Does more automation necessarily equate to better piloting skills?

[Centaurus]

The Wall Street Journal reports "US aviation regulators want cockpit automation fixes on nearly 500 Boeing 737 planes, seeking to prevent pilot error that over the years have caused fatal crashes of several jet and turboprop airliner. Foreign regulators are expected to follow the FAA's lead in demanding changes to parts of the aircraft's flight control computers, likely affecting an additional 700 or more of the widely used 737's. Chicago based Boeing previously recommended the changes but the FAA wants to make them mandatory."

So yet more automation to help incompetent pilots from flying too slowly:=. It certainly illustrates the truth of the statement attributed to a Boeing 787 test pilot, that Boeing designed the 787 on the basis it will be flown by incompetent pilots, which is why the automation in that aircraft is so sophisticated.

It would be more cost effective to all concerned that, instead of yet more automation, the FAA mandated fifty percent of all simulator recurrent and type rating training must be conducted using raw data hand flying. This would instantly obviate the need for still more expensive and unnecessary automation. :ok:

[1jz]

That's true, what happens when the worst takes place and automation fails to work. Then it would be a pair of goofs on the controls with no hands on experience.

[Cool Guys]

More Automation?

The thing with automation on a plane is you can’t just give up if the automation fails. You have to be able to fly the thing manually, with no automation. Therefore your learning curriculum doubles. You have to know 2 methods of flying. Adding more automation doesn’t negate the need to fly manually, so it adds more to the length of the study and more potential holes in the cheese. Less automation or simplifying the present automation seems like a better path.

[Wageslave]

Making the magenta line longer and thicker certainly won't improve piloting skills.

[keith williams]

I don't think that very many people would argue that more automation improves pilot skills.

But more automation has, and will continue to reduce accident rates.

Those two objectives are very different.

As long as the benefits of the reduced accident rates outweigh the costs of degraded pilot skills, increased automation will be the best way to proceed.

[+TSRA]

Those two objectives are very different.

As long as the benefits of the reduced accident rates outweigh the costs of degraded pilot skills, increased automation will be the best way to proceed.

Apologies if I'm hearing you wrong, but are you trying to suggest that manually flying an aircraft and flying an aircraft on automation are two different objectives? I would have thought the objective was rather obvious: to use both methods to achieve safe flight operations.

I would argue that, while automation has reduced the accident rate, the accidents we are seeing can be directly attributed to a lack of basic flying skills and a misunderstanding of what the aircraft is telling the pilots.

All my good instructors have always said "listen to the airplane; it's trying to talk to you." I've continued this same theme when I give training. There is no amount of computer technology available that can give me the same information has handling the aircraft (FBW aircraft aside).

I'm a big proponent of the lessons taught at American Airlines in the Children of Magenta video, and I believe that is what we should be teaching our pilots: if you're not sure what's going on, fly the damned thing yourself!

[PEI_3721]

Searches for more detail suggest that FAA action is only a rumoured proposal.
A particular aspect appears to be the 737 autothrottle where loss of speed control / inappropriate mode switching can result from a single rad alt failure; cf AMS accident.
As such, this has little or nothing to do with manual flying etc, and relates more to awareness and expectation of system behaviour. Similar aircraft, if not all recently certificated types use either self-monitoring rad alts or system crosschecks which adequately warn or disengage the autothrottle in critical flight stages, e.g. final approach. Thus crews might have a reasonably expectation of a similar standard for all aircraft, particularly if qualified on other types.
The 737 certification appears to be based on grandfather rights from previous models, which for automation / integrity design are now well out of date, and might not meet current cert requirements, e.g. 25.1302 / 1309.

A further rumour considers the 777 autothrust ARM loophole, but this is unlikely to become fact before the Asiana accident report.

[keith williams]

Apologies if I'm hearing you wrong, but are you trying to suggest that manually flying an aircraft and flying an aircraft on automation are two different objectives?

No.

What I am saying is that (1) Improving pilot skills and (2) Reducing accident rates are two different objectives. Improving pilot skills should reduce accident rates, but that does not mean that the two are the same.

The more important of the two is reducing accident rates. If we could have zero accidents with zero pilots then that is the way that aviation would go.

