Why is automation dependency encouraged in modern aviation ?
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,403
Likes: 0
Received 0 Likes
on
0 Posts
Are you reading the conversation ? If you want to correct the thrust for 800fpm to 700fpm, you have to move the thrust levers by a distance as short as an ant's bottom.
It sounds like you're working for Airbus. "The aircraft is perfect as is."
It sounds like you're working for Airbus. "The aircraft is perfect as is."
And no! I don't work for Airbus but since life is not a la carte in real life knowing what your machine does and doesn't and adjusting skills accordingly make your teeth last longer. There are two aspects to discussions, one share knowledge, individual techniques and experiences, the second is idle lament. When the industry is moving towards automated flights with least human presence I think better raw data flying tools is the last thing on their mind. There are more import things required but it's not on cards. B Max needed immediate fix they didn't do it till two crashes. B777 SFO FAA asked them to have look at the throttle hold function they aren't planning to do anything. They also have two different GA procedures one normal and another after touchdown (Emirates Dubai crash) they haven't any plans to change(personally I don't blame Boeing but hold pilots responsible for them). So wake me up when they design better raw data instrument.
Last edited by vilas; 2nd Dec 2020 at 08:25.
Thread Starter
Join Date: Dec 2015
Location: France
Posts: 507
Likes: 0
Received 0 Likes
on
0 Posts
The fact that I omitted the pitch part does not change the length by which you have to move the thrust levers, which is my point : it is too short.
It can be lament, but if you start sending it to the engineers that work across the telephone or email, it becomes customer feedback.
The point of this discussion is to debate the topic and try to find out if it could be beneficial to change a few things here and there.
If you base your answers on the fact that airbus would probably say no to any modification, even if it was asked by several airlines all agreeing on one subject, then yes the discussion becomes useless.
But if you try to ignore this (plausible) possibility, you can have a very interesting discussion like we're all having since the start of this topic.
The customer feedback is to point out the discrepancy between two discourses : there are more and more crashes related to lack of manual flying skill, or reluctance to try and put them into practise, or lack of skill in monitoring the correct behavior of the automation controling the most basic aircraft trajectory... All these could be tackled with higher practise of manual flying, but the aircraft are designed to keep the pilot more and more out of the loop.
As long as automation will sometimes fail, and as long as pilots are needed, this seems counterproductive.
It is also not lament if you propose potential solutions.
For the above mentionned problem, I would suggest adding a thrust figure on the PFD, much closer to the pilot's usual visual circuit. One N1 figure per engine, or one average figure for both engines with abnormal colors to indicate a thrust asymmetry that should be corrected by looking at the EWD.
There are two aspects to discussions, one share knowledge, individual techniques and experiences, the second is idle lament. When the industry is moving towards automated flights with least human presence I think better raw data flying tools is the last thing on their mind. There are more import things required but it's not on cards. B Max needed immediate fix they didn't do it till two crashes. B777 SFO FAA asked them to have look at the throttle hold function they aren't planning to do anything. They also have two different GA procedures one normal and another after touchdown (Emirates Dubai crash) they haven't any plans to change(personally I don't blame Boeing but hold pilots responsible for them). So wake me up when they design better raw data instrument.
The point of this discussion is to debate the topic and try to find out if it could be beneficial to change a few things here and there.
If you base your answers on the fact that airbus would probably say no to any modification, even if it was asked by several airlines all agreeing on one subject, then yes the discussion becomes useless.
But if you try to ignore this (plausible) possibility, you can have a very interesting discussion like we're all having since the start of this topic.
The customer feedback is to point out the discrepancy between two discourses : there are more and more crashes related to lack of manual flying skill, or reluctance to try and put them into practise, or lack of skill in monitoring the correct behavior of the automation controling the most basic aircraft trajectory... All these could be tackled with higher practise of manual flying, but the aircraft are designed to keep the pilot more and more out of the loop.
As long as automation will sometimes fail, and as long as pilots are needed, this seems counterproductive.
