who was supposed to be flying the plane (he was the one making the nose up inputs the whole time), and who was supposed to be monitoring (he was the one oblivious to what was happening). It's just that neither one was competent in his role.

Don't worry, though, once the stall recognition/avoidance training is firmly in place, I think we can safely predict how the next disaster will occur. There will once again be a stall warning on an aircraft somewhere, only this time the pilot will apply full thrust and pitch the plane into a steep dive, probably hitting a mountain or a skyscraper. That would seem to be the natural outcome of trying to apply procedures under extreme duress rather than learning to think about your situation and fly the plane accordingly.

[Evan]

However, there were a couple of points in the sequence when they were both trying to control the aircraft, or when one pilot took control back from the other, that it might have been more apparent what was going on in a Boeing.

Absolutely. And proper CRM explicitly requires a pilot to verbally request and hand-over sidestick control ("My airplane..." "Your airplane"...). There is also a sidestick priority button to assure that this has been communicated. There is a glareshield warning if the other sidestick is not centered. None of this prevents either pilot from making control inputs against the designated PF (the last pilot to press the priority button has priority). The aircraft was not designed to be flown by combative pilots, and definitely not by idiots. That is the part I think some people fail to understand.

After all the criticism about Airbus planes being designed to be idiot-proof, you are making the case that Boeing planes are more idiot proof.

[3WE]

...I don't think the other pilots would use this even in a Boeing scenario. They would be looking at the attitude indications on the instruments, as they should be doing. AF447 was responding to pitch-up commands. The instruments would have shown a pitch-up attitude. That alone would have told them what the PF was doing.

...After all the criticism about Airbus planes being designed to be idiot-proof, you are making the case that Boeing planes are more idiot proof. ...

Let's can the over simplification of "idiot proof" when comparing Boing and Air Bus. The bigger question is AirBus deviating from traditional control behavior and feed back.

-Do you think that both pilots having a big old yoke shoved in their stomach would have had any value to their situational awareness?

-Does the Air Bus not have a big TV screeen that is 1/2 brown and 1/2 blue? (Sorta like your typical modern boeing and all other glass cockpit planes*)?

*And sort of like this thing called an attitude indicator that is located top center of a traditional, "steam gauge" airplane?

[Graham2001]

This anecdote is older than the incident in question, but the psychology is relevant.

Supposedly during the late 60s/early 70s a Trident was flying from Frankfurt (?). While flying under the control of the autopilot the plane executed a 180 degree turn.

The captain who was in the cabin ran back to the cockpit and found the co-pilot busy punching buttons on the autopilot.

The captain took control of the plane, returned it to it's proper heading and then turned to the co-pilot and advised that they could now find out where the problem was...

The above is related from memory, as soon as I have the source I'll post the details, but the point is clear, in an abnormal situation, the first priority is to get it under as much control as you can, not play with the computer.

It's taken me a while, but I've finally managed to run down the story. It appears in the form of an anecdote attributed to a retired air accident investigator called Jim Owen on page 403 of Stephen Barlays book "The Final Call: Air Disasters...When Will They Ever Learn."

My memory had failed me partially, supposedly the 180 turn happened to a Tristar belonging to TWA and it was flying to Frankfurt, but the point remains the same.

[Myndee]

Just wanted to stop by and say hello to my favorite thread. Can't believe it is still going after all this time. I've been busy but maybe someday I can come back and hang out...miss this place, actually.

