Announcement

Collapse
No announcement yet.

So glad I'm retired!

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    [QUOTE=Evan;n1121090]True. But what aircraft does that?[/[quote]
    None, but mine was not an A vs B rant or anything oike that. Not comparing airplanes, just analyzing how automation / technology played a role.

    As I recall, the thrust lock was recognized and removed before the stall. Anyway, the pitch maneuver is what caused the stall.
    As I recall, it wasn't. The first time that the TL were touched was at the TOC when the stall warning came out and they responded by firewalling and pulling up. As I recall... (I can be wrong)

    True, but what aircraft turns off FD's in this situation? Anyway, that's not automation.
    Again, not comparing airplanes.
    How is FD not automation? It absolutely is. It is not autopilot but it is automation. It converts the bias of the expected performance vs the actual performance in attitude command cues to be followed by the pilot (human or automatic). It is much easier to keep the FD crossbar centered than to keep the crossbar of the ILS centered, for example, because the FS is already telling you the pitch and bank needed to capture and track the GS and LOC. Without the FD, a pilot needs to look at the deviation of the needle and the trend, estimate what correction is needed in heading or vertical speed, adjust the attitude for that heading and vertical speed, and constantly change those targets as you approach the center else you overshoot. The same if you want to keep an altitude and ground track. All that process is automated with the FD, you just keep the attitude where the FD tells you.

    (protection / stall warning) Certainly, but that is not automation.
    Well, it is loss the of automatic envelope protection.

    (alternative independent airspeed source)Ditto.
    Ok, not automation, but technology that will need to be taken into account if you want single pilot ops at cruise.

    I have always agreed with that. All airliners should have automated response to UAS to remain in the speed envelope until the crew can assess and take over. But none do.
    None do yet, but I think this is a must for single pilot ops at cruise.

    I think / hope that Maximus will make the link between AF447, technology/automation, and single pilot ops.

    --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
    --- Defend what you say with arguments, not by imposing your credentials ---

    Comment


    • #17
      Originally posted by Gabriel View Post
      None, but mine was not an A vs B rant or anything oike that. Not comparing airplanes, just analyzing how automation / technology played a role.
      But that's not automation playing a role. That's automation not playing a role when, in theory, it could be playing a role.

      As I recall, it wasn't. The first time that the TL were touched was at the TOC when the stall warning came out and they responded by firewalling and pulling up. As I recall... (I can be wrong)
      It was recognized on ECAM. The PNF first mentions it 12-15 secs after losing the automation. It seems to finally get taken care of (TL's moved into CLB) about 10-12 secs before the stall warning sounds.

      How is FD not automation?
      FD is guidance, not automation.

      Well, it is loss the of automatic envelope protection.
      Yes, well, everything here is about the LOSS of automation. If Maximus wants to talk about how the LOSS of automation brought down AF447, that would be more accurate, except that it would be better stated as the crew's inability to fly the plane without the automation.

      I think / hope that Maximus will make the link between AF447, technology/automation, and single pilot ops.
      AF447 shows that additional pilots provided no advantage and did not change the outcome. It's quality, not quantity, that was needed there. You, Gabriel, alone in that cockpit, would have landed it on a runway in France.

      It does show, however, that multiple pilots and CRM are needed to systematically work complex problems involving very complex systems. One pilot has to concentrate on flying and task share with another pilot concentrating on ECAM and procedure. That's why I can envision a single pilot in the cockpit if another is nearby (crew rest, front cabin) and able to quickly jump in, but I can't envision a plane with only a single pilot onboard.

      Comment


      • #18
        I am trying to reply to Evan but getting this stupid error.

        Error while saving content: %1$s

        Error information: " 500 error "
        As you can see it is not that the system just doesn't let me post.
        I remember there were some words that for some reason didn't let you post. I may have one of those in my reply.
        Anybody remember what they were?

