Announcement

Collapse
No announcement yet.

Emirates 777 crash-landed in Dubai

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

  • Originally posted by brianw999 View Post
    Really, it all comes down to one simple word....

    Aviate
    Fixed (for the context of this thread)

    --- 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


    • Originally posted by Gabriel View Post
      As an American Airlines instructor said, if the automation is not performing as expected, don't ask "what is it doing now?"
      And this is what I'm talking about. If the pilot is expecting TO/GA switches to work with the mains on the runway, he doesn't know the airplane well enough. He should KNOW that doesn't work. Because that's half the job these days.

      Comment


      • Originally posted by BoeingBobby View Post
        I have watched this video probably a dozen times. I just watched it again, just as good as I remember it.

        I've seen most of his other ones, I had never seen this one. And it's probably the most relevant of all of them these days. Since 1997 so many accidents could have been prevented by following everything he talks about there. Excellent explanations, examples and presentation.

        Comment


        • Originally posted by Evan View Post
          And this is what I'm talking about. If the pilot is expecting TO/GA switches to work with the mains on the runway, he doesn't know the airplane well enough. He should KNOW that doesn't work. Because that's half the job these days.
          That's not the point. Or not the only point at least. The pilot should expect the automation to fail at any time and for any reason, so they need to monitor its performance. In this case, even if he didn't knew that the TOGA switch doesn't work on the ground, or if he knew it but in the heat of the moment forgot it or made a mistake, since (if) he pushed the TOGA switch with the intention to activate the GA mode, the pilot should have expected the TLs to move forward, all engine parameters to go up (N1, N2, EPR, FF), engine noise increase, sensation of being pushed against the seat back, speed increase. In the scenario that you present, nothing of that happened. In the scenario that you present, it was due to the mains being on the ground, but it could have been anything else like a different human error or a technical fault (remember for example Turkish approach in Amsterdam where the AT kept the engines at idle due to a technical fault and compare it to Asiana approach in SFO where the AT kept the engines at idle due to being wrongly configured by the pilot, both with basically the same outcome, both of which could have been corrected before it was even an incident by monitoring the performance of the automation and taking manual action regardless of the cause). So the pilot should have detected that the "output" parameters (all those mentioned before) were not following the expected performance and should have taken over and corrected (which simply meant firewalling the throttles with the same hand that he should have kept on the TLs after pressing the TOGA switch). Later, after stabilizing the situation, he could reflect on "why was it doing that" and probably he would have realized that he made the mistake.

          --- 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


          • Originally posted by Gabriel View Post
            That's not the point. Or not the only point at least. The pilot should expect the automation to fail at any time and for any reason, so they need to monitor its performance. In this case, even if he didn't knew that the TOGA switch doesn't work on the ground, or if he knew it but in the heat of the moment forgot it or made a mistake, since (if) he pushed the TOGA switch with the intention to activate the GA mode, the pilot should have expected the TLs to move forward, all engine parameters to go up (N1, N2, EPR, FF), engine noise increase, sensation of being pushed against the seat back, speed increase. In the scenario that you present, nothing of that happened. In the scenario that you present, it was due to the mains being on the ground, but it could have been anything else like a different human error or a technical fault (remember for example Turkish approach in Amsterdam where the AT kept the engines at idle due to a technical fault and compare it to Asiana approach in SFO where the AT kept the engines at idle due to being wrongly configured by the pilot, both with basically the same outcome, both of which could have been corrected before it was even an incident by monitoring the performance of the automation and taking manual action regardless of the cause). So the pilot should have detected that the "output" parameters (all those mentioned before) were not following the expected performance and should have taken over and corrected (which simply meant firewalling the throttles with the same hand that he should have kept on the TLs after pressing the TOGA switch). Later, after stabilizing the situation, he could reflect on "why was it doing that" and probably he would have realized that he made the mistake.
            Well it is the point if your point is that piloting a 777 requires more than basic airmanship. Certainly he should have had a hand on the thrust levers during approach and go around, that's not the issue we are discussing. We are discussing how piloting has changed over the years to be both an airmanship job and a systems administration job—and that it requires expert knowledge in both. The automation didn't fail here. The pilot failed to understand it (apparently).

            From an airmanship, systems knowledge and CRM perspective, this is what SEEMS to have gone wrong:

            - PF failure to have a hand on the thrust levers; and to advance them manually when they didn't respond
            - PF failure to understand the limitations of the automation
            - PM failure to monitor N1 to verify TO/GA thrust (or lack thereof)
            - PF or PM premature gear retraction or mistaken gear retraction

            #2 on that list is an all too common factor these days. And it leads to pilot error in basic airmanship because "what's it doing now" is a natural human factor that distracts concentration even from otherwise very seasoned pilots. It shouldn't, and you can make that point all you want, but it does.

