4 ms·

Good thinking. Airbus planes are impervious to software glitches[1] and sensor failures[2] that result in catastrophic loss of control of the aircraft. Better

Good thinking. Airbus planes are impervious to software glitches[1] and sensor failures[2] that result in catastrophic loss of control of the aircraft. Better bring a gas mask[3] though. And bring one for the flight crew too in case they become incapacitated. [1] https://en.wikipedia.org/wiki/Qantas_Flight_72 https://en.wikipedia.org/wiki/Qantas_Flight_72 [2] https://en.wikipedia.org/wiki/Air_France_Flight_447 https://en.wikipedia.org/wiki/Air_France_Flight_447 [3] https://viewfromthewing.com/is-airbus-doing-enough-to-stop-the-fumes-that-are-making-people-sick/ https://viewfromthewing.com/is-airbus-doing-enough-to-stop-t...

inferiorhuman 4y ago
AF 447 was first and foremost a problem between the controls and the seat. The sensor problem you're referring to did not render the plane incapable of flying, but it did remove one level of idiot proofing. As a result one of the pilots pulled the stick back and didn't relent even as the plane stalled. If you did that in a Boeing plane (or any plane) with functioning sensors you'd have the same end result.
- mlyle 4y ago
  Neither did the MAX debacle render the plane incapable of flight. Both flights gave flight crews unexpected tests on unexpected situations-- something that seemed like runaway trim for the MAX, and simultaneous overspeed (better nose up) and stall (better nose down) indications for AF447.
  - rocqua 4y ago
    Test were done with MCAS regarding how soon you have to react before it becomes unrecoverable. This was 10 seconds and it was hard for boeing-trained pilots instructed in MCAS to recover from this in tests where they knew it would happen. The problem was that at some point the trim is so much that it becomes nigh impossible to hand crank the trim down, the trim flap applies so much pressure it cannot be overcome by hand. You had to disable mechanized trim because otherwise MCAS would kick in. So you would need to recognize the problem soon enough to prevent MCAS from moving the trim so far that it became unrecoverable. (Source: Netflix documentary)
    - mlyle 4y ago
      > This was 10 seconds 10 seconds is a long time of the trim wheel being cranked nose up. The same situation would apply to a stuck trim switch on the yoke. (Indeed, a stuck trim switch is worse in this respect because it cranks constantly and doesn't "yield" to the pilot pushing the switch the other way... of course, the checklist didn't presume a pilot should be diagnosing stuck AOA vs. stuck trim switch and thus recommended immediate use of the trim cutout switches instead of trying to electrically trim the other way). There's a reason that trim runaway is a "recall" procedure in the QRH that pilots are expected to know-- it gets scary. http://www.b737.org.uk/images/runawaystab2020.jpg http://www.b737.org.uk/images/runawaystab2020.jpg But it was the assumption before the MCAS disaster that pilots could deal with trim runaway, given that there was a history of flight crews dealing with trim runaway incidents in flight successfully. But MCAS was worse because: A) It did it more often, so flight crews were exposed to this dangerous situation more than ever before B) It did it in a sneakier, strange way (no constant movement of the trim wheel). Still, some real-world flight crews did overcome MCAS doing the wrong thing without crashing, even if others failed the test.
      - rocqua 4y ago
        I imagine point B is quite relevant here. MCAS going wrong is meaningfully different (and less immediately scary) from the recall procedure that the training doesn't kick in.
        mlyle 4y ago
        Sure. The main points I have, I guess, are: - It's perhaps understandable why engineering teams-- pushed by management to deliver a 737 that doesn't require retraining-- decided that if flight crews could handle runaway trim, they could handle any errant MCAS activations and that the level of redundancy and fault tolerance necessary in the system would be low. - On the other hand, real flight crews obviously do not do great when exposed to this test. - If you give a flight crew a wacky situation with conflicting information (like AF447 or the 737 MAX)-- expect failures, even if the situation is technically 100% recoverable and flyable and some crews exposed to the situation succeed. It's great that we have humans capable of reason up in the front of the plane, but reason is often faulty and doesn't get you the solution you need on the first try.
- nopzor 4y ago
  you couldn’t even physically even do that in a boeing plane, because of sensible hmi choices that boeing made early in their history. so i think your point is completely invalid.
  - inferiorhuman 4y ago
    Couldn't do what? Pull the stick back into a stall? Sure you can. Under normal law you can't do that with an Airbus, that's one of many things that caught the AF crew off guard. There's no such protection on the 737. Nobody's done that because unlike on the Airbus, it's never safe to do so. There are, however, plenty of ways to get caught off guard in a 737 and stall it. Take a look at the UTAir and Thomsonair go arounds that got their planes to about 45° nose up. Or the 737 Air Florida planted right into the Potomac because they didn't have the elevator authority to avoid the stall. Or Fly Dubai… As for human interactions, the 737 is a series of bad choices (especially the MAX). Take a look at the Helios Air crash, the pilots were confused by an alarm that could mean two different things. Unfortunately hypoxic pilots aren't known for their reasoning ability.
    - cjrp 4y ago
      Couldn't have both pilots fighting each other on the controls (hard nose-up vs. hard nose-down) without knowing about it, since the yokes are physically linked (unlike sidesticks).
      - nopzor 4y ago
        yes, this is what i meant. i should have been clearer in my original statement. here is what happened in the airbus air france incident: - one pilot had the sidestick full hard nose up, the other full hard nose down. - neither pilot realized what the other pilot was doing. on a boeing airliner: - the controls for each pilot are physically linked. - if pilots try to put “their” yolk into an opposing position, they will start having to physically wrestle for control. - this type of failure mode of the air france crash wouldn’t have happened in a boeing airplane; the pilots would have realized what was going on within seconds. general note: - when flying, it is highly unusual for both pilots to be giving input into the controls at the same time. one pilot “has the airplane”, the other is there to assist, help with the radios, planning, etc. the handoff between who “has the airplane” is very deliberate. this was another failure mode of the air france flight. - but my point still stands, the fact that the controls are linked makes the gp point a bit moot.

