Why is automation dependency encouraged in modern aviation ?

[sonicbum]

Originally Posted by Centaurus

I remember talking to a friend of mine who was a Boeing 737 flight simulator instructor at Boeing Seattle. He told me his story of talking to the then chief test pilot of the Boeing 787 who said "We designed the 787 knowing that it will be flown by incompetent pilots. For this reason we have all the automatics protections designed into the aircraft."

Doesn't that say it all?

Very unlikely he said that.
ICAO defines an incompetent pilot as someone who needs additional training to reach proficiency before resuming flight operations. I am quite sure the Boeing 787 chief test pilot is aware of that.

[Uplinker]

I think the general philosophy is to design for the average pilot on a fleet rather than the most competent?

An incompetent pilot* should not be allowed anywhere near a real aircraft in the first place.


*(which is an oxymoron anyway),

[sonicbum]

Originally Posted by KayPam

(Does boeing allow to fly a RNAV trajectory on raw data only?)
But didn't manufacturers (at least airbus) create incompetent pilots by depriving them of the means to fly 95% of the trajectories that they fly ?
The industry as a whole encourages to use automatics because they have a precision that pilots will never have. But this does not mean that pilots should not practise their skills to the best level that they can.

Why ?
You've seen that you can do it.
I've seen many pilots do it. The first captain that I flew with after my line check disconnected everything while in descent FL100. (He was able to do that because we not flying an RNAV trajectory obviously..)
And if you don't do it enough, your skills will erode.

The one thing that I learnt during basic IFR training is that manual flying takes less and less resources the more you do it.
It should indeed be mandatory to know how to handle things manually, to be able to detect when the FG does sh*t and to handle the situation correctly when it goes wrong.
What if a pilot that never flies manually encounters a situation where the airplane reverts to direct law ? It happens, just any failure downgrading to alternate, then gear down will leave you in direct law.

However, I disagree about the sim part.
Flying raw data in the sim is good practise for real flying. The only problem is the amount of sim practise. Two line checks per year, the majority of which is spent managing failures (leaving only 2 hours of manual sim flying per pilot per year) is obviously not enough.
But if you work tens of hours in the sim (in combination to normal line flying), you will be a decent pilot in the aircraft.
When my colleagues and I did our base training, we had spent 16 hours each (or so) preparing for it in the sim. When we touched the real aircraft for the first time, the most surprising thing was the ground handling qualities (Airbus itself admits that they don't really study "ground handling qualities"), not the stick and rudder part. We were obviously not perfect, but we all had a decent level, at least given the fact that we never had touched a jet aircraft before.

Few points :

1) manufacturers create safe and efficient airplanes by using the best combination of technology available at that time and upgrading it as time goes by. They also train pilots -if required- to be competent and proficient.

2) You can't practice ALTN or Direct Law on the line for obvious reasons ; hence flying manually during normal line ops will generally do more harm than good (increases workload, reduces spare capacity for other tasks, etc..). It is beneficial -generally speaking- only in the case of cadets to get a better feeling on the machine.

3) From a training and confidence point of view, is it a good idea to visit the SIM twice a year for a total of 3 or 4 days ? NO.
BUT governing bodies across the world prescribe it as an acceptable minimum, and since operators do not like to throw money out of the window, they generally comply with that.
Think about crosswind landings : you can hand-fly as much as you like on the line, but if you never encounter crosswind conditions, it is likely that your crosswind landing technique will need a brush-up. The only way to practice it would then be in the SIM, otherwise when You are out and face a max crosswind landing when the toughest thing you have done in the past 6/7/8 months is 10 kt cross you will be definitely sweating.

[vilas]

I just know that a few minutes in the sim every few months would leave me woefully unprepared should the lights start flashing on a stormy night

Check Airman you missed something from what I said. I said about raising the standard of pilots in raw handling before they come on board and more frequent visits to sim to retain that and not for few minutes, I didn't say that. Real aircraft raw data approach on two engine is easier that raw data with OEI and ATHR OFF in sim. Isn't it all about scan? If someone is terrible on two engines in real aircraft will be all over in the sim also. Those who are starting fresh will need more practice than those who are trying to just maintain it. Besides if opportunity presents practice by all means.
Centaurs, it all started with Bernard Ziegler who said "I am making an aircraft that will not allow pilots to make mistakes. Any one can fly it even my concierge can fly it".