It is also not lament if you propose potential solutions.
For the above mentionned problem, I would suggest adding a thrust figure on the PFD, much closer to the pilot's usual visual circuit. One N1 figure per engine, or one average figure for both engines with abnormal colors to indicate a thrust asymmetry that should be corrected by looking at the EWD.
I am adding the phrase "ants bottom" to my lexicon of words meaning small distances. Brilliant ! 
- (the other of course being "a gnats cock").
- (the other of course being "a gnats cock").
Join Date: Sep 2016
Location: USA
Posts: 803
Likes: 0
Received 0 Likes
on
0 Posts
However, wouldn’t it be better technique to make the adjustment at the same time? Why wait until speed decays and then correct it? Why not maintain the correct speed to begin with?
—
To the general discussion about raw data enhancements, yeah there have been some suggestions that are interesting thought experiments (like a cross rate indicator tacked onto the localizer, or zoomed in artificial horizon, etc.) but would probably get laughed out of the design room. If it’s not based on something that existed on steam gauge instruments, it’s probably pie in the sky fantasy. It’s like a steampunk alternate history where EFIS was never invented and mechanical instrument design continued. Never gonna happen in even the wildest possibility of real reality.
But OTOH, there have been some glaring omissions that I consider a near criminal link in the feedback loop of cutting the pilot out of flying because it’s too difficult to fly. Like my plane the CRJ, there are no tick marks on the N1 gauge! I’m not even asking about 10ths, I would settle for quarters or anything to give me a rough at-a-glance reference. So I’ve resorted to using the “N1” label at the bottom left of the right gauge, as the only tick mark. So my initial settings are on it, just a bit under it, or needle pointing down at a 45 degree angle, etc. Like FullWings says, inventing workarounds for designers’ failure to consider that humans might use it some day. Or maybe they thought the numerical readout would suffice, and have never experience the difference between quickly and fluidly setting an analog-style gauge compared to interpreting numbers, when you’re multitasking with 5 simultaneous things... especially when this numerical readout is 45 degrees away from everything else you’re looking at.
Of course it’s no problem when the autopilot’s flying it and all you have to do is set the thrust to work the speed tape...
And yes it should be on the PFD too, that’s a great suggestion (but I’m not holding my breath). It’s in every sense a primary flight control as is the artificial horizon. To anyone inclined to reply that the result (airspeed) is more important than the input (thrust) so should be looked at instead, consider that the same relationship exists between everything else and the attitude, so by the same logic the artificial horizon should be removed also.
Last edited by Vessbot; 2nd Dec 2020 at 16:24.
Join Date: Jun 2007
Location: Wanderlust
Posts: 3,403
Likes: 0
Received 0 Likes
on
0 Posts
However, wouldn’t it be better technique to make the adjustment at the same time? Why wait until speed decays and then correct it? Why not maintain the correct speed to begin with?
Join Date: Jul 2006
Location: Somewhere over the rainbow
Posts: 946
Likes: 0
Received 0 Likes
on
0 Posts
Yeah ideally both at the same time, I agree. Quite difficult to estimate tho.I think instinctively I do it most of the time.Usually if my speed is VAPP and VAPP + 5 I don’t really touch the levers unless I feel like it will go below VAPP. Can’t be bothered too much unless it’s one of those days I try to be on the dot. LOL
Join Date: Sep 2016
Location: USA
Posts: 803
Likes: 0
Received 0 Likes
on
0 Posts
Several posts described the difficulties of flying with raw data, relating this with 'deficiencies' in modern instrument displays. These posts suggest novice-like behaviour and difficulties with situation awareness; projection, predictive ability. No aspersions on the ability of skill of the authors, but an observation relating to the wider industry; aviation operates with a lower level of expertise than previously, yet it is 'safer'.
Improved technology, relyability, etc, reduces the need for higher levels of expertise in both normal and abnormal operations. Pilot training is matched to operational need, generally scenarios which have reduced opportunity to develop expertise.