inferiorhuman 4y ago

AF 447 was first and foremost a problem between the controls and the seat. The sensor problem you're referring to did not render the plane incapable of flying, but it did remove one level of idiot proofing. As a result one of the pilots pulled the stick back and didn't relent even as the plane stalled. If you did that in a Boeing plane (or any plane) with functioning sensors you'd have the same end result.

mlyle 4y ago
Neither did the MAX debacle render the plane incapable of flight. Both flights gave flight crews unexpected tests on unexpected situations-- something that seemed like runaway trim for the MAX, and simultaneous overspeed (better nose up) and stall (better nose down) indications for AF447.
- rocqua 4y ago
  Test were done with MCAS regarding how soon you have to react before it becomes unrecoverable. This was 10 seconds and it was hard for boeing-trained pilots instructed in MCAS to recover from this in tests where they knew it would happen. The problem was that at some point the trim is so much that it becomes nigh impossible to hand crank the trim down, the trim flap applies so much pressure it cannot be overcome by hand. You had to disable mechanized trim because otherwise MCAS would kick in. So you would need to recognize the problem soon enough to prevent MCAS from moving the trim so far that it became unrecoverable. (Source: Netflix documentary)
  - mlyle 4y ago
    > This was 10 seconds 10 seconds is a long time of the trim wheel being cranked nose up. The same situation would apply to a stuck trim switch on the yoke. (Indeed, a stuck trim switch is worse in this respect because it cranks constantly and doesn't "yield" to the pilot pushing the switch the other way... of course, the checklist didn't presume a pilot should be diagnosing stuck AOA vs. stuck trim switch and thus recommended immediate use of the trim cutout switches instead of trying to electrically trim the other way). There's a reason that trim runaway is a "recall" procedure in the QRH that pilots are expected to know-- it gets scary. http://www.b737.org.uk/images/runawaystab2020.jpg http://www.b737.org.uk/images/runawaystab2020.jpg But it was the assumption before the MCAS disaster that pilots could deal with trim runaway, given that there was a history of flight crews dealing with trim runaway incidents in flight successfully. But MCAS was worse because: A) It did it more often, so flight crews were exposed to this dangerous situation more than ever before B) It did it in a sneakier, strange way (no constant movement of the trim wheel). Still, some real-world flight crews did overcome MCAS doing the wrong thing without crashing, even if others failed the test.
    - rocqua 4y ago
      I imagine point B is quite relevant here. MCAS going wrong is meaningfully different (and less immediately scary) from the recall procedure that the training doesn't kick in.
      - mlyle 4y ago
        Sure. The main points I have, I guess, are: - It's perhaps understandable why engineering teams-- pushed by management to deliver a 737 that doesn't require retraining-- decided that if flight crews could handle runaway trim, they could handle any errant MCAS activations and that the level of redundancy and fault tolerance necessary in the system would be low. - On the other hand, real flight crews obviously do not do great when exposed to this test. - If you give a flight crew a wacky situation with conflicting information (like AF447 or the 737 MAX)-- expect failures, even if the situation is technically 100% recoverable and flyable and some crews exposed to the situation succeed. It's great that we have humans capable of reason up in the front of the plane, but reason is often faulty and doesn't get you the solution you need on the first try.
nopzor 4y ago
you couldn’t even physically even do that in a boeing plane, because of sensible hmi choices that boeing made early in their history. so i think your point is completely invalid.
- inferiorhuman 4y ago
  Couldn't do what? Pull the stick back into a stall? Sure you can. Under normal law you can't do that with an Airbus, that's one of many things that caught the AF crew off guard. There's no such protection on the 737. Nobody's done that because unlike on the Airbus, it's never safe to do so. There are, however, plenty of ways to get caught off guard in a 737 and stall it. Take a look at the UTAir and Thomsonair go arounds that got their planes to about 45° nose up. Or the 737 Air Florida planted right into the Potomac because they didn't have the elevator authority to avoid the stall. Or Fly Dubai… As for human interactions, the 737 is a series of bad choices (especially the MAX). Take a look at the Helios Air crash, the pilots were confused by an alarm that could mean two different things. Unfortunately hypoxic pilots aren't known for their reasoning ability.
  - cjrp 4y ago
    Couldn't have both pilots fighting each other on the controls (hard nose-up vs. hard nose-down) without knowing about it, since the yokes are physically linked (unlike sidesticks).
    - nopzor 4y ago
      yes, this is what i meant. i should have been clearer in my original statement. here is what happened in the airbus air france incident: - one pilot had the sidestick full hard nose up, the other full hard nose down. - neither pilot realized what the other pilot was doing. on a boeing airliner: - the controls for each pilot are physically linked. - if pilots try to put “their” yolk into an opposing position, they will start having to physically wrestle for control. - this type of failure mode of the air france crash wouldn’t have happened in a boeing airplane; the pilots would have realized what was going on within seconds. general note: - when flying, it is highly unusual for both pilots to be giving input into the controls at the same time. one pilot “has the airplane”, the other is there to assist, help with the radios, planning, etc. the handoff between who “has the airplane” is very deliberate. this was another failure mode of the air france flight. - but my point still stands, the fact that the controls are linked makes the gp point a bit moot.