Announcement

Collapse
No announcement yet.

Skills, Training, CRM, Culture... Safety

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

  • #46
    Originally posted by Evan View Post
    ...I think this crew may have lost that critical instinct.
    Worse, the automation may be letting crews that never had that instinct to get into cockpits in the first place.

    Comment


    • #47
      Even, I mostly agree with your last scenario: They knew that they were not stabilized and that they were required to go around, but they also knew that they could still correct it and get a good landing. So they decided to disregard the requirement. Except that the automation didn't work as they expected and, in the last moments, they didn't keep a good eye on the airspeed and a good hand on the throttle.

      So you have:
      1- Disregard of procedure.
      2- Wrong understanding of automation, or forgetting to re-configure it.
      3- Not proper monitoring of flight and engine parameters.

      Those are three holes in three different slices of cheese. All those failures were required for the accident to happen. Where two of them are mistakes, lack of training, etc., the first one is not.

      FOQA can detect the previous instances where 1 happened but 2 or 3 didn't, and hence it all ended in a happy landing.

      Of course, as Dispatch Dog said, this only works in a company with a good culture of safety.

      --- 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


      • #48
        Originally posted by Dispatch Dog View Post
        Worse, the automation may be letting crews that never had that instinct to get into cockpits in the first place.
        No, wrong... Management is letting thise crews into the cockpit. Automation is a tool. We have to stop blaming the automation. Airlines have to be required to train pilots on respecting the basic rules and also what their role is when under automation: to manage, to monitor, and to provide redundancy.

        A stabilized approach on automatics requires the pilot to maintain situational awareness of these basic parameters:

        • Horizontal situational awareness :
        − By closely monitoring the flight path;
        • Speed awareness :
        − By monitoring speed deviations;
        • Vertical situational awareness :
        − By monitoring the vertical flight path and the rate of descent;
        • Energy awareness :
        − By maintaining the engines thrust to the level required to fly a 3-degree approach path at the final approach speed (or at the minimum ground speed, as applicable).

        As Gabriel proposes here, they may have had awareness on the first three, bit I think they botched they fourth point for sure.

        Comment


        • #49
          I read what Gabe has to say. I read what Evan has to say. Then they both seem to iterate.

          I'm beginning to get the idea that four folks were on the flight deck each waiting for someone else to do something. Is it that remote a possibility that a newly minted check pilot was waiting for the pilot flying to do something but that the captain was waiting for a command from the check pilot?

          I was blessed at one point to have had one of Delta's senior check pilots as my instructor. He took me on after I had a rough month with the ego maniac instructor from hell. He constantly belittled his students, not in a constructive way but would also show disrespect by not showing for lessons and then I'd learn he was playing a few sets of tennis, not even having the decency to call and cancel (hell at least lie). The Delta chap was also an Air Force check pilot for "Thuds" so he was friendly but all business. I was on the verge of tossing in the towel but after a week with "Larry" was starting to regain a glimmer of hope. During one of our afternoons, I looked over my shoulder to find him "fast asleep" and with that I called his name ... "sure you want't to be napping now ha ha?" ... he just deadpanned "You fly fine, let me get ten or fifteen minutes more down time". You can bet your sweet arz that "Larry" wasn't napping, but till I figured that out, I'm sure that was one of the many tricks in his bag to give a guy back his dignity and confidence. Just a few posts back I mentioned changing the tail wheel and main wheel on a Bird Dog. I was out of the military and certainly not legal to be working on a civilian plane. It was fly or not fly that day and he watched me like a hawk then signed off. He was also airframe and power plants rated.
          Live, from a grassy knoll somewhere near you.

          Comment


          • #50
            One of the four pilots was not in the cockpit.

            Comment


            • #51
              Deadstick, if not for the CVR's you could make that argument for the whole lot of em'.
              Live, from a grassy knoll somewhere near you.

              Comment


              • #52
                Originally posted by guamainiac View Post
                Deadstick, if not for the CVR's you could make that argument for the whole lot of em'.
                Actually, not.

                This argument could have been made more for the pilots in the Turkish flight, for example, where they were in AP / AT and didn't note a thing until the stickshaker.

