Why is automation dependency encouraged in modern aviation ?

[Check Airman]

I think you’ll find that you can do more work when using both arms.

[FlyingStone]

Originally Posted by KayPam

Sorry if it's not clear or grammatically incorrect, but I said "constant force feedback". The spring in the sidestick does provide a constant force feedback, doesn't it ?
Or would it be more grammatically correct to say "constant force feel" ?

Where’s the “feedback” coming from? Does the Airbus sidestick provide you with feedback so you can tell the difference in load on the flight control surfaces between 0kts and Vmo?

[Capewell]

there is no ' force feedback' at all on the Airbus sidestick. it is spring loaded so you need to apply progressively more force to it until it hits the stops

[Vessbot]

*sigh*. There's 2 parallel lines of discussion going on -

1) The terminology of "force feedback" - of which the usage is incorrect, since there's no information from the result feeding "back" to the input.

2) The actual intent of KP's post, despite the wrong word being used (so let's just forget it was used, and pretend it was "constant force"), the implications of the constant force ergonomics on cues to the pilot. Personally I think it's a double edged sword. It prevents overwhelming forces that can add stress and confusion to the pilot (this was KP's post) but at the same time (I think, based on pure speculation, not having flown such a system) it hides meaningful information, which can add its own source of confusion.

[Capn Bloggs]

Originally Posted by Vilas

Airbus GA is no problem or for that matter even flying because aircraft trims itself. It even applies to alternate law

That wasn't exactly advantageous in AF447. Had it been a normal aircraft, the crew would have had to consciously trimmed back into the stall. The A330 did it automatically, to full nose up, which is just what they didn't want or need.

[Vessbot]

Originally Posted by Capn Bloggs

That wasn't exactly advantageous in AF447. Had it been a normal aircraft, the crew would have had to consciously trimmed back into the stall. The A330 did it automatically, to full nose up, which is just what they didn't want or need.

That may be somewhat of a factor, but the opposite would not have been a slam dunk prevention. Cf. Asiana 214 which was a "normal aircraft" yet the pilot's consciousness did not extend to the elevator force provided to them by the system, and still drove it into a near stall. (Without the ground there to cue a reaction first, there's every reason to think they would have continued pulling back all the way to the stall warning/shaker/limiter/whatever. edit: they did get the shaker)

Any stupid thing that an Airbus airplane is able to do automatically, a pilot is able to do manually. Well, at least this stupid thing. Otherwise there would have been no stall accidents prior to the 1980's.

[Capn Bloggs]

Originally Posted by Vessbot

(Without the ground there to cue a reaction first, there's every reason to think they would have continued pulling back all the way to the stall warning/shaker/limiter/whatever.)

I doubt it. They ONLY reason they kept puling back was to reach the runway which was in plain view straight ahead. The AF447 crew had no outside reference to entice them to pull back, they were simply climbing and slowing down, and as they did the jet trimmed back for them. In a normal aeroplane, the nose would have dropped if the pilots had done nothing.

[Vessbot]

Originally Posted by Capn Bloggs

I doubt it. They ONLY reason they kept puling back was to reach the runway which was in plain view straight ahead.
The AF447 crew had no outside reference to entice them to pull back, they were simply climbing and slowing down, and as they did the jet trimmed back for them. In a normal aeroplane, the nose would have dropped if the pilots had done nothing.

You're comparing the two different reasons for pulling back, and that difference doesn't really matter. The point is that they did, and the likeliness of that being mitigated by auto trimming vs. not. We can see that lack of auto trimming is far from a guarantee against the pilot applying obscene amounts of force manually. I grant that the stick force, though useless in the Asiana case, does give some chance of providing a cue. But how much chance? I dunno. (Overall I'm in favor of the speed stable feedback design, but that's cheap words since I've only ever flown that kind.) Also for AF447 speculating on what the nose would have done "if the pilots had done nothing" is moot, as they didn't do nothing, they pulled all the way back and the result would have been largely the same if it was in a conventional airplane. (Same for the less famous crash of Air Asia 8501, a virtual duplicate of AF447).

I think a likelier benefit to both of those (447, 8501) would have been if the controls were physically interlinked, and the tunnel visioned CA might have been clued in to what the FO was inputting. In both cases, the CA was making some half-hearted forward pushes, so he had the right idea. But there was no meaningful nose-down response due to the FO overpowering him hard against the aft stop. If the CA knew the reason, they could have yelled for the FO to stop and/or push forward, as he pushed harder forward himself.

