Very good AF 447 article

_dof_

Spotted a link to this article on pprune, from Vanity Fair of all places..

vanityfair.com/business/2014/10/air-france-flight-447-crash

Excellent article, probably the best aviation article I've seen in mainstream press.
Irish pilots even get a mention towards the end.

fr336

Well that was a very interesting read for me, though nothing will change as the moneymen hold the pursestrings and the world not just aviation gets more and more automated. What's a crash here and there to them when they are saving millions. Technology can be good 99% of the time imo, often better than humans for obvious reasons, but the 1% is the difference between life and death.

Irish Steve

The bean counters have a lot to answer for, the 447 accident could have happened at any time, the fact that it happened to the flag carrier of the country that builds Airbus has come as a big embarrassment to all concerned, and it has exposed the weaknesses of the training regime in some of the carriers. There have been other incidents of a similar nature, but the system has been able to find ways to make them less embarrassing.

The manner in which pilots are trained has changed significantly over the last 20 years, and one of the downsides to that is that many people at the sharp end of the modern EFIS jet have never spent any time flying using "raw data", with next to no electronics massaging the information for them, and that absence means that their skills when things go pear shaped are less well developed.

The other issue is that "self improvers" are frowned upon by the modern system, it used to be the case that if someone wanted to learn more about the aircraft they were flying, if a simulator slot was "spare", it was not a problem to let them use that slot for learning more about the extremes of how the aircraft handles when certain faults occur, but that's all changed, spare simulator slots are less frequent than they used to be, and a pilot who want to learn more is likely to be regarded as having a problem that he (or she) is not prepared to discuss rather than wanting to improve the skill level, and everything now has to be done using "standard operating procedures", which is fine in theory, and next to useless when the reality is that the operation has suddenly and perhaps catastrophically become very much NON standard. At that point, you are down to how well the crew understands the (complex) systems and interactions of the aircraft they are now left to fly with very little automation to help them.

The justification for some of these changes are that they happen so rarely, it's uneconomic to train pilots to be able to deal with them. Probably has something to do with the bean counters never being within a million miles of having to comfort grieving relatives after a fatal accident, and anyway, the investigation will probably find that it was "pilot error". Yeah, right, we're really convinced by that excuse.

As an example, after a massive uncontained engine failure on an A380, which caused widespread and significant damage to the aircraft, a very experienced "heavy" crew (4 on board rather than the standard 2) needed nearly 5 HOURS to resolve all the check list issues and problems caused by that failure before they were able to determine that it was going to be safe to land the thing. If the weather in the area had been more hostile, that's a very long time to be flying a damaged and possibly crippled aircraft while trying to sort out the problems.

Bean counters also don't like people flying the aircraft by hand, it might be fractionally more expensive, and that only serves to make the problem worse, there is no alternative to keeping skills sharp other than actually flying the aircraft.

Welcome to the modern world of aviation.

cml387

Excellent article.

For the other side of the story I'd recommend Fly By Wire , a book written about how the Airbus flight software (and some good luck) ended up with the Miracle On The Hudson

keith16

Irish Steve wrote: »

many people at the sharp end of the modern EFIS jet have never spent any time flying using "raw data", with next to no electronics massaging the information for them, and that absence means that their skills when things go pear shaped are less well developed.

Article was a good read and a good response Steve. I'm particularly interested in what you say, quoted above.

Raw data is one thing, but I guess the AF 447 pilots did have information massaged for them.

In my mind, it wasn't "massaged" enough. They lost airspeed indication only, correct? Yet this was interpreted as a loss of airspeed.

So despite the data being massaged for them, my sense is that they only had information, not necessarily insight.

billie1b

cml387 wrote: »

Excellent article.