[FlyingStone]

Originally Posted by KayPam

But, it would be completely feasible to allow rnav approaches without FDs : just display an HSI : horizontal situation indicator, which would indicate the desired RNAV track, and a lateral deviation bar in nautical miles (instead of degrees on a VOR), and eventually a timer that indicates when to start the turn (like Garmin does). That's it.

The old 737 EFIS classics had this (optionally, I believe). With the world going towards PBN and curved (RF) procedures, this wouldn't be particularly useful.

Originally Posted by pineteam

Yes by regulation you can't do raw data take off for RNAV SID but it's not an excuse. Just request conventional departure!

Simple, just don't fly into any of noise sensitive or busy airports, or airports without conventional SIDs. Easy peasy.

Originally Posted by KayPam

What if a pilot that never flies manually encounters a situation where the airplane reverts to direct law ? It happens, just any failure downgrading to alternate, then gear down will leave you in direct law.

You can do 10 go-arounds in normal law on the A320, yet the one in direct law will have to look significantly different when it comes to control inputs. Apples and oranges.

I'm a big fan of hand flying and raw data (latter prohibited by employer unfortunately), but the reality is that (before Covid) we operate in ever busier airspace and at the end of the day, people in the back or shippers of freight pay us to deliver them or their goods safely and efficiently from A to B, not to have fun or to prove to somebody that we still have "what it takes". Could I handle flying raw data in and out of London TMA during a busy morning period? Yes. Is it a smart thing to do in real life, assuming you've got AFDS serviceable? I hope we all agree what the answer should be here.

PBN is the way forward for noise/fuel/emissions/cost/safety (not necessarily in this order) reasons, whether we like it or not.

[vilas]

and at the end of the day, people in the back or shippers of freight pay us to deliver them or their goods safely and efficiently from A to B, not to have fun or to prove to somebody that we still have "what it takes"

Absolutely! I​​​​​​ keep saying this all the time. Line flying is not a training flight. Some training is inevitable but should be kept to minimum. We are there not for ourselves but for a purpose.

[pineteam]

Come on guys flying raw data is not training. It’s maintaining his flying skills. Just because the aircraft is full of automations does not mean you should use it all the time. In that case we should perform autoland all the time since the aircraft can do it beautifully. Of course trying to do raw data it in a busy airspace after 10hr of duty would not be smart to say the least. But on a good cavok day with a sharp set of crew I really don’t see any safety issue down there..

[Two's in]

Why is automation dependency encouraged in modern aviation ?

Because killing passengers is bad for business.

[Check Airman]

Originally Posted by Two's in

Because killing passengers is bad for business.

If raw data equates to killing passengers, that person shouldn’t occupy a seat in the cockpit.

[Check Airman]

Originally Posted by vilas

Absolutely! I​​​​​​ keep saying this all the time. Line flying is not a training flight. Some training is inevitable but should be kept to minimum. We are there not for ourselves but for a purpose.

Stateside, we do a great deal of training on the line. The sim only goes so far. You start your line training with a training CA. Thereafter, the training continues with line captains / FO’s who teach you how it’s done. The “training” footprint would be enormous otherwise.

[Check Airman]

Originally Posted by vilas

Check Airman you missed something from what I said. I said about raising the standard of pilots in raw handling before they come on board and more frequent visits to sim to retain that and not for few minutes, I didn't say that. Real aircraft raw data approach on two engine is easier that raw data with OEI and ATHR OFF in sim. Isn't it all about scan? If someone is terrible on two engines in real aircraft will be all over in the sim also. Those who are starting fresh will need more practice than those who are trying to just maintain it. Besides if opportunity presents practice by all means.
Centaurs, it all started with Bernard Ziegler who said "I am making an aircraft that will not allow pilots to make mistakes. Any one can fly it even my concierge can fly it".

We agree that more frequent sims would be good. My point is that it’s a perishable skill though, so let’s say we visited the sim twice as often, it’d still leave your skills lacking on that last flight the day prior to the sim.