        --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
        --- Defend what you say with arguments, not by imposing your credentials ---

        Comment


        • #19
          so, i've probably opined this before, but i'll do it again. "automation" as you folks like to call it is absolutely wonderful. until it isn't. automation works so well (when it's working) because it is doing 1000's of calculations and measurements and potentially adjustments per SECOND. far better than any human can. that's not to say it can "fly" better, but it certainly can compute better and react faster.

          to achieve this level of fabulousness, the complexity of the systemS are enormously complex, and are entirely dependent on multiple functioning sensors, which as those begin to fail for one reason or another, cause the automation to shut itself off, often at inopportune times. our trusty fail-safe devices are two people we trust have been (in theory) trained to FULLY understand these enormously complex systems, how they work, when and why they fail, and oh yeah, how to fly the bloody plane without said fabulous automation.

          i'm convinced that very few pilots really understand these deeply complex systems and how they interact. rather, they are taught the various flight "laws" and that as the system degrades so does the automation. and as much as we armchair, monday morning quarterbacks can sit here and say how obviously wrong a bunch of highly trained pilots were for relentlessly pulling up while there plane sank at an incredible rate into the atlantic, they didn'f fail (and die) SOLELY because they pulled up relentlessly. they failed because they really did not understand the enormously complex set of systems in the aircraft that had failed to a degree. i dont need to say yet again why they died and took 100's with them.

          maybe, just maybe, the designers are partially to blame for building a system that can fail. what i mean is, if we can sit here and say "known pitch and power, yada, yada, yada, why can't the automation be designed to, in let's say a UAS scenario, to fly known pitch and power??? duh! if the pilots are expected to do so, why not demand the same of the multi-billion dollar automation? long ago i called this the "easy switch."

          i'd like to add that i have a ton of respect for pilots, even if they screw up and kill lots of people every once in a while. i've sat in the cockpit of several modern planes during flight and although i'm pretty smart, i'm not sure i could do what they do. maybe. but maybe not.

          ok, rant over

          Comment


          • #20
            Originally posted by TeeVee View Post
            so, i've probably opined this before, but i'll do it again. "automation" as you folks like to call it is absolutely wonderful. until it isn't. automation works so well (when it's working) because it is doing 1000's of calculations and measurements and potentially adjustments per SECOND. far better than any human can. that's not to say it can "fly" better, but it certainly can compute better and react faster.

            to achieve this level of fabulousness, the complexity of the systemS are enormously complex, and are entirely dependent on multiple functioning sensors, which as those begin to fail for one reason or another, cause the automation to shut itself off, often at inopportune times. our trusty fail-safe devices are two people we trust have been (in theory) trained to FULLY understand these enormously complex systems, how they work, when and why they fail, and oh yeah, how to fly the bloody plane without said fabulous automation.

            i'm convinced that very few pilots really understand these deeply complex systems and how they interact. rather, they are taught the various flight "laws" and that as the system degrades so does the automation. and as much as we armchair, monday morning quarterbacks can sit here and say how obviously wrong a bunch of highly trained pilots were for relentlessly pulling up while there plane sank at an incredible rate into the atlantic, they didn'f fail (and die) SOLELY because they pulled up relentlessly. they failed because they really did not understand the enormously complex set of systems in the aircraft that had failed to a degree. i dont need to say yet again why they died and took 100's with them.

            maybe, just maybe, the designers are partially to blame for building a system that can fail. what i mean is, if we can sit here and say "known pitch and power, yada, yada, yada, why can't the automation be designed to, in let's say a UAS scenario, to fly known pitch and power??? duh! if the pilots are expected to do so, why not demand the same of the multi-billion dollar automation? long ago i called this the "easy switch."

            i'd like to add that i have a ton of respect for pilots, even if they screw up and kill lots of people every once in a while. i've sat in the cockpit of several modern planes during flight and although i'm pretty smart, i'm not sure i could do what they do. maybe. but maybe not.

            ok, rant over
            I agree, that is the CURRENT state of the situation, but it doesn't need to be.

            The design philosophy changes radially when your first design input is "no pilots". Then you are not allowed to use pilots as the redundancy for system fails. It can be done,
            And let's face it, human pilots don't have a great safety record when the fit hits the sham and the automation suddenly says "that's it, your plane" and handles a crippled plane to the human.