For the other side of the story I'd recommend Fly By Wire , a book written about how the Airbus flight software (and some good luck) ended up with the Miracle On The Hudson

In my opinion the Hudson miracle was nothing but pot luck and a great pilot, to me it was more down to wind/weather conditions and water current/tides being perfect, I reckon it could have been any type of aircraft and they would have gotten the same results

cml387

billie1b wrote: »

In my opinion the Hudson miracle was nothing but pot luck and a great pilot, to me it was more down to wind/weather conditions and water current/tides being perfect, I reckon it could have been any type of aircraft and they would have gotten the same results

Did you read the book?

billie1b

cml387 wrote: »

Did you read the book?

Honestly no, I just downloaded it on iBooks, i'll start it later

donvito99

cml387 wrote: »

Excellent article.

For the other side of the story I'd recommend Fly By Wire , a book written about how the Airbus flight software (and some good luck) ended up with the Miracle On The Hudson

An excellent read, I'd also recommend Aloft by the same author, William Langewiesche.

pclancy

Excellent article.

Now, I'm not thinking this is related to the crash but...

"at that moment the cockpit door opened and a flight attendant walked in, asking that the temperature in the baggage hold be lowered because she was carrying some meat in her suitcase"

Can any commercial pilots comment on that kind of request happening often? Weird.

adamski8

Good read. Communication was terrible alright. Didn't seem like anyone knew what each other were talking about or worried about. 1st person didnt fully explain and 2nd didnt asl what they meant and that was even before the incident started.

Bonin really seemed to have a urge to go to 36000 which seemed really odd.

Really weird they thought they haf lost airspeed. Surely they knew the engines were working and they were at the right pitch etc. It sounds to me like one of the passengers trying to fly it. And shouldn't they be pulling out a checklist first not just randomly moving controls around and seeing what happens?

PukkaStukka

Irish Steve wrote: »

....
As an example, after a massive uncontained engine failure on an A380, which caused widespread and significant damage to the aircraft, a very experienced "heavy" crew (4 on board rather than the standard 2) needed nearly 5 HOURS to resolve all the check list issues and problems caused by that failure before they were able to determine that it was going to be safe to land the thing. If the weather in the area had been more hostile, that's a very long time to be flying a damaged and possibly crippled aircraft while trying to sort out the problems.

This is a very valid point. I presume you are referring to the Quantas. In the reconstruction of that incident, the amount of errors being generated after the failure was unbelieveable to the point where they were being created faster than the crew member could read them. Was the pilot fighting an issue or debugging a whole IT system? They were so lucky that they had the check captain and engineer with them that day to share an insane workload being created by a system whose purpose is, among other things, to lighten it!

ectoraige

Excellent articles, raises since good points about the challenges automation brings when it relies on a manual fallback. It's tragic that it could have been averted so easily. I can't help think of Bonin's children growing up knowing how easily their father could have stopped the problem. I must confess I was surprised by the apparent lack of pitch data available to the crew, surely when it's smart enough to sound a stall warning, it might consider showing this over other data.

Nonoperational

Great read. Thanks.

adamski8

Did anyone in the industry think the conversations were strange throughout or is this how some pilots would talk to each other? I mean there wasn't much jargon banded about. Much more like half sentences and ambitious remarks!

Nwm2

_dof_ wrote: »

Spotted a link to this article on pprune, from Vanity Fair of all places..

vanityfair.com/business/2014/10/air-france-flight-447-crash

Excellent article, probably the best aviation article I've seen in mainstream press.
Irish pilots even get a mention towards the end.

"Simulator studies have shown that Irish pilots, for instance, will gleefully throw away their crutches, while Asian pilots will hang on tightly. It’s obvious that the Irish are right, but in the real world Sarter’s advice is hard to sell. "

_dof_

donvito99 wrote: »

An excellent read, I'd also recommend Aloft by the same author, William Langewiesche.

Thanks, will buy that, interestingly from Amazon:

About the Author
William Langewiesche is the author of seven previous books: Cutting for Sign, Sahara Unveiled, Inside the Sky, American Ground, The Outlaw Sea, The Atomic Bazaar, and, most recently, Fly By Wire. He is the international editor for Vanity Fair.

So that probably explains the quality of an aviation article in Vanity Fair.

