Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The crash of Air France flight 447 (2021)

admiralcloudberg.medium.com

Post reply on HN
Print view View original

201–210 of 231 posts

Re: The crash of Air France flight 447 (2021)

#201

Post by afterburner »

mrguyorama wrote: post #185
afterburner wrote: post #159

Earlier quoted context omitted.

Perhaps the physical effort of struggling against his co-pilot would have made him realize he was doing something wrong. It's too easy to pull back full on a fully non-mechanical control, and just lock up there.

No, this situation is the same as when lifeguards are warned about saving panicking and drowning people, in that if you don't handle them in specific ways they will happily drown you in their panic. You cannot rely on "shaking someone out of it" in these circumstances because there is no reliable way to force the brain to reject it's panic instincts and return to rational thought. The only solution would have been ph…

Locking him out of the controls would have been enough (if the technology allowed it). Even just relieving him of his duty would have been enough. If Captain Dubois had been there early enough and decided to relieve him, Bonin would have almost certainly obeyed.

If he doesn't, then I guess then you can resort to force.

Re: The crash of Air France flight 447 (2021)

#202

Post by cyberax »

local_issues wrote: post #80
turnsout wrote: post #56

Earlier quoted context omitted.

To be fair, it sounds like this crash caused airlines and regulators to demand much more adverse event training, including high-altitude stall recovery. But according to the article, Airbus hasn't addressed the split control issue…

There isn't a "split control issue," the split control setup is fundamental to the way the Airbus control schema works. Changing it would be a non-starter.

Why? The split control never made any fucking sense. It resulted in several crashes where inputs were averaged.

Re: The crash of Air France flight 447 (2021)

#203

Post by nwallin »

local_issues wrote: post #80
turnsout wrote: post #56

Earlier quoted context omitted.

To be fair, it sounds like this crash caused airlines and regulators to demand much more adverse event training, including high-altitude stall recovery. But according to the article, Airbus hasn't addressed the split control issue…

There isn't a "split control issue," the split control setup is fundamental to the way the Airbus control schema works. Changing it would be a non-starter.

You say this as if the split control setup is a feature and not a bug. Can you go into why this is the case?

For me at least, it seems like the condition where both pilots are inputting control on the stick or rudder is always an error condition. Even if they're both inputting the same inputs.

The scheme directly contributed to the crash for two reasons; obviously if the co-pilot wasn't pulling back on the stick, they would have been able to get out of the stall. But additionally, the two crew members were spending crew resources on redundant tasks: both of them thought that they were the ones aviating and the other was in charge of figuring out why the flight computer seemed to be behaving badly. Possibly if one of them did dedicate their attention to the flight computer, they'd have figured out what the actual flight parameters were.

Re: The crash of Air France flight 447 (2021)

#204

Post by cyberax »

aaronmdjones wrote: post #117
throwaway110823 wrote: post #63

Earlier quoted context omitted.

Someone in my extended family works at Airbus as an engineer (hence the throwaway account) When I asked him about this accident, I brought up this specific point - the mechanical non-linkage. Sure, the Airbus is fly by wire (there are no "mechanics"), but you can still program one joystick to mimic the other joystick, right? As far as I know, the Airbus plane actually averages the inputs..?? [0] Anyway, he sorta-angr…

The 777 and 787 are fly-by-wire aircraft with mechanically-linked yokes. In extreme situations (one yoke is actually stuck and won't move at all), there is a breakaway on that link, and if you exert enough force upon it (like, really hulk at the thing), you can snap it. I don't know whether this has been necessary in flight, but the facility is there. The control surfaces are split between the yokes, so for example t…

> I don't know whether this has been necessary in flight, but the facility is there.

It happened in one accident, where pilots were literally tugging the yokes in opposite directions. They were literally struggling against each other to fix the attitude.

Re: The crash of Air France flight 447 (2021)

#205

Post by cyberax »

_cenw wrote: post #57
kloch wrote: post #52

It's tempting to blame the inexperienced first officers, their training, or the CRM failures on this flight, but the blame for this crash has to land squarely on the atrociously bad UX of the Airbus cockpit. Sensory overload and no clear readout of what is actually broken (pitot tube icing in this case) is bad but the fly-by-wire joystick configuration is what really doomed this flight. In Boeing airliners and many o…

* The overhead panel would've shown FAULT on the ADIRs. * The ECAM did show IF SPD DISAGREE: ADR CHECK PROC. * The plane did shout DUAL INPUT several times, and a button to lock out the other pilot is right on the stick. You can hear it being used by Bonin. I'd say this is a failure of lack of crew resource management (not even technical ability) more than anything else.