                But these ones were hand flying and making active corrections to the glide slope.

                In the Turkish flight, nobody was flying the plane.
                In this case, they were flying and consciously making illegal decisions.

                That's why I think that it's worse. Intentional negligence is worse than being incompetent.

                --- 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


                • #53
                  Just wanted to drop in and say I am 100% in agreement with your original post Gabriel, and I hope the industry as a whole is going to seriously look into these issues moving forward. We need to train aviators.

                  Comment


                  • #54
                    Fascinating thread. As I've said a few times, I don't fly planes, I don't spend hundreds of hours studying things like plane types and flying techniques.

                    Where I do have some years of experience is in the specialty of quality control. I started working in that area in the early 80's and continued in various capacities till the middle 90's. In the process, I saw a lot of things that happen when a company wants a better reputation for quality.

                    In early stages, some companies measure quality in a macro way as mistakes that got out and affected the consumer. Those things affect the reputation of a company, which will have an impact on business.

                    Usually if a company doesn't give up for some reason, they start to think about catching errors before there is customer impact. Studying error patterns to see how quality problems develop. In the early stages, measurement tools are sometimes crude. By the last decade, software had developed to the place you could get stats on EVERYTHING. Speed of response to errors, turnaround to a fix, fixes that work as a percentage of the whole. Fixing problems is a process of innovation. Until there is a track record on specific fixes for specific problems, we never know which are dependable fixes, which are just lucky coincidences.

                    Getting back to aviation. I've run into people very impressed with "how many fatal accidents in the last decade". I've always thought that was a pretty lame statistic. That is a grotesquely bad scenario. To actually have the blood of your customers on your hands. I think the legal department came up with that one.

                    Because, really, people don't willingly get on a plane that they think will do a belly landing, a water landing, really anything but the near-perfect landing. So really, for a quality analyst, all landings except the near-perfect, are meat to be digested and used to revise every pertinent company standard. Aviation fans may find this rigorous, but in banking we did this, and we can't even kill people!

                    Those "safe landings" are often very expensive. You may have all your passengers out alive, but if your plane is in the Hudson River, what kind of expense hit does that mean for your company? Or is it that the insurance company pays off, so quality is irrelevant? I think if you want to discuss company culture, you want to take the Japanese attitude of decades back that ANY fault is not acceptable. You do not, like some manufacturers did back then, work to a statistic of "acceptable faults". If you've set your standard that low to start with, you've already sent a message to your employees that the company is "cool with mistakes".

                    Comment


                    • #55
                      EC,

                      Airlines with a good safety culture do just that. It's called FOQA.
                      They download the data of every flight and process it to look for anything out of "norm". Bank angle past a certain limit? Flagged. Sink rate on final beyond certain limit? Flagged. Approach not stabilized? Flagged. Hard touchdown? Flagged. Too low fuel on landing? Flagged.

                      They follow the trends of each characterized deviation, and look for correlation with other factors (do this happen statistically more often in one airport, one airplane type, one hour of the day...?).

                      Then they try to correct these issues before they have to do it anyway after it causes an accident.

                      --- 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


                      • #56
                        Gabriel, is this an actual factual representation of a carrier.

                        I'm impressed.
                        Live, from a grassy knoll somewhere near you.

                        Comment


                        • #57
                          Originally posted by guamainiac View Post
                          Gabriel, is this an actual factual representation of a carrier.

                          I'm impressed.
                          Is that a question?
                          Is it about the FOQA program?
                          If yes and yes, then yes. A lot of airlines use it. I'd risk to say that most do, since it's a requirement for Europe, for all alliances (like SkyTeam or OneWorld), and to pass the IOSA audit what is required to be a IATA member.

                          Now, how many of them take it seriously and how many just to meet the requirement? I don't know.

                          This is an interesting post by a professional pilot. It starts with an unstabilized approach and ends with FOQA. Really very relevant for this thread and the Asiana one.