[EconomyClass]

Air France is amending its flight-safety analysis procedures, introducing a new position to quicken its response and assessment of incidents.
The carrier undertook an in-depth examination of its operations and decision-making processes in the wake of the mid-Atlantic crash of flight AF447 in June 2009.
Investigators have yet to release a final report into the accident but have attributed the loss of the Airbus A330 to a failed stall recovery after an encounter with in-flight icing.
Air France says the creation of a new "gatekeeper" role follows recommendations from external specialists who reviewed the carrier's safety procedures.
In the event of an incident or other flight occurrence, the gatekeeper would contact the captain, first officers and other crew to "rapidly obtain additional information" which the carrier believes would be "essential" to understand any anomalous situation.

http://www.flightglobal.com/news/articles/air-france-revamps-flight-safety-analysis-procedures-369608/

[Evan]

Air France says the creation of a new "gatekeeper" role follows recommendations from external specialists who reviewed the carrier's safety procedures.
In the event of an incident or other flight occurrence, the gatekeeper would contact the captain, first officers and other crew to "rapidly obtain additional information" which the carrier believes would be "essential" to understand any anomalous situation.

So... up until now nobody was doing this?

[Spectator]

So... up until now nobody was doing this?

No, up to now the ECAM messages were used for advance notification to maintenance, and the pilots flew the plane.

Saying that contacting the crew is "essential" to understanding a situation implies that they never really understood the prior problems (read pitot icing incidents) properly. The change is from reactive to proactive safety.

[Evan]

No, up to now the ECAM messages were used for advance notification to maintenance, and the pilots flew the plane.

Saying that contacting the crew is "essential" to understanding a situation implies that they never really understood the prior problems (read pitot icing incidents) properly. The change is from reactive to proactive safety.

The pitot issue (transient UAS at cruise altitude) was a recognized issue, hence the recommendation to refit with the newer pitots prior to the AF447 crash. So the mechanical fix was proactive. What wasn't recognized was the lack of training on UAS procedure, despite a dialogue between Airbus and their customers describing the experiences. Hopefully, this "gatekeeper" is there to recognize procedural failures, although theoretically they shouldn't need a special "gatekeeper" to recognize this. And hopefully they are putting more emphasis on upset training, CRM and adherence to procedure.

[Spectator]

The pitot issue (transient UAS at cruise altitude) was a recognized issue, hence the recommendation to refit with the newer pitots prior to the AF447 crash. So the mechanical fix was proactive. What wasn't recognized was the lack of training on UAS procedure, despite a dialogue between Airbus and their customers describing the experiences.

Agreed 100%. When I referred to "pitot icing incidents" I was thinking more about the handling of them than the mechanical issues. No one seemed to identify until after AF447 that pilots were not adhering to the prescribed procedures.

[MCM]

I wonder if the "Gatekeeper" is actually the same "gatekeeper" that is the key person in the privacy system relating to quick access recorders and accessing flight data. That is a system used by a number of airlines around the world.

[Evan]

I wonder if the "Gatekeeper" is actually the same "gatekeeper" that is the key person in the privacy system relating to quick access recorders and accessing flight data. That is a system used by a number of airlines around the world.

How does that system work?

It sort of sounds to me like this is more about protecting AirFrance than protecting their passengers. They must be reluctant to repeat the AF447 ACARS leak that started all the discussion.

[MCM]

Nope its not like that at all.

FOQA (Quality Assurance data such as Quick Access Recorder information etc) data is identity protected, and is used for statistical monitoring only.

The concept is that that FOQA is to monitor operational standards and to monitor trends, and is not to "spy" on individual crew. So, for example, an airline will use this sort of data to see if there is an unusually high number of unstable approaches on a particular runway, and ideally to then investigate why that is occurring.

However, due to the deidentified nature, it would not be possible to actually contact the crew - which may slow down a proper understanding of the incident.

The gatekeeper is a person - usually a member of the pilot association (be it union or otherwise), who is the only person who can link specific data to specific crew. They can then contact the crew to find out more information, without the "who" ever being known.

Naturally, the gatekeeper has to be someone trusted by both the company and pilot group.

Its a system that, certainly in my very limited experience, seems to work quite well. Pilots are happy they are not being spied on, and companies get the safety data they need.

Note: this is purely for FOQA monitored data. Normally any noteworthy events would be reported by the crew.