In reality though, you can practice a lot of raw data flying on the line in normal operations. Why visit the sim for something that can safely be done on the line?

PS- I didn’t mean to imply that you said a few minutes a year was sufficient. Apologies. You’re of course correct that somebody who’s new needs more raw data flying. I’d submit that even somebody who’s experienced needs it as well. I’ve watched things fall apart (on the line and in the sim) when an experienced pilot loses the FD.

[Olympia463]

Hand Flying

I'm a glider pilot, a very old glider pilot in fact. I have 2200 sorties in my log book, on most of which I was hands on 100%, except when my pupil was doing the flying. Lest you think I never did long flights, I did stay up for 5 hours once in thermals (a Silver badge requirement), and made many many other flights in excess of three hours.. I retired from gliding over forty years ago, but in 2007, after a 30 year break, I went back to a local gliding club for a weeks flying to see if flying was like roller skating - something you never forget how to do. And it IS like roller skating - I re-soloed in the early part of that week and spent the rest of it cruising around the beautiful hills of Scotland. My point is that manual skills are never forgotten once you have them. But, and is is a big but. you have to acquire these skills. I would suggest that airline pilots should join a gliding club and either fly the gliders or tow them. I'd favour doing both actually. That way you will keep your manual skills in good trim. It is very significant that Sully was a glider pilot at one time, and the 'Gimli glider' was also conveyed to the ground in one piece by two ex-glider pilots. You could say that every flight in a glider is a controlled crash, but that isn't strictly true. Gliding is real flying - you become part of the plane - something you will never be if the plane does all the flying for you. Lest you think I flew only in good weather in the summer, I flew in every month in the year in the UK and in every kind of weather and had no accidents.

I hear the cry of you modern pilots missing the handling of the controls. The idea that automation will ever totally replace human pilots is rubbish, but you guys have got to make an effort to maintain your raw flying skills at all costs for the day or night when something you were not expecting happens.



I

[KayPam]

Originally Posted by vilas

Check Airman you missed something from what I said. I said about raising the standard of pilots in raw handling before they come on board and more frequent visits to sim to retain that and not for few minutes, I didn't say that. Real aircraft raw data approach on two engine is easier that raw data with OEI and ATHR OFF in sim. Isn't it all about scan? If someone is terrible on two engines in real aircraft will be all over in the sim also. Those who are starting fresh will need more practice than those who are trying to just maintain it. Besides if opportunity presents practice by all means.
Centaurs, it all started with Bernard Ziegler who said "I am making an aircraft that will not allow pilots to make mistakes. Any one can fly it even my concierge can fly it".

Did you know that Ziegler killed several people while flying recklessly close to a ski lift ?
AF447 proved that his goal was not reached. A few ice crystals and you need a competent pilot in the front seat.

Originally Posted by FlyingStone

You can do 10 go-arounds in normal law on the A320, yet the one in direct law will have to look significantly different when it comes to control inputs. Apples and oranges.

I'm a big fan of hand flying and raw data (latter prohibited by employer unfortunately), but the reality is that (before Covid) we operate in ever busier airspace and at the end of the day, people in the back or shippers of freight pay us to deliver them or their goods safely and efficiently from A to B, not to have fun or to prove to somebody that we still have "what it takes". Could I handle flying raw data in and out of London TMA during a busy morning period? Yes. Is it a smart thing to do in real life, assuming you've got AFDS serviceable? I hope we all agree what the answer should be here.

True, direct law will feel very different. But if you're used to setting a pitch/power couple, and monitoring all flight parameters with an efficent scanning, then it shouldn't be too difficult to manage.

At my airline we clearly have two types of airports. Large capital airports like Paris, London, Amsterdam, etc.. and also smaller regional airports in our home country or elsewhere, like Porto, Nice, Bucarest..

It is important to underline that I don't want to hand fly just for fun. It is important to practise hand flying for safety reasons.
There were several incidents where pilots put the aircraft in severely complicated positions, due to them following the flight guidance system at a time where it malfunctioned. The capture of a false glide is a very common example.

So I'm not talking about hand flying raw data a VOR approach into London. But any unconstrained airport should be seen as an opportunity to sharpen our skills.

Originally Posted by Two's in

Because killing passengers is bad for business.