Turbulent Bill

PukkaStukka wrote: »

This is a very valid point. I presume you are referring to the Quantas. In the reconstruction of that incident, the amount of errors being generated after the failure was unbelieveable to the point where they were being created faster than the crew member could read them. Was the pilot fighting an issue or debugging a whole IT system? They were so lucky that they had the check captain and engineer with them that day to share an insane workload being created by a system whose purpose is, among other things, to lighten it!

The flight computers should really parse the errors just to show the first few generated, with an indication that more were subsequently found. At least that way the pilots have a better chance of understanding the root cause, which would put the rest of the warnings in context.

In fairness to Airbus, the excellent Vanity Fair article shows how the automated systems did degrade gracefully - you couldn't ask for much more in the circumstances.

ectoraige

Turbulent Bill wrote: »

The flight computers should really parse the errors just to show the first few generated, with an indication that more were subsequently found. At least that way the pilots have a better chance of understanding the root cause, which would put the rest of the warnings in context.

In fairness to Airbus, the excellent Vanity Fair article shows how the automated systems did degrade gracefully - you couldn't ask for much more in the circumstances.

The problem though is that on many occasions it's not immediately clear how the root cause could lead to a catastrophic loss. There's been plenty of "swiss-cheese" incidents where it took months of stimulation and testing before investigators understood what was happening.

Dropping from autopilot means the avionics designer has reached a point where they fell a human is better able to determine the best course of action. The priority for the PF is to know what is and isn't working at that point in time. One they are flying and can communicate, then they can start investigating causes. Clearly on this flight they weren't seeing the effects of their inputs - they weren't even wholly aware of what their inputs were.

_dof_

ectoraige wrote: »

Clearly on this flight they weren't seeing the effects of their inputs - they weren't even wholly aware of what their inputs were.

One of the most bizarre things is the lack of airmanship when the aircraft was stalling.
Even a PPL student would be expected to have developed the instinct that to recover from
a stall you must lower the nose first, so even if they are bombarded by warnings and confusing information, that stall warning should really have cut through all other warnings and in their brains and triggered an instinctual behaviour.

Maybe it's the over reliance on the automation and that in Normal Law the systems won't allow a stall to happen so they lost that instinct? Are stalls and flying in Alternate Law even covered in recurring training/checks for the Airbus family? It's probably considered too remote a possibility to cover in training?

Then the collapse of CRM compounded that initial mistake. so that the most basic of challenge and responses "I have control", "You have control" never happened. So you have two pilots both trying to control the aircraft with different inputs, with the systems trying to merge the conflicting control inputs into something that neither pilot would have recognised as their input. So I think you're right, the weren't aware of what their inputs were.

The Boeing advocates will argue that the sidesticks with no direct physical connection or force feedback allowed this to happen, and in a Boeing, it couldn't happen, because each pilot would have been able to feel what control inputs the other pilot was making. I suppose that could be argued, but even in the Airbus, it should have been second nature for one pilot to relinquish control to the other one, if CRM was working.

Maybe there's a psychological aspect in that the relief pilot wasn't seen by the copilot or himself as the commander (in the absence of the captain), so he was reluctant to step in and firmly take control. Or does the relief pilot assume final authority in the absence of the captain?

donkey balls

Have to say I never knew that the side sticks on the buses operated independently:o,For some strange reason i thought that all aircraft control yokes are the same regarding one moves the other also moves.

Irish Steve

The first thing that came to my mind was why they never went near any of their checklist items, one of which should have been unreliable airspeed, as it's a known issue and one that is covered in check lists. For whatever reason, they seemed to be unable to function as a crew from the outset of the problem, and there appeared to be no strategy to analyse and resolve the problems they were facing.

The most sickening aspect of it all is that by the time they hit the ocean, the aircraft was again fully functional, but they had made such a mess of the handling at the beginning of the incident, the automation was no longer able to give them any meaningful information, as it did not believe the parameters it was getting, due to the extremes that were being reported.