[vilas]

Originally Posted by Capn Bloggs

That wasn't exactly advantageous in AF447. Had it been a normal aircraft, the crew would have had to consciously trimmed back into the stall. The A330 did it automatically, to full nose up, which is just what they didn't want or need.

Let's not get 447 into this. First in Airbus you never keep the stick out of neutral on permanent basis. Any Airbus pilot knows if you keep pulling back she will keep trimming up. It's not a secret. The way to stop is leave the stick once the desired pitch is reached. Unfortunately just like the SFO crew never checked their speed the 447 crew never looked at the PFD. Even a B757 which doesn't auto trim has crashed with UAS. The subject is developing/ maintaining raw data flying skill. No one says pilot shouldn't have it. The difference is on practicing how much to, what to and where to. Naturally every one has different views. But no point quoting accidents that have nothing to do with the subject.
​​​​​​

[Vessbot]

Originally Posted by vilas

Let's not get 447 into this. First in Airbus you never keep the stick out of neutral on permanent basis. Any Airbus pilot knows if you keep pulling back she will keep trimming up. It's not a secret. The way to stop is leave the stick once the desired pitch is reached. Unfortunately just like the SFO crew never checked their speed the 447 crew never looked at the PFD. Even a B757 which doesn't auto trim has crashed with UAS. The subject is developing/ maintaining raw data flying skill. No one says pilot shouldn't have it. The difference is on practicing how much to, what to and where to. Naturally every one has different views. But no point quoting accidents that have nothing to do with the subject.
​​​​​​

I think it has much to do with the subject. In answering "how much to" fly raw data, doesn't it weigh in as a criterion, for the result to be a successful suppression of the urge to pull up 10 degrees and 6000 fpm (or, even more disturbingly in AirAsia 8501, 45 degrees and 11000 fpm) from cruise?

[Check Airman]

Originally Posted by vilas

But no point quoting accidents that have nothing to do with the subject.​​​​​​

I'd say AF 447 has EVERYTHING to do with the subject. A pilot who never does any raw data flying is a slave to the FD/FMA. The instrument scan becomes a stare. On AF447, the FD commanded a nose up input, and the PF said something like "I've been pulling the whole time".

We'll never know for sure if he was trying to follow the FD, but regular raw data flying allows you to more easily look behind the FD, and also allows you to more readily realise when it's time to turn it off.

Over the ocean in essentially night IMC is when some people would probably say is not ideal to practice raw data flying. However it's exactly the situation encountered by AF447.

If you don't feel comfortable hand flying an approach in IMC, you should let your regulator know there needs to be a DAY VMC ONLY limitation on your ATPL.

[vilas]

Originally Posted by Check Airman

I'd say AF 447 has EVERYTHING to do with the subject. A pilot who never does any raw data flying is a slave to the FD/FMA. The instrument scan becomes a stare. On AF447, the FD commanded a nose up input, and the PF said something like "I've been pulling the whole time".

We'll never know for sure if he was trying to follow the FD, but regular raw data flying allows you to more easily look behind the FD, and also allows you to more readily realise when it's time to turn it off.

Over the ocean in essentially night IMC is when some people would probably say is not ideal to practice raw data flying. However it's exactly the situation encountered by AF447.

If you don't feel comfortable hand flying an approach in IMC, you should let your regulator know there needs to be a DAY VMC ONLY limitation on your ATPL.

It's not that simple. UAS is one of the most challenging failured because it doesn't come in standard format like hydraulic or engine failures(Ask Uplinker. He will tell you)There is definite procedure for it and if you didn't know it then when it happens you are not likely to discover by accident(pun! May be). There were more than 22000 posts on 447 so I won't go over it. There's no evidence that 447 pilots didn't know how to fly raw data. OK! they followed the FD and stalled. Did they know stall recovery or they needed actual practice at 370? They had to know how to deal with UAS which they didn't. They didn't even know stall recovery procedure. They knew FDs don't recover you from stall or did they? So let's forget 447. Nobody is saying don't fly raw data approaches. But automation is there and installed for day to day use and not only as a standby to daily raw data practice flying. It's a commercial flight and not a training flight. Everything has to have a sense of proportion.

[Check Airman]

Why do commercial flights and training flights need to be mutually exclusive? When you go to the hospital, the doctors who see you may well be in training. Even after training, as consultant or attending physicians, don't you agree that they're honing their skills with each new patient?

Nobody's suggesting we go around pulling breaker to see what happens. I don't even consider flying around with 0 automation to be training, really- it's more about maintaining a firm grasp on an ablatable skill set- one which we may need to call upon without as much as a moment's notice.

