Announcement

Collapse
No announcement yet.

Air Canada pulls a Hans Solo

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

  • #61
    Originally posted by 3WE View Post
    No, not human nature...the violation of the basic principle of fly the GD airplane.
    THAT IS HUMAN NATURE! What will it take to get that through your head 3WE? When events vastly differ from expectations, and up is down and left is right, we see the violation of the basic principle of fly the GD airplane.

    That way we prevent this—the only way—is to prevent situations whereby expectations are so wrong.

    You think type ratings cover these things (and they should) but the evidence is contrary in these cases. I'm sure type ratings cover the standard profiles, but in each of these cases there was a deviation from the standard profile. In the case of Asiana, they got behind in the final, high on the slope and opted to use a supplemental vertical mode that they didn't understand in a phase of flight where it was completely inappropriate. They may have done this because it is appropriate to use that mode at higher altitudes to rapidly lose height but only with a safe level-off or transition altitude factored in. But they weren't sufficiently trained to know this, or to set the FCU altitude safely and to understand the interaction of the AP and the AT when using that mode.

    And yes, fly the GD plane means monitor. We've got to impress upon these pilots that the nature of the job under automation is to fly the automation using the instruments. Not to take a break on final.

    So we have ignorance on supplemental modes and complacent trust in automation. Then it's just: what the... (this is not supposed to happen!) TOGA... seawall... panic.. pull up relentlessly... prang!

    The "this is not supposed to happen" part is where the fundamental stuff turns to smoke.

    If this veteran 747 pilot had decided to handfly the approach, I'm sure the worst thing that would have come out of it is a go-around. Because his expectations would have been more-or-less correct.

    Comment


    • #62
      Originally posted by Evan View Post
      THAT IS HUMAN NATURE! What will it take to get that through your head 3WE? When events vastly differ from expectations, and up is down and left is right, we see the violation of the basic principle of fly the GD airplane.
      No, Evan...NOT the inversion of up and down and left and right...

      More mundane things like warning noises.

      Yeah, sure, they can be loud and I'll give some one a split second freak out...

      But maybe count to 10...feel that the plane is still going forward...and think about a robust heading and power setting. Don't yank the yoke back...If you are on an ILS at 200 & 1/2- maybe your hands are in place and your mind is reviewing the feeling and importance of watching the AI (however it's depicted).

      I loved the Sully movie of how the guy starts thumbing through the QRH for DUAL engine failure. (Not realistic, but hey you IN FACT have a big ass procedure book because there's a shit pot of ways your automation can crap out!)

      If your Airbus goes crazy and you don't know if it's UAS or a bomb has blown a hole in the cargo hold...FUNDAMENTALS says don't freak but instead fly the damn plane...nothing type specific about that at all whatsoever.

      But you just want to cram more acronyms down their throat, instead of reminding them that pausing to check on a robust attitude and power setting (maybe even a gentle descent), does not harm you as you take 30 seconds to realize it IS UAS and not a bomb, and THEN you can look up UAS in the QRH...
      Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

      Comment


      • #63
        Originally posted by BoeingBobby View Post
        You have to be shitting me, they have auto-land in 172's now?
        Not exactly. Just that touching down in a 172 at 60 or 65 knots in a 3 deg slope approach, with some 300fpm, and no flare will likely be not fatal neither for the plane nor the persons on board.

        --- 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


        • #64
          Originally posted by Evan View Post
          ...who can't land a 777 for what reason? Obviously you can't weed out those who lack the skills to fly the 777 BEFORE you begin teaching them those skills. But you CAN weed out the one's who lack aptitude, concentration, discipline, prudence and intellect in the first place. So that would be step one.

          Step two: weed out the things that erode a culture, like pilot gradient and favoritism towards gonzo ex-military pilots (because I believe we are talking about Asiana here). And weed out the idea that big hours on a different type qualifies them for something they don't have the training for.

          Step three: weed out the ones who have impeccable hand-flying skills but don't have the time-of-day to bother learning about systems. ("My airplane... Ok, what's it doing now?!" "Uh... captain, you have to pull the knob. You're pushing it.")
          Evan, I agree with all that you say but I don't agree with you not saying what you don't say.

          How about step zero, weed out those who lack the stick-and-rudder skills necessary to HAND fly ANY airplane safely and confidently? Remember that this guy mentioned that he was nervous because he didn't like hand-flying approaches? (even when it was a long straight-in final to a long, wide runway without obstacles in a beautiful CAVOK day with barely any wind or turbulence?)