On the contrary automation dependency is starting to kill passengers. Turkish at amsterdam, asiana at SFO..

[Vessbot]

Originally Posted by vilas

Absolutely! I​​​​​​ keep saying this all the time. Line flying is not a training flight. Some training is inevitable but should be kept to minimum. We are there not for ourselves but for a purpose.

I disagree with this, and strongly. Line flying is where the real training happens: every day, continuously. It’s where the habits are ingrained, skills developed, and the automatic reactions (mental and physical) set up for success or failure.

Actual training in the sim is just an initial scaffold to hold together a flight from beginning to end while following the SOP and profiles, and learning how to handle a few chosen non-normals, without everything falling apart. And then you do a few flights with a check pilot in the real plane to do your first few with a heightened level of safety backstop.

But imagine yourself having come fresh off training in your first airline operation, compared to say, a few thousand hours later. And think about the difference in how much you’re able to contribute then vs. now, in terms of situational awareness, predicting routine minor things going wrong, how overwhelmed you were, etc. For an even more stark comparison, imagine your then-self in command of the airplane!

What’s the difference between then and now? Of course, all of our cumulative experiences every day on the line. I’m sure you’ll agree that as far as SA, SOP, weather, traffic, ATC, and all that stuff we add about a thousand times more worth to the crew now vs. then, as a consequence of all that gathered experience.


But in what area do we not gather experience? Flying the airplane! If all the flying we ever do after getting signed off is below 1000 feet, configured, trimmed, thrust already set, and already delivered onto the loc and GS, then where do we gather this experience from? We don’t.

——

General reply to the thread:

Why is automation dependency encouraged, and hand flying proficiency paid lip service to? I know that everyone has the concept of proficiency in their brain, but it’s just an abstraction with no easy way to address. Put a few sentence preamble in front of every document to encourage hand flying? Sure. But what does that really do? It’s like knowing it’s bad to eat too much sugar. You know to keep it down overall, but how does that calculate on a particular day when you feel like a snack? You have the snack, on the this-day basis, with no contradiction between the this-day principle and the overall principle, because what you do on this day is a drop in the bucket that can’t possibly affect the overall amount in a meaningful way. But when the this-day urge happens every day, the calculus shifts without you realizing, and the overall principle is ultimately violated.

Just the same, every time on “this day” a reason (many times, actually an excuse) can be found why automation is the appropriate way to go. Every time, until after a while it turns out that an overall amount to maintain hand flying proficiency is never reached. This day, a specific reason - or 10 specific reasons - can be found to address the factors and decide to use the automation. But overall, by the nature of overallness, no specific reason can be called on to say that it’s the day to fly the plane. Always it can be pushed off to some vague next time, but the next time never comes.

To put it another way, it’s a problem with two competing opposite demands. On one hand, the safety, precision, and consistency of automation usage, and on the other hand, the safety of the pilot being proficient and comfortable flying the plane. But the triggers to specifically address those opposte demands aren’t symmetrical. The specificty of all the reasons to all upon automation usage ensure that it’s called upon virtually every time, while the triggers to call upon hand flying practice are too vage - really, non-existent. So, the situation is what it is.

[Check Airman]

Originally Posted by Vessbot

I disagree with this, and strongly. Line flying is where the real training happens: every day, continuously. It’s where the habits are ingrained, skills developed, and the automatic reactions (mental and physical) set up for success or failure.

Actual training in the sim is just an initial scaffold to hold together a flight from beginning to end while following the SOP and profiles, and learning how to handle a few chosen non-normals, without everything falling apart. And then you do a few flights with a check pilot in the real plane to do your first few with a heightened level of safety backstop.

Thank you for putting this more eloquently than I could. Perhaps there's a cultural component to the difference in philosophies. Stateside, a lot of CA's will conclude the initial briefing with "let's keep it safe, and have fun". That's not an invitation to fly under the golden gate bridge, but if i haven't done a localizer approach in a while, we may request one even though there's an ILS available. Whether you'd consider that practise or training, the point is that it hones the skills for the day when you need to fly a localiser approach.