[AirRabbit]

The thing with automation on a plane is you can’t just give up if the automation fails. You have to be able to fly the thing manually, with no automation. Therefore your learning curriculum doubles. You have to know 2 methods of flying. Adding more automation doesn’t negate the need to fly manually, so it adds more to the length of the study and more potential holes in the cheese. Less automation or simplifying the present automation seems like a better path.
Well said, my friend. And while it isn’t often mentioned out loud, the primary reason that manual skills seem to suffer (and they DO suffer) is because most organizations are extremely reluctant to bear the expense that these longer training times generate, and logically the greatest expense is most likely generated by the increased time spent in the simulator. Additionally, I think that there are 2 factors that are most prominent in such requirements: first, airlines are seeking the authority to operate into destinations under increasingly arduous circumstances; and second, the background (both operational and training) as well as the practical experience of the pilot work-force today, in general, is transitioning to a makeup of lesser practically experienced applicants. This has a direct result of training pilots, who, by virtue of their more limited experience, are, out of necessity, more and more dependent on what they learn in their initial training exposure when starting to operate for an airline. That translates directly into a required increase in training time.

It’s my opinion that, in at least most cases, this increased training time should be included as a requirement levied by the regulator. However, to my chagrin, the regulator, at least in some cases, does not want to publicize their “requiring” organizations to spend more money than those organizations have already allocated to conduct the amount of training presently conducted under existing regulatory requirements. Further, some regulatory authorities have authorized a complete replacement of the existing regulatory requirements, relying on what I believe is a wholly misplaced and unreal confidence in how training programs are developed and administered. First, the pilots newly hired today are not like the pilots newly hired 30 years ago – and that isn’t meant as a derogatory comment toward today’s new pilots – just an acknowledgement of the differences between what was “typical” training and experience obtained by pilots 30 years ago and the training and experience presented by pilot applicants today. When these issues are combined with the dramatic increase in the automation found in ever-increasing amounts in the airliner flight decks today, it is clearly necessary to ensure that pilots receive the appropriate training required to not only operate but understand all of the equipment in today’s aircraft, including not only how to use it, but when and when not it should be used … as well as obtaining the experience today’s equipment operation clearly demands.

AND – anyone here can call me barbaric or insensitive if appropriate, but I just don’t think that individual airline management personnel, individual pilot organizations, or individual federal regulatory authorities, on their OWN, can determine what is appropriate. I believe that a mutually honest and open exchange of information and a determination of levels of training and how such training can and should be determined to have been accomplished is necessary. Additionally, I think there are efforts currently underway to do just that – but those efforts have not received anywhere near enough publicity or participation. The International Airline Pilots Association, the International Civil Aeronautics Organization, and the Royal Aeronautical Society in the United Kingdom, have collaborated to look into just exactly these issues – but I am not convinced that sufficient participation by ALL of the necessary organizations is being accomplished. I would urge any and all participants here to find out how they can participate in these efforts and then commit the time and effort to do so.


I don't think that very many people would argue that more automation improves pilot skills.
But more automation has, and will continue to reduce accident rates.
Those two objectives are very different.
As long as the benefits of the reduced accident rates outweigh the costs of degraded pilot skills, increased automation will be the best way to proceed.

It may be that some automation contributes to a decrease in accident rates … but I believe that is for a limited set of circumstances and has, up to now, been limited to operations into very specific and very limited environmental situations … and I don’t mean just weather … I mean a combination of weather and physical circumstances. I believe that an improper or ill-advised reliance on automation may wind up increasing accident prone circumstances to develop – which investigations may or may not be able to identify. And I continue to worry that instructors today are much too prone to let the simulator do the teaching while they, themselves, spend their time programming the simulator to do what he/she wants it to do. This is another aspect of what those organizations I mentioned above are addressing and yet another reason to urge participants here to seek out and become involved in those activities.

[glendalegoon]

NO, it does not.

[Centaurus]

the primary reason that manual skills seem to suffer (and they DO suffer) is because most organizations are extremely reluctant to bear the expense that these longer training times generate, and logically the greatest expense is most likely generated by the increased time spent in the simulator

AirRabbitt - that was very well put. Your views tie in nicely with those of a friend of mine who, until recently, worked for Boeing in USA. For many years I have been fortunate enough to have had pleasant and insightful email conversations with him, a senior Boeing simulator instructor based in Seattle. He has listened with commendable patience to my ranting on the need for the manufacturer to initiate and encourage pilots to keep in regular practice at manual flying in both the simulator and the real thing.

While he agreed in general with my thoughts, he said the current situation will not change until individual regulators have the technical knowledge and the will to enforce changes to the airlines under their watch. He doubted that would happen. And from what I have seen, he has proved right; certainly in my part of the aviation world.

[Jwscud]

It's also in my company's ops manuals, with some restrictions about weather, airspace and so on.

[BraceBrace]

It is in many airlines ops manuals, and while I agree it doubles the required training, it is (mainly) not the companies fault that manual flying skills are degrading.

When you are doing an official check, you do as you are supposed to and use automation to the maximum extent possible. Airmanship is what counts and I have nothing against this because it WILL make your life more easy.