          You are not really answering 3WEs question:

          Originally Posted by 3WE

          Do we need better screening to weed out the guys who can't hand land a 777 on a sunny evening with gentle winds...
          He didn't say proficiently fly the 777 taking advantage of all its advanced systems. He said "hand land a 777 on a sunny evening with gentle winds".
          I can do that (sorry BB, I do honestly believe I can, I could in a level D 737 sim without any outstanding effort and I don't see why would it be any different in any other airplane type).
          I would not let the speed go down and the nose go up at the same time in an uncontrolled fashion.
          And I know shit (much less than you, for example) about the systems in the 777. But I do know how to HAND FLY and use the elevator and throttle(s) in a coordinated way to manage things like vertical speed, airspeed, glide slope, angle of attack, and energy.
          You are not addressing that part.

          --- 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


          • #65
            Originally posted by Gabriel View Post
            Evan, I agree with all that you say but I don't agree with you not saying what you don't say.

            How about step zero, weed out those who lack the stick-and-rudder skills necessary to HAND fly ANY airplane safely and confidently? Remember that this guy mentioned that he was nervous because he didn't like hand-flying approaches? (even when it was a long straight-in final to a long, wide runway without obstacles in a beautiful CAVOK day with barely any wind or turbulence?)
            Ok, so you are telling me that it is possible to get a ATPL without the stick-and-rudder skills necessary to HAND fly ANY airplane safely and confidently. How do you get there without anyone noticing that? I mean, if there is any truth to that, we have a HUGE problem on our hands.

            Both you and 3WE are coming at the Asiana crash as if the reason for the crash was a reluctance to handfly on a nice CAVOK kind of day. It wasn't, and ANY airman must be able to fly that approach down to the mandatory disengagement altitude (or autoland) on automation on the type he is piloting. That's what it's there for. I have no idea as to whether he could or could not handfly that, but I would be VERY, VERY surprised if he couldn't. Again, the point is, he didn't have any reason to.

            My point is that these crashes don't occur because of rusty handflying skills. They happen because of poor understanding of what the automation can and can't do and how and when to use it. They also happen because of poor training for monitoring that automation along with the instruments. They happen in that blended regime when the automation is relying on actions from the pilot who doesn't realize this, or when it hands things over to the pilot and the pilot is not expecting this. They happen when the automation does something unexpected, not because there is anything wrong with it but because the pilot doesn't know what to expect. Because he's not properly trained on it. Because it's not a DC-3 and you need a lot more than stick and rudder skills to know how it works.

            Seriously, go back to DC-3's and we will eliminate this problem. Pilots will be much more in the game, will fly more instinctively and confidently. They will confidently fly into mountains and buildings and into each other as well, just like they used to. They will succumb to somatogravic illusions in IMC and fly into the sea. They will use more fuel and flying will be more turbulent and inefficient. And pilots will be exhausted after a nine-hour flight, and maybe just fly into the trees out of fatigue. Or they will survive crashes and say things like "I really have no idea why I did that."

            3WE thinks pilots just forget to fly the GD plane or pull up relentlessly because they're just morons who don't know any better. You and I both know it is human factors that cause these errors. And our three best defenses against human factors are procedural training, proficiency on automation and duty-time regulations.

            Comment


            • #66
              I will make one observation about the software profession which has some analogs here. As software has advanced we have created a lot of frameworks that "simplify" the act of programming. In these frameworks, the system takes care of a lot of things that the developer used to have to worry about. These frameworks work really well for simple programs, but as soon as you begin to push the envelope of complexity, it usually falls down and requires a person to hand tailor the system and code to work properly, or perform well.

              So what happens is that you suddenly make the development job available to a lessor skilled population and still be productive. However, when the system becomes more complex, the problems become MUCH more difficult to solve, because you're no longer just dealing with complex computer concepts, you're also dealing with this unique framework which is still brokering/translating/transforming your code before it gets to the machine. So, now to solve the problems at the limits of complexity, you have know the core basics of the machine underneath (which makes it hard for the lessor skilled population right off the bat) but you have to know exactly how that translation layer in the middle is working, and each framework has it's own idiosynchronicities. Those frameworks are often quite complex and understanding how they work in exception cases is often only discovered through experience. The end result is a complex system that is impossible to work on for the lessor skilled, and harder to work on for the properly skilled and trained.

              However, these days, no one would build most things without those frameworks. It's way to expensive to do things from scratch and takes too long. Many of us use these frameworks in the most basic of ways to keep them as predictable as possible.

              I think there are a lot of parallels here. Basic understanding of flying is pretty important, but these new planes are full of complex translation systems and when things happen near the edge of the envelope, it seems to me that the pilot has to know a lot more stuff now about the specifics of those complex systems.

              Comment


              • #67
                Originally posted by Evan View Post
                Ok, so you are telling me that it is possible to get a ATPL without the stick-and-rudder skills necessary to HAND fly ANY airplane safely and confidently.
                Welcome, you finally arrived to the party.