The thing is, people in high-stress situations tend to miss audio cues. That's why Boeing uses stick shakers for the stall warnings.

And there were _several_ incidents with Airbuses that were caused by dual inputs.

Re: The crash of Air France flight 447 (2021)

#206

Post by rnoises »

On a slightly different note, this sort of thing is what makes me question whether we can have an orderly progression from the current Level 2 self-driving in cars to Level 5. When a plane does most of its flying by itself, if it can cause this much chaos by handing control back to a trained pilot without much warning, how much worse would it be when a car hands control back to a inexperienced, distracted driver in tricky driving conditions... especially considering that reaction times need to be much faster on the road?

Re: The crash of Air France flight 447 (2021)

#207

Post by _cenw »

mardifoufs wrote: post #195
_cenw wrote: post #75

Earlier quoted context omitted.

I'd be hesitant to compare these. The aircraft still operated as designed. It detected that the input data was bad and turned the automation off. The pilot flying just didn't seem to grasp what that means. Most Tesla Autopilot failures aren't drivers that don't know how to drive, it's the automation making bad inputs.

No, some warnings were also thrown away and never shown to the pilots because the automated systems thought they were erroneous (since they were so outside the flight envelope IIRC).

If by warning you mean the audible stall warning - that's because below 50 knots, the airspeed is considered unreliable / erroneous, because to get to that speed without flaps you'd have to be stalling the aircraft for over a minute (getting that exact warning) and surely nobody would manage to do that...

As I said elsewhere, the airplane was quite literally falling as fast as it was moving forward. I don't really fault the plane for considering that condition erroneous. And it didn't throw away a warning, it instead marks the speed tape with a bright red "SPD" instead of showing any airspeed. This is likely the better alternative over getting persistent stall warnings if one of your airspeed indicators fails.

Here's how that condition would've looked: https://docs.flybywiresim.com/pilots-corner/a32nx-briefing/p...

Re: The crash of Air France flight 447 (2021)

#208

Post by cyberax »

ceejayoz wrote: post #33
netsharc wrote: post #14

Earlier quoted context omitted.

I suppose you've heard it all, like the fact that there are more than 10000 planes currently in the air and we haven't heard about a plane crash for weeks now. If you go to https://globe.adsbexchange.com and zoom out you can see all the planes around the world, flying safely. I suppose the only way through the fear is to fly and realize at the end that you're fine (although obviously IANAP)

Also note that the last deadly airline crash in the US was all the way back in 2009. We've gone 14 years without a single one, which is a pretty remarkable safety record for the FAA and NTSB.

> Also note that the last deadly airline crash

Do you mean a total loss of a passenger airline?

Because there was:

- "Prime Air" crash in 2019 ( https://en.wikipedia.org/wiki/Atlas_Air_Flight_3591 ), it was a total loss of a cargo airplane.

- A deadly crash in SF in 2013: https://en.wikipedia.org/wiki/Asiana_Airlines_Flight_214

And a couple more deadly incidents.

Re: The crash of Air France flight 447 (2021)

#209

Post by ceejayoz »

cyberax wrote: post #208
ceejayoz wrote: post #33

Earlier quoted context omitted.

Also note that the last deadly airline crash in the US was all the way back in 2009. We've gone 14 years without a single one, which is a pretty remarkable safety record for the FAA and NTSB.

> Also note that the last deadly airline crash Do you mean a total loss of a passenger airline? Because there was: - "Prime Air" crash in 2019 ( https://en.wikipedia.org/wiki/Atlas_Air_Flight_3591 ), it was a total loss of a cargo airplane. - A deadly crash in SF in 2013: https://en.wikipedia.org/wiki/Asiana_Airlines_Flight_214 And a couple more deadly incidents.

Sorry, should've said "of a US airline".

I wouldn't count cargo planes if I were evaluating risk as a passenger.

Re: The crash of Air France flight 447 (2021)

#210

Post by deepspace »

Tronno wrote: post #98

In post-mortems involving pilot error, what strikes me is that airliner cockpit design sounds so convoluted, unintuitive, and just plain bad . The burden of disentangling this bad design is always placed on the pilots - through extensive training and rote memorization - which inevitably fails under stress. In this particular example, consider: - the opposing pilot inputs being signaled only by a pair of little green…

Not relevant in this example, but another terrible user interface in the cockpit is the FMC. That thing with the '70s era green screen, where pilots program routes with cryptic codes, seemingly inherited from the Apollo guidance computer.

So many accidents seem to happen when pilots receive a route change and have to hastily re-program the FMC.

Post reply on HN