There are fewer 'surprises' requiring revised awareness and understanding; thus even moderately surprising situations have greater impact because of their rarity and lower expert ability to manage them.
In addition, older generations of crew are part of this evolution, there is reduced expertise in training and mentoring.
Automation dependency is a fact of modern operations.
The belief that more hand flying will improve the level and extent of expertise is not supported by science. Many texts identify the need to experience challenging situations to develop non-normal manual and mental flight skills, to achieve a higher level of expertise.
Developing expertise requires time and opportunity, neither being readily available in modern aviation. The current standard appears good enough when compared with safety data - the industry is safe; it could be safer, but beware false beliefs or generating negative training.
'Accelerated Proficiency and Facilitated Retention' - Accelerated Learning, Expertise, Learning Retention, Skill Acquisition, Skill Proficiency, Training. https://apps.dtic.mil/dtic/tr/fulltext/u2/a536308.pdf
Summary page 8 "It typically takes years of experience for professionals to master their domain. … domain complexity and the need for extended and continuing practice at rare and difficult cases."
Expertise and proficiency page 29 -
Levels of proficiency Table3 2.1 - 2.2 page 31 - 33.
Characteristics of effective scenario-based training Table 4.8 page 148
Improved technology, relyability, etc, reduces the need for higher levels of expertise in both normal and abnormal operations. Pilot training is matched to operational need, generally scenarios which have reduced opportunity to develop expertise.
There are fewer 'surprises' requiring revised awareness and understanding; thus even moderately surprising situations have greater impact because of their rarity and lower expert ability to manage them.
In addition, older generations of crew are part of this evolution, there is reduced expertise in training and mentoring.
Automation dependency is a fact of modern operations.
The belief that more hand flying will improve the level and extent of expertise is not supported by science. Many texts identify the need to experience challenging situations to develop non-normal manual and mental flight skills, to achieve a higher level of expertise.
Developing expertise requires time and opportunity, neither being readily available in modern aviation. The current standard appears good enough when compared with safety data - the industry is safe; it could be safer, but beware false beliefs or generating negative training.
'Accelerated Proficiency and Facilitated Retention' - Accelerated Learning, Expertise, Learning Retention, Skill Acquisition, Skill Proficiency, Training. https://apps.dtic.mil/dtic/tr/fulltext/u2/a536308.pdf
Summary page 8 "It typically takes years of experience for professionals to master their domain. … domain complexity and the need for extended and continuing practice at rare and difficult cases."
Expertise and proficiency page 29 -
Levels of proficiency Table3 2.1 - 2.2 page 31 - 33.
Characteristics of effective scenario-based training Table 4.8 page 148
Originally Posted by PEI 3721
The belief that more hand flying will improve the level and extent of expertise is not supported by science.
Hi PEI 3721,
To take one example of instrumentation; in our modern big jets would we be happy raw data flying an NDB course with the NDB pointer on one small round instrument, the heading bug on another small round instrument; no means of bugging the required NDB course, and having to regularly sync the directional gyro with the E2 compass?
No. Yet, those of us of a certain vintage did it at flight school - within 5°, whilst also flying a non stable, non AP, non A/THR light aircraft.
So now that we can overlay the NDB required course bar and the raw data NDB needle onto the (automatically synced), ND compass display making NDB tracking 10 times easier and 10 times more accurate; does that make us bad pilots?
When cars were first developed the car radiator temperature gauge was on the top of the radiator, out on the front of the bonnet. Modern cars have the temperature gauge next to the speedo in front of the driver - much easier to see and monitor. Does that instrument improvement make us worse drivers?
Bad pilots are those who 'blindly' follow a FD and don't look through it at the raw data and the pitch, bank, speed and V/S readouts.
To take one example of instrumentation; in our modern big jets would we be happy raw data flying an NDB course with the NDB pointer on one small round instrument, the heading bug on another small round instrument; no means of bugging the required NDB course, and having to regularly sync the directional gyro with the E2 compass?