[vilas]

Originally Posted by Check Airman

Why do commercial flights and training flights need to be mutually exclusive? When you go to the hospital, the doctors who see you may well be in training. Even after training, as consultant or attending physicians, don't you agree that they're honing their skills with each new patient?

Nobody's suggesting we go around pulling breaker to see what happens. I don't even consider flying around with 0 automation to be training, really- it's more about maintaining a firm grasp on an ablatable skill set- one which we may need to call upon without as much as a moment's notice.

I am sure you can give better example than Doctors. Sure! They practice their skills on sick patients but not on a healthy person making him sick just for practice. I understand what you are trying to say but I don't see it as that much of a problem to be so passionate about to suggest instruments to be redesigned to allow pilots to fly raw data. In Airbus it's very rare to do at moment's notice. AP off not a big deal it will hold it's flight path. In Airbus one is better off keeping one's hands off till you get your wits about. Again 447! in alternate law all they had to do was just keep wings level, ask for the Capt to return and gracefully exit the scene. QZ8501 is even more bizarre. They engineered a crash by trying to be engineers by fiddling with CBs (which is forbidden), they caused alternate law, instead of controlling the bank they caused the pitch up which on it's own the aircraft wouldn't, they stall the aircraft, they don't recover. There's a lot to learn from this but raw data has nothing to do. Let's leave the accidents. Absurd but someone may say had the 447 PF reported sick it would have prevented the accident. Isn't it true?

[vilas]

Kay Pam I say again practice double SFCC slat channel Fail approach and all the scan required will be there.

[Checkboard]

UAS = Unmanned Aerial System? Upper Air Stalls? University Air Squadron?

[hec7or]

UAS Unreliable Air Speed

[KayPam]

I'm not sure accidents have nothing to do with the subject..
They show that a problem exists.

Have you ever seen a general aviation accident where the aircraft stalled from a high cruise altitude and never recovered ?
This does not mean, obviously, that general aviation is exempt from piloting errors, but they have much less training (PPL vs CPL or ATPL + TR), much less practise (average of 12 hours vs 600 hours per year), and only pilot required.
So flying errors are much more suprising for commercial aviation. It looks like we don't make the most of our advantages (two pilots, and potentially far higher practise)

[PEI_3721]

To roundoff a few previous comments:-
KP, #166,'The 9 competencies are' - a model with neat definitions, but what exactly are these aspects in practice, how do we explain them, train them. '… better workload management (because of a higher workload which will make you practise this competency) and situational awareness'. Yet humans continually have difficulty with identifying high workload or when we have 'good' awareness' or even the correct awareness; or when will understanding be sufficient vs an 'optimum' ideal.
Our views are closer than might appear; not what they are, but how we get to them.

Vessbot, #169 re 'skills being transferable'. Ideally it would be convenient to acquire skills in one situation that would transfer to others. However, the research report indicates that some skills may not transfer; page 63 onwards.
Perceptual-motor skills, manual flight handling, feel of the aircraft, could transfer between approach and GA, but less so the mental skill of knowing when to GA, or knowing that the aircraft will 'intersect the ground 2300 feet short of the runway', knowing that the need is to 'level the wings', - the orientation (the understanding) part of OODA.
Note that the report relates advance proficiency with expertise, a deffiniton not alway used in aviation.
Also, and importantly, that task is considered in different ways; manual task includes both cognitive and motor skills - tactical, whereas flying involves manual tasks and higher order cognitive skills in awareness, understanding, and decision making - strategy. Higher order skills should transfer, but rarely do - we may not have them to begin with (training) or with the basics they are not practiced, improved, higher levels of expertise.
Re 'regulator beliefs', see refs - as we choose to interpret them.

A conclusion of sorts: the issue is less of 'know what' but more of 'know how', tacit knowledge, experience from being there, doing it, remembering that something has been achieved.

Like riding a bike; tell me, I fall off, but having fallen off and continuing to seek the skill, there comes success, but still being unable to explain how this was achieved.

Refs:- https://www.icao.int/Meetings/a40/Do.../wp_296_en.pdf
https://ad.easa.europa.eu/blob/SIB_2.../SIB_2013-05_1
https://www.faa.gov/other_visit/avia.../SAFO17007.pdf

[FlyingStone]

Originally Posted by KayPam

Have you ever seen a general aviation accident where the aircraft stalled from a high cruise altitude and never recovered ?
This does not mean, obviously, that general aviation is exempt from piloting errors, but they have much less training (PPL vs CPL or ATPL + TR), much less practise (average of 12 hours vs 600 hours per year), and only pilot required

GA aircraft flown by PPLs typically aren't complex, heavy multi engine swept-winged jets flying at FL400.
​​​