            Comment


            • Originally posted by Evan View Post
              Well it is the point if your point is that piloting a 777 requires more than basic airmanship. Certainly he should have had a hand on the thrust levers during approach and go around, that's not the issue we are discussing. We are discussing how piloting has changed over the years to be both an airmanship job and a systems administration job—and that it requires expert knowledge in both. The automation didn't fail here. The pilot failed to understand it (apparently).

              From an airmanship, systems knowledge and CRM perspective, this is what SEEMS to have gone wrong:

              - PF failure to have a hand on the thrust levers; and to advance them manually when they didn't respond
              - PF failure to understand the limitations of the automation
              - PM failure to monitor N1 to verify TO/GA thrust (or lack thereof)
              - PF or PM premature gear retraction or mistaken gear retraction

              #2 on that list is an all too common factor these days. And it leads to pilot error in basic airmanship because "what's it doing now" is a natural human factor that distracts concentration even from otherwise very seasoned pilots. It shouldn't, and you can make that point all you want, but it does.
              Okay Evan, Let us say you have made your point. What is YOUR recommendation to solve this problem?

              Comment


              • Originally posted by Evan View Post
                Well it is the point if your point is that piloting a 777 requires more than basic airmanship.
                Sorry, but it requires only airmanship if, as I(and others) do, you include in its definition, all the skills, capabilities, knowledge, behaviors, attitude, mindsets and judgement that are necessary to fly a plane competently and safely. You seem to be mistaking airmanship for stick-and-rudder (and throttle) skills.

                Certainly he should have had a hand on the thrust levers during approach and go around, that's not the issue we are discussing. We are discussing how piloting has changed over the years to be both an airmanship job and a systems administration job—and that it requires expert knowledge in both. The automation didn't fail here. The pilot failed to understand it (apparently).
                The automation failed to perform as the pilot expected (for whatever reason) and the pilot failed to detect it and take effective action in a timely fashion.
                In this scenario that failure was triggered by pilot error. Would you be any happier if this failure had been triggered by a system malfunction or design bug?

                From an airmanship, systems knowledge and CRM perspective, this is what SEEMS to have gone wrong:

                - PF failure to have a hand on the thrust levers; and to advance them manually when they didn't respond
                - PF failure to understand the limitations of the automation
                - PM failure to monitor N1 to verify TO/GA thrust (or lack thereof)
                - PF or PM premature gear retraction or mistaken gear retraction

                #2 on that list is an all too common factor these days. And it leads to pilot error in basic airmanship because "what's it doing now" is a natural human factor that distracts concentration even from otherwise very seasoned pilots. It shouldn't, and you can make that point all you want, but it does.
                And # 2 could have been "the pilot knew the system perfectly and totally understood the implications, but he initially reacted by instinct and muscle memory (instead of a deeper cognitive process) by doing to what is most often practiced in the simulator and most often happens in real life: the go-around is started before the mains touch down. Perhaps he even made the mental decision to start the go around before the main touched down but in the second or so that it takes to go from decision to effective action (press the TOGA switch) the airplane touched down.

                I am not saying that this is what happened, but it could have happened. As well as a system failure is not impossible.

                I fully agree that systems training and understanding is fundamental and necessary. But an eventual lack of that is far from explaining, by itself, what happened here.
                I remember a similar discussion with AF, where you claimed that the main cause was that the pilots didn't follow the UAS procedure. And I said ok, but that by itself is far from explaining why the pilot pulled up so hard that he triggered the stall warning and then entered a 1.5G, 15deg ANU, 7000fpm, 2000ft climb and then, when the stall warning sounded again at the top of the climb (and kept sounding for 1 1/2 minutes), he reacted by "pulling up all the time" (in the pilot's own words).

                --- 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


                • Originally posted by BoeingBobby View Post
                  Okay Evan, Let us say you have made your point. What is YOUR recommendation to solve this problem?
                  Oh no, for the love of Saint Pelagea of Thessaloniki, please, no!

                  Comment


                  • Originally posted by Gabriel View Post
                    Sorry, but it requires only airmanship if, as I(and others) do, you include in its definition, all the skills, capabilities, knowledge, behaviors, attitude, mindsets and judgement that are necessary to fly a plane competently and safely. You seem to be mistaking airmanship for stick-and-rudder (and throttle) skills.
                    When I say 'basic airmanship', I am referring to the universal skills required to fly the 777 or the Tomahawk. What I am saying is that '777 airmanship' must include a thorough systems awareness as well.