No. Yet, those of us of a certain vintage did it at flight school - within 5°, whilst also flying a non stable, non AP, non A/THR light aircraft.
So now that we can overlay the NDB required course bar and the raw data NDB needle onto the (automatically synced), ND compass display making NDB tracking 10 times easier and 10 times more accurate; does that make us bad pilots?
When cars were first developed the car radiator temperature gauge was on the top of the radiator, out on the front of the bonnet. Modern cars have the temperature gauge next to the speedo in front of the driver - much easier to see and monitor. Does that instrument improvement make us worse drivers?
Bad pilots are those who 'blindly' follow a FD and don't look through it at the raw data and the pitch, bank, speed and V/S readouts.
Last edited by Uplinker; 4th Dec 2020 at 11:06.
Thread Starter
Join Date: Dec 2015
Location: France
Posts: 507
Likes: 0
Received 0 Likes
on
0 Posts
Several posts described the difficulties of flying with raw data, relating this with 'deficiencies' in modern instrument displays. These posts suggest novice-like behaviour and difficulties with situation awareness; projection, predictive ability. No aspersions on the ability of skill of the authors, but an observation relating to the wider industry; aviation operates with a lower level of expertise than previously, yet it is 'safer'.
Just because someone suggests a possible improvement about ergonomics, that says nothing about their skills.
Airline pilots should be competent, have all the skills that you list, yes.
Does that prevent an aircraft manufacturer from making their products easier to use ?
With your reasoning, we could very well accept that the aircraft manufacturer makes a very complicated aircraft and reject the blame on the pilots for not being competent enough.
For example, we could accept the overhead panel to be designed for minimum cost rather than for presenting a coherent view to the pilot. We would get an overhead with buttons put in a seemingly random or counterintuitive fashion. Would you blame a pilot for launching the wrong fire agent, if agent 1 from engine 1 was next to engine 2 fire button ? Or would you blame the manufacturer for an absurd design ?
You could also imagine reverted controls, all kinds of absurdity, you would still blame the pilots protesting about this as "not skillful enough" ?
It is absurd.
Aircraft manufacturers and pilots work hand in hand for safety.
Pilots have to be the most competent that they can and Airbus has to deliver an aircraft that is as easy to use as possible.
So yes, I still don't understand airbus' position. If manual flying is important, as they say, to maintain flying skills, why don't they give the pilot a crosstrack deviation, allowing them to practise RNAV in raw data ?
Join Date: Jul 2006
Location: Somewhere over the rainbow
Posts: 946
Likes: 0
Received 0 Likes
on
0 Posts
Raw data departure using Nav or Arc mode is easier than raw data approach so as long as you can fly raw data approach you should be just fine for a raw data departure. Of course a raw data departure using solely Rose VOR would be challenging if not used to do so. I agree it would be nice to have cross track deviation but why not like I said in an earlier post fly a conventional departure instead? All the airports I'm flying have conventional SIDs. You can always request one. I do that sometimes in my own base.
Last edited by pineteam; 4th Dec 2020 at 12:47. Reason: added one sentence
Join Date: Jul 2006
Location: Somewhere over the rainbow
Posts: 946
Likes: 0
Received 0 Likes
on
0 Posts
Thread Starter
Join Date: Dec 2015
Location: France
Posts: 507
Likes: 0
Received 0 Likes
on
0 Posts
Raw data departure using Nav or Arc mode is easier than raw data approach so as long as you can fly raw data approach you should be just fine for a raw data departure. Of course a raw data departure using solely Rose VOR would be challenging if not used to do so. I agree it would be nice to have cross track deviation but why not like I said in an earlier post fly a conventional departure instead? All the airports I'm flying have conventional SIDs. You can always request one. I do that sometimes in my own base.
Rose VOR is completely manageable if anticipated correctly, it was considered basic during my MCC training.
It is quite fun to consider that the first time I flew the A320 sim, it was all raw data flying and navigating. It didn't seem to be a huge problem. But it is never done like this on the line.