                How do you get there without anyone noticing that?
                By focusing on doing what it takes to take a person from zero hours to the right seat of an airliner in the shortest time possible. You need to study the AFM from cover to covert (what can be done using several hours per day NOT flying), you need to study ALL the POSSIBLE test questions and their answers (there is a finite pre-set, publicly available, list of questions with their choices and the correct one from which the FAA pick a bunch for your exam), you need to master exactly all the test maneuvers, of which landing of course is one of them, and then you need to collect hours instructing in exactly the same skills. So yes, you basically need to know how to fly an airplane, but the focus is not there. No time to "play" with the plane and understand her "language" (i.e. to understand what the plane is telling you with its performance and the force feedback, and how to tell her what you want her to do), no time for playing "what if" beyond the minimum requirement, no time for practicing lots of different approach scenarios and landing techniques that are not part of the test scope. Many of those things you may never use later in real ATP life, but they give you a background of informal knowledge and understanding the plane, any plane.

                Both you and 3WE are coming at the Asiana crash as if the reason for the crash was a reluctance to handfly on a nice CAVOK kind of day.
                No. Rather that his reluctance (not only in "he did not hand-fly it" but even more in saying that he was nervous with that) is a symptom of the kind of stick-and-rudder airman that we had there. Probably a good stick-and-rudder airman would not have let the approach deteriorate anywhere close to where it did EVEN IF HE DID THE SAME MISTAKES that the Asiana pilot did. He would have said "something is wrong, and at this point I don't care what it is, click click, clack clack", much earlier that this pilot did, when the plane was basically already stalling. Even, listen: You WILL HAVE PILOT MESSING UP WITH AUTOMATION. While we need to minimize that risk, we also need a second layer of cheese behind that one.

                (CAPS are for emphases, not shouting)

                --- 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


                • #68
                  Originally posted by Gabriel View Post
                  So yes, you basically need to know how to fly an airplane, but the focus is not there. No time to "play" with the plane and understand her "language" (i.e. to understand what the plane is telling you with its performance and the force feedback, and how to tell her what you want her to do), no time for playing "what if" beyond the minimum requirement, no time for practicing lots of different approach scenarios and landing techniques that are not part of the test scope. Many of those things you may never use later in real ATP life, but they give you a background of informal knowledge and understanding the plane, any plane.
                  Ok, now tell me about recent crashes that were the result of knowing the fundamentals of flying an airplane but not her "language" (and not her automation "language").

                  Probably a good stick-and-rudder airman would not have let the approach deteriorate anywhere close to where it did EVEN IF HE DID THE SAME MISTAKES that the Asiana pilot did. He would have said "something is wrong, and at this point I don't care what it is, click click, clack clack",
                  Then he has NO BUSINESS being in the cockpit of a modern airliner. Period. There is no excuse for "something is wrong, and at this point I don't KNOW what it is" in this scenario.

                  You WILL HAVE PILOT MESSING UP WITH AUTOMATION. While we need to minimize that risk, we also need a second layer of cheese behind that one.
                  I completely agree with this. I've never not agreed with this. I've never said that hand-flying skills were obsolete or unnecessary. I INSIST on solid hand flying skills. All I've been saying all along is that this is not what is behind most of these accidents. The larger problem is ignorance and/or misuse of automation, especially in the blended or transitional realm. So, it you can land your Piper Cub on the Golden Gate Bridge in a snowstorm, this does not qualify you to fly a commercial airliner.

                  Comment


                  • #69
                    Originally posted by Schwartz View Post
                    I think there are a lot of parallels here. Basic understanding of flying is pretty important, but these new planes are full of complex translation systems and when things happen near the edge of the envelope, it seems to me that the pilot has to know a lot more stuff now about the specifics of those complex systems.
                    No, not "a lot more stuff", nowhere near analogous to framework complexity. Engineers developed these systems to be as intuitive and pilot-friendly as possible, in many cases to mimic traditional aviation, and to degrade back to traditional aviation methods, not away form them, when they fail. Pilots need to know more specifics about automation interactions and behaviors but—when we are talking about the things that cause accidents— it's not a daunting or unreasonable amount. It does not create the kind of detrimental complexity that your analogy suggests. At any point a pilot can take full manual control of the aircraft and fly it intuitively.

                    For example, the Asiana crash was not due to excessive complexity. FLCH is a very simple mode to understand. But you can't use it with complete ignorance.

                    Comment


                    • #70
                      Originally posted by Gabriel View Post
                      Blah, Blah, Blah.
                      Good stuff, Concur, and yes, not only did Mr. Asiana appear to not know how to land an airplane, he was afraid to do so, AND it would seem his PM didn't know how to land one either!