            As I was trying to say in the post that I was not allowed to post, yes, there will be accidents that could have been avoided by a competent human. But there will be orders of magnitudes more accidents avoided by removing the imperfect human from the equation. Computers don't get fatigued, don't get task-saturated, on-t get tunnel vision, don't panic, don't suffer from startle, don't forget procedures, don't get getthereitis, are immune to spatial disorientation or loss of situational awareness, don't get distracted, don't commit suicide or murdercide, don't intentionally fly into skyscrapers, don't decide to violate rules or procedures, and don't relentlessly pull up when the stall warning goes out. There, you have just removed 99% of the accidents. Add a 5% of new "original" accidents and it is still a huge win.

            --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
            --- Defend what you say with arguments, not by imposing your credentials ---

            Comment


            • #21
              Originally posted by Gabriel View Post
              As I was trying to say in the post that I was not allowed to post, yes, there will be accidents that could have been avoided by a competent human. But there will be orders of magnitudes more accidents avoided by removing the imperfect human from the equation. Computers don't get fatigued, don't get task-saturated, on-t get tunnel vision, don't panic, don't suffer from startle, don't forget procedures, don't get getthereitis, are immune to spatial disorientation or loss of situational awareness, don't get distracted, don't commit suicide or murdercide, don't intentionally fly into skyscrapers, don't decide to violate rules or procedures, and don't relentlessly pull up when the stall warning goes out. There, you have just removed 99% of the accidents. Add a 5% of new "original" accidents and it is still a huge win.
              There is a ‘legacy’ mentality holding back progress. Legacy desktop computers were notorious for crashing or blipping out. Embedded computers such as those found on aircraft have been far more reliable but still fail on occasion. But a system involving three computers, each with dual channel, self-monitoring capability using solid state memory and being routinely checked, serviced and upgraded simply does not get into a state where it can no longer safely conduct the flight. Unless...

              Unless it is blinded. Sensors are the weak point and sensors of the same design exposed to the same environment do not create redundancy. So, until there are ways to overcome this issue (I think there could be), we cannot have fully-automated commercial aircraft.

              What we we need is a new mentality reflecting the modern realities. Flight should be conducted through closely monitored automation and automation should be robust enough to handle both normal and abnormal/ emergency ops. Manual piloting should be the very last resort—the equivalent of mechanical backup today, something that will most likely never occur.

              The conundrum is: how to keep pilots skilled at flying if they never fly? There are ways, in high-fidelity sims and recurrent non-revenue flight training. This is how many emergency responders train for things they have never had to do but might be required to do one day. I think it is entirely possible, although your worst-case scenario emergency landing might not be a greaser down on the numbers. Anything you walk away from should be quite possible.

              Of course, the other conundrum: why would anybody want this job? is what Bobby is pointing out. It would certainly appeal to a different mindset, but the view from the office would still be quite fantastic

              Comment


              • #22
                Originally posted by TeeVee View Post
                ...i'm convinced that very few pilots really understand these deeply complex systems and how they interact. rather, they are taught the various flight "laws"...
                Enjoyed the rant.

                Regarding this snip- I think it's impossible for pilots to fully understand the systems...it's too much. Thus we teach more generalized laws. Word choice is brutal here- there are no good words. I recall military pilots having to recite the thickness of brakes and discs, and from too many years of being on these fora, get glimpses of the very complex things pilots DO know- even though I'm sure they don't know everything. And for "generalized laws", I'm sure there's ATP's here ready to run a 787 up my backside.

                My head does explode when pilots do things that go against what (seems to me) is uber uber uber basic things (the dead-horse pull-up). Or forget and think that logical systems work in illogical ways. (Is this overload?)

                Design wise, I also pontificate (as an outsider) that snippits of Airbus this and Airbus that and "normal law", and red, green, blue and yellow, flaps 1... The deal with all of this is that when everything craps out, it's an airplane. But Airbus tries too hard NOT_to be an airplane, where as Boeing (with it's huge load of automatic stuff, sort of acts like and airplane). With the Boeing philosophy, you are at least experienced with operating an airplane, versus operating a drone that suddenly says, computer broken, your aeroplanie.