                    Look, we are not arguing about that here. Yes, it always comes first and despite what happened here the pilot should have had his hand on the thrust levers, the PM should have had his eyes on the needles, both should have recognized the lack of response and the PF should have pushed the thrust levers manually. We agree on that.

                    What led me into this line of discussion was the question as to whether the old-school pilots were better pilots because they had less automation. I answered that the entire job of piloting has changed and that basic airmanship alone no longer qualifies you to command a digital flight deck. If you are weak on systems, you need to get out of there until you aren't

                    The automation failed to perform as the pilot expected (for whatever reason) and the pilot failed to detect it and take effective action in a timely fashion.
                    Again (circular argument), the automation didn't fail to perform, the pilot failed to understand it.

                    In this scenario that failure was triggered by pilot error. Would you be any happier if this failure had been triggered by a system malfunction or design bug?
                    I want to add here that the 777 autopilot is not only fail-passive, it is fail-operational. The engineers have gone to extraordinary lengths to make certain it "performs as expected", but a single pilot with flawed expectations of the system can prang the whole thing in an instant.

                    I remember a similar discussion with AF, where you claimed that the main cause was that the pilots didn't follow the UAS procedure. And I said ok, but that by itself is far from explaining why the pilot pulled up so hard that he triggered the stall warning and then entered a 1.5G, 15deg ANU, 7000fpm, 2000ft climb and then, when the stall warning sounded again at the top of the climb (and kept sounding for 1 1/2 minutes), he reacted by "pulling up all the time" (in the pilot's own words).
                    Improvisation, made necessary by a lack of procedural training and botched due to human factors and/or bad piloting. Following procedures minimizes the effects of these two things

                    Comment


                    • Originally posted by BoeingBobby View Post
                      Okay Evan, Let us say you have made your point. What is YOUR recommendation to solve this problem?
                      I'm not a 777 pilot. I don't even fly it in the sim. I'm just fascinated by aviation investigations, aerospace engineering, and have a natural curiosity that leads me to do my research and learn how these aircraft function, how the are designed and why these accidents continue to happen. So I study up on them in my spare time. I don't get paid to do that. It's not my profession. But it's not that hard either.

                      My first speculation on this thread (#10) pretty much nailed the issue here (if the reports thus far are correct) because I knew the system limitations and the potential pitfalls they represent. Now if I—a non-777-rated non-pilot—knew this why can't I expect a fully rated, salary-earning professional pilot to know these things?**

                      I don't think it is too much to ask that EVERY type rating be contigent on a very thorough and proven understanding of the automation behaviors and limitations. I also don't think it too much to ask that professional pilots put some of their personal time away for boning up on the FCOM, FCTM and the various industry publications that support a greater understanding of the aircraft they are entrusted with. As I said before, piloting is a much more intellectual profession than it once was. And the responsibility is always a dreadful one.

                      Since I've been on this board we have seen a number of fatal accidents arising from automation or systems "not behaving as expected" yet not misbehaving at all. There can be no argument that this is a leading causative factor in the degradation of situational awareness and departure from stable flight.

                      My solution is a stronger focus on education and re-education to familiarize pilots with the limitations and behaviors of any system that can be considered critical to stable flight control. There's a lot to know on these FBW jets, but there's clearly a need to know and if it takes more time and more money to train pilots adequately than so be it. I'm really tired of being stunned by the lack of demonstrated knowledge possessed by supposedly type-rated pilots. I get the solid impression of a "need-to-know" policy in pilot training that skips over things that modern pilots need-to-know.

                      And yes, of course training more discipline for the basics such as having your hands, your eyes and your attention where they belong, especially in critical phases such as go-around.

                      **I posted a while back that this PF was apparently a freelancing or borrowed Qantas pilot and that Qantas doesn't have any 777 pilots. One bit of speculation I put out was that Emirates was not providing adequate training for the type to pilots with no on-type experience. This could explain the flawed "expectations" he might have had re: the TO/GA switches. If this is the case, we also need to address this issue with stricter regulation.

                      Comment


                      • Originally posted by Evan View Post
                        I'm not a 777 pilot. I don't even fly it in the sim. I'm just fascinated by aviation investigations, aerospace engineering, and have a natural curiosity that leads me to do my research and learn how these aircraft function, how the are designed and why these accidents continue to happen. So I study up on them in my spare time. I don't get paid to do that. It's not my profession. But it's not that hard either.

