ATR, Nepal, Jan 2023

[3WE]

Ban ALL turns to final, and just land on a conveyor belt..

Fixed.

[ocelot]

Do you feather props with the same motions as the NEARBY flap handle (you French aeroengineers did WHAT?)

Look at the picture: almost exactly the same motion from flaps 15 -> 30 as condition from auto -> feather. As someone said on avh, not a masterpiece of engineering design.

TBH, if I were designing that cockpit I wouldn't have condition levers on the pedestal at all; it's not like you're actually adjusting the prop pitch with them while you're flying and they are large and take up valuable real estate. But nobody asked me

[ocelot]

Another point about muscle memory is: it will tell you how to move the levers under your hand, so the fact that they engage differently (and are a different shape, and two of them instead of one) won't necessarily stop you.

Ideally if you do that the contradiction between what you just did and what you meant to do will register, but usually that sort of thing registers as that "something isn't right" feeling, and if you're flying in the wrong seat it's probably all too easy to write it off as that instead of immediately crosschecking everything to try to figure out what you _did_ do just in case it was bad.

Having the plane announce what it's doing as you push buttons seems like a good thing in general. But it won't be a panacea; you'll get used to it reacting to what you're doing and not necessarily notice right away if it says something you didn't expect, especially if you're already fatigued. However, if _both_ pilots hear it that's an extra safety factor.

[elaw]

Some interesting analysis from a pilot/instructor on ATR aircraft: https://youtu.be/wIlO-TBDyaw

[Gabriel]

Some interesting analysis from a pilot/instructor on ATR aircraft: https://youtu.be/wIlO-TBDyaw

Yep, and he can't believe that he did that mistake. Position in the quadrant, size, height, shape, count, and way to operate (lift from the sides vs squeeze the triggers underneath) all seem very different for this expert with tons of hands-on experience.

I repeat: I believe that the problem was unconsciously intending to operate the wrong levers, rather than going for the right lever and missing.
I repeat: No amount of differentiation will avoid that. Like when the flaps were retracted upon the gear up call just after lift off. Levers positions, shapes, and operation just don't get much different than that.

[3WE]

Some interesting analysis from a pilot/instructor on ATR aircraft: https://youtu.be/wIlO-TBDyaw

Great find. He mentioned that they were low on downwind…and it does look tight in terms of a short final…so hurry-up could be a factor…

Repeating- I thought flaps selected AND INDICATED was more universal…and it appears he didn’t follow though.

New thought- doesn’t the ATR give a nice whine and feeling of braking when you go to full flaps?

Being mildly dyslexic- the instructors mind could have said “two right levers”, pull the far one”, and Ocelots double muscle memory mechanism kicked in.

Comments on the video hinted that the instructor might have done it deliberately.

The video made an excellent point that the actual conversation might be key.

/Rambling

[ocelot]

I've been thinking not so much missing as transposing incorrectly for left/right. Which is easy to do if you're flying the wrong side _and_ you've been doing it all day and you're no longer watching out for the differences. Then once you grab the wrong lever your hand knows how to move it, especially if you're not really paying much attention because there's too much else going on.

But, just reaching for the wrong handle is plausible too, and this is not something we're ever going to find out.

[elaw]

Some more thoughts on the subject from one of our favorite aviation speakists: https://youtu.be/p5m8kvZqwC4

[Gabriel]

Some more thoughts on the subject from one of our favorite aviation speakists: https://youtu.be/p5m8kvZqwC4

Exactly.

[3WE]

I appreciated his stories of dudes doing the wrong thing on not_adjacent controls.

Makes me rethink the AI “monitoring pilot” who says, “Feathering the props is highly irregular, Dave”.

Also, another twist on this incident was the gal saying “no power”…my fantasy is that two pilots could take a quick look at everything and find the problem a little quicker?

Or, we resign ourselves that one in a few million times, excrement transpires, in spite of our best efforts…

/Pontification.

[elaw]

Not to worry... from what I read, by next year ChatGPT will be flying all the airplanes and our troubles will be over.

[3WE]

So I went to AvHeraldie and yeah, Evanie is sooooo happy that the report said “high workload, right turns and especially not_stabilized criteria. were contibuting.”

His viewpoint still irritates me. I didn’t read of any banks greater than 30 degrees. And I don’t_concur that flying an ATR into a nice airport “172 style” is “high workload”.

Apparently, they did roll out on final below the exalted 500 ft stabilized level, but maybe that’s because the power was off? And I guess both should have said “go around” instead of “continue” (paraphrasing).

Yeah, adherence to procedure is important, but do spare me the intentional bank/accelerated stall. I reviewed the video and the nose up plowing still bugs me…not faulting the crew- they were simply trying to maintain altitude.

[elaw]

Possibly an interesting twist? https://www.youtube.com/watch?v=u989uFh8H_8

[3WE]

Frenchman speakist.

Very thorough (of course, they tend to be.)

I dunno…it seems like they flew “powerless” for quite a few seconds, and it was a shame they couldn’t fix it.