                Now, all of the Airbus bashing aside, a particular Airbus captain I "know" has claimed that his rule of thumb is if things start going bad, turn it into a Cessna and fly it. (So, outsiders claiming that Airbus doesn't operate like an aeroplanie, is all wet.)

                What should we do? Well, I think it's good that pilots generally like to hand fly a fair bit (although, clearly there's those that don't). No problem with computers monitoring stuff...just like you say, they are fast at math and logic. The interface? I mean that very broadly. A key moment on this fora was Bobby dissing TPMS as another distraction at a critical time that will someday CAUSE a crash...but but but, it's trying to help you... I know that engineers work really really really hard designing stuff to "make sense" and work well "even at the worst possible time"...and what do we get: MCAS...(rant all day long, but I'm sure many people thought they did the right thing...and other people were good parrots, too).

                /rant-induced rant (and yes, I went all over the place)
                Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                Comment


                • #23
                  The Maximus video is out, and it is quite disappointing. It does focus in the differences between the spring-loaded, no feedback, not-interconnected sidestick vs the force-feedback, conspicuous, interconnected yokes, and how that could have made the difference for the other pilot to realize what the flying pilot was doing, which I think is a fair statement. But other than that, it has simplifications, missing pieces of information and just incorrect information.

                  https://www.youtube.com/watch?v=3hZ_CKYiKv8

                  --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                  --- Defend what you say with arguments, not by imposing your credentials ---

                  Comment


                  • #24
                    Originally posted by Gabriel View Post
                    The Maximus video is out, and it is quite disappointing. It does focus in the differences between the spring-loaded, no feedback, not-interconnected sidestick vs the force-feedback, conspicuous, interconnected yokes, and how that could have made the difference for the other pilot to realize what the flying pilot was doing, which I think is a fair statement.
                    Is it? If the aircraft is in a nose-high attitude, which the PNF can see clearly by the PFD, is it really necessary to feel the other pilot's commands? Obviously, they are nose-up commands. The thrust setting is visible to both pilots. How can the monitoring pilot not recognize what the flying pilot is doing at that point? And, in the event of dual input, is the problem a design problem or a CRM breakdown/absence problem? If it's not 'your airplane' you don't make control inputs, period. I call bullshit. This was a human error and training deficiency accident and here comes the little dot.

                    Comment


                    • #25
                      i tried watching the video and gave up. this guy is soooooooo boring i dont know how anyone can sit through anything he narrates. not to mention the fact that he's simply reading other articles word for word which i find more than annoying.

                      Comment


                      • #26
                        Originally posted by Evan View Post
                        Is it?
                        Yes, it is. The assertion that "differences between the spring-loaded, no feedback, not-interconnected sidestick vs the force-feedback, conspicuous, interconnected yokes could have made the difference for the other pilot to realize what the flying pilot was doing" is a fair statement.

                        If the aircraft is in a nose-high attitude, which the PNF can see clearly by the PFD, is it really necessary to feel the other pilot's commands? Obviously, they are nose-up commands. The thrust setting is visible to both pilots. How can the monitoring pilot not recognize what the flying pilot is doing at that point? And, in the event of dual input, is the problem a design problem or a CRM breakdown/absence problem? If it's not 'your airplane' you don't make control inputs, period. I call bullshit. This was a human error and training deficiency accident and here comes the little dot.
                        3WE is right, you ARE black-and-white, uh?