                        My first speculation on this thread (#10) pretty much nailed the issue here (if the reports thus far are correct) because I knew the system limitations and the potential pitfalls they represent. Now if I—a non-777-rated non-pilot—knew this why can't I expect a fully rated, salary-earning professional pilot to know these things?**

                        I don't think it is too much to ask that EVERY type rating be contigent on a very thorough and proven understanding of the automation behaviors and limitations. I also don't think it too much to ask that professional pilots put some of their personal time away for boning up on the FCOM, FCTM and the various industry publications that support a greater understanding of the aircraft they are entrusted with. As I said before, piloting is a much more intellectual profession than it once was. And the responsibility is always a dreadful one.

                        Since I've been on this board we have seen a number of fatal accidents arising from automation or systems "not behaving as expected" yet not misbehaving at all. There can be no argument that this is a leading causative factor in the degradation of situational awareness and departure from stable flight.

                        My solution is a stronger focus on education and re-education to familiarize pilots with the limitations and behaviors of any system that can be considered critical to stable flight control. There's a lot to know on these FBW jets, but there's clearly a need to know and if it takes more time and more money to train pilots adequately than so be it. I'm really tired of being stunned by the lack of demonstrated knowledge possessed by supposedly type-rated pilots. I get the solid impression of a "need-to-know" policy in pilot training that skips over things that modern pilots need-to-know.

                        And yes, of course training more discipline for the basics such as having your hands, your eyes and your attention where they belong, especially in critical phases such as go-around.

                        **I posted a while back that this PF was apparently a freelancing or borrowed Qantas pilot and that Qantas doesn't have any 777 pilots. One bit of speculation I put out was that Emirates was not providing adequate training for the type to pilots with no on-type experience. This could explain the flawed "expectations" he might have had re: the TO/GA switches. If this is the case, we also need to address this issue with stricter regulation.

                        Every six months all part 121 pilots go for a simulator check. This includes a 3 or 4 day ground school where systems are covered in detail, plus all of the regulatory crap. i.e. TSA, hazmat training, CRM, TEM. This is followed by a 2 day simulator session. First day in called a PC warm up and the second day is the PC, which is basically a type rating oral and check ride. This is followed by an RFT ride 6 months later. Not sure how much more training they can do or have time for and still have the pilots be productive. Humans are still going to fuck up from time to time, it is just a design flaw.

                        Comment


                        • Originally posted by BoeingBobby View Post
                          Every six months all part 121 pilots go for a simulator check. This includes a 3 or 4 day ground school where systems are covered in detail...
                          I suppose "in detail" is a matter of opinion. I'm on my third 121 carrier, and I can tell you that a good half of my systems knowledge came from my own self-study and reading additional materials. Granted, I'm somewhat of a systems geek.

                          I'm scared to say this, but I do agree with Evan, at least to a certain extent. It's not uncommon that carriers teach only the bare minimum to pass the oral and the type ride, but nowhere near as much nuts-and-bolts stuff as they could be teaching. Many pilots I've flown with (like myself) compensate by additional reading and research. Some don't.

                          I haven't the foggiest how systems training is at EK.

                          Comment


                          • Originally posted by ATLcrew View Post
                            I suppose "in detail" is a matter of opinion. I'm on my third 121 carrier, and I can tell you that a good half of my systems knowledge came from my own self-study and reading additional materials. Granted, I'm somewhat of a systems geek.

                            I'm scared to say this, but I do agree with Evan, at least to a certain extent. It's not uncommon that carriers teach only the bare minimum to pass the oral and the type ride, but nowhere near as much nuts-and-bolts stuff as they could be teaching. Many pilots I've flown with (like myself) compensate by additional reading and research. Some don't.

                            I haven't the foggiest how systems training is at EK.

                            Do some research on what it take to get an ATPL, most of us in the US would never pass the exam.

                            Comment


                            • Originally posted by BoeingBobby View Post
                              Do some research on what it take to get an ATPL, most of us in the US would never pass the exam.
                              That's a whole separate conversation. There is an amazing amount of completely irrelevant stuff on the EASA exam.

                              Comment


                              • Originally posted by ATLcrew View Post
                                That's a whole separate conversation. There is an amazing amount of completely irrelevant stuff on the EASA exam.
                                Not sure if is correct or not, but Evan stated that the PIC was a Quantas pilot. If that is the case, he has one. And on a different note, yes in the recurrent ground school they go over just the basics. I don't know how it was where you are now, but at Atlas the basic indoc class is 7 weeks long and goes pretty deep into systems. Yes it requires the student to do a lot of self study to get it all but the original type oral is about 4 hours and is pretty heavy on systems.

                                Comment

                                Working...
                                X