                      Evan said (paraphrased). "C'mon guys, they were taught hand flying"

                      Valid point...BUT

                      1. As you (Gabriel) state, there's evidence that this teaching often (not always) neglects the cowboy fundamentals.

                      2. We have a decent list of accidents where it appears the pilot AND co-pilot don't know much about fundamentals.

                      3. To Evan's statement that they were taught fundamentals...an equally valid counter statement is that: More recently, they were given repeated, intense, type-specific, procedural-heavy training, tested for competence on that training, and even subjected to "IOE" type things...could be that the original training fades away. We even have Bobby's story (the believable part), that a FO asked why in the hell he was hand flying after 250 feet.

                      IMPORTANT DISCLAIMER: No doubt, the insiders are laughing at our outsider pontification from our parent's basements. They and the majority of their colleagues know the basics better than us. However, I think we can still ask them, how in the hell do these eye-rolling crashes happen and should something be done to prevent them. (And I'm not sure this mis-alignment belongs on that list).
                      Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                      Comment


                      • #71
                        Comments below.

                        Originally posted by Evan View Post
                        Ok, now tell me about recent crashes that were the result of knowing the fundamentals of flying an airplane but not her "language" (and not her automation "language"). Here we go...those crashes make YOU all torqued off and livid. Gabriel and I don't like them either. And the people killed are dead. Just like the people killed when pilots do things that go against crazy basic fundamentals. Gabriel and I get torqued off and livid about those. When those cases occur, you tend to cite some checklist item, or automation. EDIT: Just like two posts above...you cite that they didn't know about FLCH...no acknowledgement that 1) the dude was afraid to land and 2) the dude apparently didn't think that regularly glancing at airspeed was important...autoland or handland or autopilot to 200 feet or 707 steam or DC-3 or 182-glass or Piper cub.

                        Then he has NO BUSINESS being in the cockpit of a modern airliner. Period. There is no excuse for "something is wrong, and at this point I don't KNOW what it is" in this scenario. Really? There is no excuse for 'something is wrong'? I repeat- get out of Mom's basement and take the bicycle around the block...Go drive 5000 miles in a car...Try flying an airplane for yourself...Take a reality pill.

                        ***The larger problem*** Tell that to the families of dead people from "the smaller problem"...especially the ones killed in extremely preventable crashes...like those where the "how-to-stall-almost-any-airplane procedure" was demonstrated to a high degree of proficency.
                        Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                        Comment


                        • #72
                          Originally posted by Evan View Post
                          3WE thinks pilots just forget to fly the GD plane or pull up relentlessly because they're just morons who don't know any better.
                          You have some combination of these problems: 1) Bad reading comprehension, 2) Extreme bias, 3) Horrible reasoning ability.

                          I don't know which ones or the mixtures.

                          Similarly, I don't know why pilots pull up relentlessly- except that pulling up not_relentlessly is one way to make the plane go up. I don't know why two Asiana pilots were not monitoring airspeed on short final...I think airspeed on short final is pretty important, at least on these particular types of aircraft listed here: [Just about all of them].

                          I tend to blame the intense, recurrent, procedural-oriented training (that you love so much), and that that causes some pilots to forget that relentless pull ups are often ineffective at making the plane go up or stay up. (Oh, let's not forget that your beloved Airbus allows relentless pull ups...except when it doesn't...and what do you want to remember "when up becomes down and left becomes right"?)
                          Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                          Comment


                          • #73
                            Originally posted by Evan View Post
                            Ok, now tell me about recent crashes that were the result of knowing the fundamentals of flying an airplane but not her "language" (and not her automation "language").
                            It's not one OR the other. It is one AND the other.

                            Air France, Colgan, Diet Pepsi four one oh, Asiana, Turkish, Emirates...

                            All these could have been avoided by proper understanding and use of the type-specific automation. All of these could be avoided, AFTER THE TYPE-SPECIFIC AUTOMATION FAILED to perform as expected (due to issues with the expectations, not with the automation), by proper understanding and use of the UNIVERSAL DON'T-CARE-WHAT-TYPE BASIC FLYING SKILLS. And I don't buy "startle" as a magic bullet for all those. In fact, for none of those. Startle could have been a factor in the first second or two of SOME of those.

                            --- 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


                            • #74
                              Originally posted by Gabriel View Post
                              ***And I don't buy "startle" as a magic bullet for all those.***
                              I wonder what the proper, type-specific procedure is for startle factor is for all of the thousands of types and sub types of aircraft.

                              "Keep flying the damn plane using familiar, healthy attitudes and power and speed" is apparently wrong.
                              Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                              Comment


                              • #75
                                Originally posted by Gabriel View Post
                                It's not one OR the other. It is one AND the other.
                                Les règles de l'aviation de base découragent de longues périodes de dur tirer vers le haut.

                                Comment

                                Working...
                                X