[Check Airman]

Originally Posted by FlyingStone

I'm a big fan of hand flying and raw data (latter prohibited by employer unfortunately), but the reality is that (before Covid) we operate in ever busier airspace and at the end of the day, people in the back or shippers of freight pay us to deliver them or their goods safely and efficiently from A to B, not to have fun or to prove to somebody that we still have "what it takes". Could I handle flying raw data in and out of London TMA during a busy morning period? Yes. Is it a smart thing to do in real life, assuming you've got AFDS serviceable? I hope we all agree what the answer should be here.

I've never flown to or from LHR, but assuming it's similar to busy airports here in the US, why not? There are times when JFK or ORD are busier than others- if you feel the need to use automation at that point- have at it. But I disagree with a blanket statement that we don't hand fly at certain airports.

By way of example, going to LAX from the east can be challenging, with multiple speed and altitude constraints on the approach. It's sometimes easier to just press the red buttons than twisting, pushing, pulling, selecting and managing. On top of that, it's heaps of fun!

[Check Airman]

Originally Posted by FlyingStone

The old 737 EFIS classics had this (optionally, I believe). With the world going towards PBN and curved (RF) procedures, this wouldn't be particularly useful.

I've flown curved legs without a FD. The aircraft had a CDI. It was a complete non-event. As pineteam said earlier, had Boeing and Airbus installed a CDI on the nav display, we could easily do "raw data" RNAV procedures.

[theFirstDave]

I am a (retired) Controls Electrical Engineer (Sr. Controls and Instrumentation). I designed and implement mega-million dollar systems. But I also know how to wire panels, bend conduit and solder breadboards. My job didn't need me to do it, but as I came up through the ranks (starting as an electrician) I kept my skills up. Others never acquired the manual skills as they jumped from college to design.

I am also a Private Pilot. All my flying is manual. I use VOR's, no ATH, no AP, no FD. Yes, at sometimes I am near overload. I won't fly into some places without a second set of hands / eyes with me. The most amount of automation I have is a Garmin GPS that also will warn me of terrane.

So, how to get the ATP guy some experience other than engaging AP at 400' and just twirling dials until you're on short final? Get yourself in a a piston plane (SE or light twin) and haul VFR around the LAX area. That'll get your PIC skills a workout and give you a boat load of situation awareness, energy management experience and more.

At the very least, it'll be fun again.

[KeepItStraight]

Aircraft are designed by engineers and flown by pilots.

I know I'm generalising but with no disrespect to either profession both professions can have quite a different perspective on how an aircraft should be operated or how valuable (or useless and distracting) certain whiz bang equipment can be in differing circumstances.

[FlightDetent]

Originally Posted by Check Airman

I've flown curved legs without a FD. The aircraft had a CDI. It was a complete non-event. As pineteam said earlier, had Boeing and Airbus installed a CDI on the nav display, we could easily do "raw data" RNAV procedures.

Correction: You could easily fly the "raw data" curved RNAV - on a good day. Us, the global pool of average pilots having weak days now and then, could not without adversely affecting some of the statistics.

Either case, such an exercise would be worthwhile to the pilot and unnecessary burden to the flight operations at the same time.

Think of your best talented A-10 pilot. Is he an expert aviator or a Senior Ordonance Delivery Executive? Both in my opinion and one skill set would not suffice if the other was deficient. Understandably the hangar talk is mostly who can best fly the ship tail first. As you go higher up the chain of command and responsibility, the pendulum swings to the other side of relevant. The mission requires both and don't despise the people on the top for caring almost solely about the overall production efficiency.

Is automation dependency encouraged? What a beautiful statement, clear and strong, showing where the line in the sand is. If we (had) reach(ed) the point where perceived automation over-use causes less trouble [(killed customers)^3 / cash / emissions] than the incumbent option - the future of air transport will see the piloting element dissolved.

No, it's not my operational philosophy; but yes it is what I see when reading the landscape of our much-beloved industry. Systems and engineering designs can be improved asymptotically to perfection, whereas reliability and performance level of humans - as a group - have their upper limit, distinctly below of what is achievable with technology. Habsheim and MAX do not enter the debate, those were both rapist acts.

Loved and will long for Skiathos and Santorini, quite happy to stay not qualified for Madeira, or Innsbruck. Mostly for the fuel decisions when approaching the IAF hold.