I’d like to suggest improved procedures, but sometimes, trucks attack bicycles.

[Gabriel]

And I guess both should have said “go around” instead of “continue” (paraphrasing).

They did, eventually. It didn't work because a) too late and b) go around doesn't work very ell with feathered props, you tend to stall-spin-crash-burn-die.

[Not_Karl]

...sometimes, trucks attack bicycles.

Ban them both and also ALL aeroplanies.

They did, eventually. It didn't work because a) extreme bank angle and, as lesser, contributing factors b) too late and c) go around doesn't work very ell with feathered props, you tend to stall-spin-crash-burn-die.

Fixed.

[Gabriel]

...sometimes, trucks attack bicycles.

Ban them both and also ALL aeroplanies.

They did, eventually. It didn't work because a) extreme intentional bank angle, b) flying a visual pattern in extreme visual conditions and, as lesser, contributing factors c) too late and d) go around doesn't work very ell with feathered props, you tend to stall-spin-crash-burn-die.

Fixed.

Fixed

[Not_Karl]

...sometimes, trucks attack bicycles.

Ban them both and also ALL aeroplanies.

They did, eventually. It didn't work because a) extreme intentional bank angle required to align with the runway, b) flying a visual pattern in extreme visual conditions and, as lesser, contributing factors c) too late and d) go around doesn't work very ell with feathered props, you tend to stall-spin-crash-burn-die.

Fixed.

Fixed

Expanded.

[3WE]

And I guess both should have said “go around” instead of “continue” (paraphrasing).

Evan used the term “consumed”. I don’t disagree. But I think it’s more “consumed” with a mildly challenging landing, as opposed to “blasting wantonly into danger against all good wisdom.”

Today I will fault this procedural deviation: “Flaps 30,” “Selected and indicated.. Instead it was, yeah, I grabbed that lever and muscle memoried it.

Also, I’m feeling sympathetic: A beeping light that says no electrical power doesn’t tell ME the props are feathered (of course, this isn’t a 172M).

[elaw]

Also, I’m feeling sympathetic: A beeping light that says no electrical power doesn’t tell ME the props are feathered (of course, this isn’t a 172M).

I'm definitely with ya on that one in particular. We're conditioned to view "electrical" as a specific category of trouble, and "incorrectly set propellers" definitely does not fall into that category.

IMHO they really should have a light marked "Hey, McFly, your props are feathered, you're not on the ground, and your airspeed is rapidly decaying toward a stall... is that what you really want to do?"

Of course that would probably not fit so the indicator instead would have to be labeled "H, MF, YPAF, YNOTG, &YAIRDTAS - ITWYRWTD?". That would not be confusing at all*.

* Edit: scratch that idea. I just realized it's too easy to confuse with a METAR.

[Gabriel]

Also, I’m feeling sympathetic: A beeping light that says no electrical power doesn’t tell ME the props are feathered (of course, this isn’t a 172M).

I'm definitely with ya on that one in particular. We're conditioned to view "electrical" as a specific category of trouble, and "incorrectly set propellers" definitely does not fall into that category.

IMHO they really should have a light marked "Hey, McFly, your props are feathered, you're not on the ground, and your airspeed is rapidly decaying toward a stall... is that what you really want to do?"

Agreed. Regardless of whether the feathering was manual or automatic, there should be a couple of prominent flashing lights
coordinated with an aural alarm that will only stop if the condition is rectified, or if it is acknowledged by pressing said lights.

Of course that would probably not fit so the indicator instead would have to be labeled "H, MF, YPAF, YNOTG, &YAIRDTAS - ITWYRWTD?". That would not be confusing at all*.

* Edit: scratch that idea. I just realized it's too easy to confuse with a METAR.

Funny, I thought NOTAM. But I think just L PROP FTHR and R PROP FTHR would suffice.

[3WE]

Evan free advice mode:

Retrofit planes that vocalize input:

Flaps 30 set, 30 indicated.
Propellor feather

And I had a huge circular argument with Gabriel on Artificial Intelligence ENABLING this. Again- not AI per se, but LESS CRYPTIC talking interfaces, and some programming.

What if the Air France plane said: Unreliable airspeed. Your airplane in 10 seconds…instead of world ending beeping and flashing lights?

[Gabriel]

What if the Air France plane said: Unreliable airspeed. Your airplane in 10 seconds…instead of world ending beeping and flashing lights?

And what about the system shouting STALL STALL!!! instead of having a neutral buzzer, and have it sound interruptedly for 1 and 1/2 minute?
Oh wait...

No, I am joking. We agreed on the plane talking more clearly to the pilots. The disagreement was that for you the implementation path was AI and for me it was traditional coding, but we wished for the same effect.

[3WE]

The disagreement was that for you the implementation path was AI and for me it was traditional coding, but we wished for the same effect.

No.

[ocelot]

There's no reason to get AI stuff involved just to have the computer talk. Computers have been talking for almost 40 years. Anyone remember "Kennedy Approach"?