                        Is it really necessary to feel the other pilot's commands? No. Not necessary. But might it have helped?
                        How can the monitoring pilot not recognize what the flying pilot is doing at that point? Well, ask him. Oh, I forgot, he is dead, perhaps because he did not recognize it.
                        And, in the event of dual input, is the problem a design problem or a CRM breakdown/absence problem? It is a CRM problem. That doesn't mean that some other thing can't help to address it. The question is really not what caused it but what could have helped.
                        If it's not 'your airplane' you don't make control inputs, period. Except when the other guy is doing crap. If that happen, AND IF YOU RECOGNIZE IT, yes a "my controls" call is in order, but if the other guy doesn't let go you may need to fight for it.
                        This was a human error and training deficiency accident and here comes the little dot. Little dot my arse. Human error, forget about it. The steep improvement in commercial aviation came when we stopped blaming humans for erring, and instead starting focusing in making the systems error-resistant or error-resilient (the first one makes the error difficult to occur in the first place, the second one makes the error not critical). Training? For sure, the training must have been wrong since the Cessna 172.... OR.... the guy panicked and acted totally irrationally and contrary to his training. That things can happen and it is one of the most complex ones. You never know how a human will react under extreme stress and scare, no matter how good and extensive the training. It is true that repeated exposure in training increases the chances of a controlled, rational and correct response, but there is no guarantee.

                        Other than all of the above, we already discussed several design improvements that could have increases the probability of saving this plane. Does it mean that the lack of such improvements caused the accident? No, in the same way that the lack of GPWS never caused a CFIT accident, but the presence of GPWS avoided many.

                        Bottom line, the lack of a repeated and at times sustained squeeze of the non-flying pilot's balls by a control column certainly didn't cause the accident, but the presence of such could very well have avoided it.

                        --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                        --- Defend what you say with arguments, not by imposing your credentials ---

                        Comment


                        • #27
                          Originally posted by Gabriel View Post
                          Bottom line, the lack of a repeated and at times sustained squeeze of the non-flying pilot's balls by a control column certainly didn't cause the accident, but the presence of such could very well have avoided it.
                          Agreed but that's not the bottom line here. Absolutely yes, anything that provides greater SA could have helped. Is that his claim? No.

                          He entices us from the get go with a statement about how the automation was at fault, thus inplying that the design is somehow unsafe. Typical populist nonsense.

                          If Airbus ditched their FBW and moved back to control columns it would make the design safer for certain ultra-rare scenarios brought on by the failure of CRM and cockpit gospel rules. It would also make it less safe in many other ways. It Boeing ditched their control columns for sidesticks and hard envelope protections, the same is true in reverse. What's the point? Neither are unsafe! Neither have ever caused a crash.

                          Maximus is just trolling the know-nothing population with shock 'journalism' in a mad grasp for attention and the revenue it brings him. Why are we even discussing this tripe?

                          Comment


                          • #28
                            Originally posted by Evan View Post
                            Maximus is just trolling the know-nothing population with shock 'journalism' in a mad grasp for attention and the revenue it brings him. Why are we even discussing this tripe?
                            i agree 100%

                            Comment


                            • #29
                              Originally posted by Evan View Post

                              Agreed but that's not the bottom line here. Absolutely yes, anything that provides greater SA could have helped. Is that his claim? No.

                              He entices us from the get go with a statement about how the automation was at fault, thus inplying that the design is somehow unsafe. Typical populist nonsense.

                              If Airbus ditched their FBW and moved back to control columns it would make the design safer for certain ultra-rare scenarios brought on by the failure of CRM and cockpit gospel rules. It would also make it less safe in many other ways. It Boeing ditched their control columns for sidesticks and hard envelope protections, the same is true in reverse. What's the point? Neither are unsafe! Neither have ever caused a crash.

                              Maximus is just trolling the know-nothing population with shock 'journalism' in a mad grasp for attention and the revenue it brings him. Why are we even discussing this tripe?
                              Ok. So you agree 100% with my original comment then. Thank you.

                              --- Judge what is said by the merits of what is said, not by the credentials of who said it. ---
                              --- Defend what you say with arguments, not by imposing your credentials ---

                              Comment


                              • #30
                                Originally posted by Gabriel View Post

                                Ok. So you agree 100% with my original comment then. Thank you.
                                Yes, except that you call it 'quite disappointing' and I call it 'tripe'.

                                Comment

                                Working...
                                X