                          --- 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


                          • #58
                            Originally posted by Gabriel View Post
                            EC,

                            Airlines with a good safety culture do just that. It's called FOQA.
                            They download the data of every flight and process it to look for anything out of "norm". Bank angle past a certain limit? Flagged. Sink rate on final beyond certain limit? Flagged. Approach not stabilized? Flagged. Hard touchdown? Flagged. Too low fuel on landing? Flagged.

                            They follow the trends of each characterized deviation, and look for correlation with other factors (do this happen statistically more often in one airport, one airplane type, one hour of the day...?).

                            Then they try to correct these issues before they have to do it anyway after it causes an accident.
                            Remember back with AF447 when people discussed continuous transmission of flight data to the ground so that the lost black boxes wouldn't hold up the investigation? I remember people saying that bandwidth made it impossible (though a Canadian company claimed it had the hardware ready). I believe (though I'm certainly able to be wrong) that someone claimed that only mechanical data was used after successful flights so that the maintenance department could look for evidence of failing components. And that certainly is part of a quality program. But really all data should be measured against the standards set in training. Pilot efficiency can degrade just like the mechanical components. You could require a complete physical before every flight, but I think all the data gathered by the black boxes are another kind of physical in that they show how the most important component in the plane, the human, is progressing. Everyone makes a mistake now and then. Mechanics are variable too. But if you let your software profile all the errors in a flight, I really believe you could determine, for one thing, if all the pilots are deviating or it is just this one. I did lots and lots of this kind of analysis. We put together training classes where we learned there were repetitive mistakes by many staff members.

                            And then there's other stuff. Maybe your documentation is short of bullet proof. Maybe it isn't easy to use. Skilled employees can often cover this sort of thing. So you have to dig into the data to find out how good your infrastructure is.

                            Comment


                            • #59
                              Hi Guys,

                              Long time stalker - first time poster.

                              I have a couple of points I'd like to make about this issue;

                              1) IF the data is available for each flight and IF the decision 'gates' are considered unbreakable rules then why don't the FAA (or whomever for each country) impose fines/suspensions for breaches of these conditions. I'm thinking of a system SIMILAR to a driver running a red light in their car (sure there are some HUGE differences) but this would reinforce the pilots 'respect' for doing the correct thing. (I missed being stablised by 500', I best go around or I'll be docked another 2 points on my license, etc.)

                              2) I have worked in a V safety concious environment and we had a talk from an executive from DuPont, please bear with me on this - it's a great story and I believe has merit.
                              Years ago DuPont made explosives in overseas factories and occasionally a factory would blow up. While the loss of the lives of the overseas employees was not seen as a huge problem it became apparent that broken factories hurt the bottom line for the company. In order to fix the situation DuPont made their managers (and families) live inside the walls of the factory compound. Overnight they went from not giving a damn to NEVER HAVING A FACTORY EXPLOSION AGAIN.

                              So my question here is... Do the execs from each airline fly their own airline? Not on a 'special' plane that is treated 'differently' from the rest of their fleet. If not, why not, maybe then they'd think twice about saving a dollar by reducing fuel loads etc.

                              Thanx for the rant

                              VAZ

                              Comment


                              • #60
                                Originally posted by vaztr View Post
                                IF the data is available for each flight and IF the decision 'gates' are considered unbreakable rules then why don't the FAA (or whomever for each country) impose fines/suspensions for breaches of these conditions. I'm thinking of a system SIMILAR to a driver running a red light in their car (sure there are some HUGE differences) but this would reinforce the pilots 'respect' for doing the correct thing. (I missed being stablised by 500', I best go around or I'll be docked another 2 points on my license, etc.)
                                To ensure collaboration and an atmosphere of commitment rather than fear, the raw data is strictly confidential, depersonalized, and treated on a statistical basis. Now and then they will use an example of a real flight to make a point on an awareness seminar or something similar, but all the info that could help identify the involved crew (date, sometimes airport, sometimes airplane) is removed.

                                The main difference with the driving offense that you mention is that an airline is a company with the power to define their own policies of hiring, training, retention, to write their procedures and to manage their safety culture. That doesn't happen with the car drivers as a group.

                                --- 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

                                Working...
                                X