However, simulator training prior to those checks have always been open to both sides (manual or automated flying). Instructors have always been very "supportive" in my company. There was always the possibility to try, and with some spare time available, there was always the question "would you like to do something extra?". Usually, these questions are quickly answered with a "NO".

It's actually the way of thinking of many of my crew members. The (understandeable) request of airlines to adhere to very basic rules like "stable" approaches, with obligatory visits to the office or even simulator if you screwed up those basis requests, seems to have changed the behaviour of many pilots. All of a sudden, pilots looked at magenta FMC lines like gods hand that has to be followed. Then it even got worse because it's a vicious circle going down. Autopilots and autothrottles have to stay on as long as possible in VMC conditions under all beautifull sunshiny days. Anything else, and they will very clearly tell you they "don't feel comfortable" or even consider it "unsafe". We, pilots, don't feel comfortable enough doing it "raw". The result is you are blocked by your own collegues from doing anything else but follow automation, not by the company.

Then you go to the simulator, where the instructor gives you programs that deal with FMC failures or something similar. The result is not really worth publishing. The debrief will be something like "yeah, but that was a hard program with lots of stuff to do." Yet in theory, there was not a lot to do, you only had to revert to very basic modes. Yet you never used them anymore because of your magenta addiction, and end up feeling even more uncomfortable than ever before.

[grounded27]

This is no different than any change to an FCC, a good example is the delayed ground spoiler response on the MD-11 to lessen the tendency for the aircraft to pitch up upon ground spoiler deployment.

If you wish to ask yourself why automate as opposed to train, why do we have automation in the first place? Pilots were first told it was to lessen their workload thus increasing safety... The truth is a shift in liability from the pilot & airline to the manufacturer is a large selling point.

The pilots screwed themselves, many pilot union contracts out there limit the duration of a flight with an autopilot inoperative. Automation is the future, can't stop it. We have created for ourselves an environment today where I would prefer an aircraft landing it's self over a 22 year old kid that is paying to sit in a seat that gives him the illusion that he is an aviator. Or the 55 year old captain, so freaking bored with his job that his complacency is certainly less safe than an autopilot!

[AirRabbit]

…a senior Boeing simulator instructor based in Seattle … has listened with commendable patience to my ranting on the need for the manufacturer to initiate and encourage pilots to keep in regular practice at manual flying in both the simulator and the real thing.

While he agreed in general with my thoughts, he said the current situation will not change until individual regulators have the technical knowledge and the will to enforce changes to the airlines under their watch. He doubted that would happen. And from what I have seen, he has proved right; certainly in my part of the aviation world.

Thanks for the kind words, sir … but I am of the opinion that the regulators have, and have had – for some time now – the necessary “technical knowledge” to make the kinds of changes that we’re discussing. In my opinion (deeply held and verified in more instances than I’d care to try to number), the individuals who happen to be employed by Regulatory Authorities, are human (beleaguered with the same frailties that beleaguer other humans) and, therefore, are more interested in or focused on their own successes and failures that might or will aid or prevent their own advancement through the ranks to reach a point that will allow a comfortable retirement; and, of course, while seeking that financial goal, they would much prefer the acknowledgement of their peers, their subordinates, and their superiors on a personal satisfaction level to the degree that they would do almost anything that they believe, or hope, would maximize those potentials … and they would do nothing that they think might negate any of that official or unofficial personal praise - thereby ensuring that the result is greatly weighted in favor of one's own advancement, even if that means a degradation in the originally so-called "meaningful" circumstance. Again, from my perspective, just about the ONLY way to get past those personal hopes and fears is to have such regulatory changes considered, determined, and recommended by a consortium of professionally competent personnel made up of the regulators and the regulated … to provide the best possible opportunity for whatever it is that is ultimately recommended to actually be a success … such that all who participated can bask in the limelight of that success, and, just as importantly, that any failure or “dud” (i.e., mediocre result – because surely there will be some of those as well) can be marked-up against “…those who wanted to do the wrong thing” … whatever that might have been. So, when those kinds of training consortiums are developed, I think all of us owe it to ourselves and to those who follow, to do all we each can do to become involved and vigorously participate in such activities so that we have the very best opportunity to develop and recommend good, logical, and successful advancements in practices, procedures, and equipment, all of which increase the safety and reduce the incidents/accidents to (hopefully) long-ago-forgotten remembrances. If we fail to do these things the best we can hope for is that those who will, through their “positional authority” on our industry will, somehow, determine that those good, logical, and successful advancements are coincident with their own perception of their own successes. Personally, I’m not optimistic about those kinds of coincidences.