Announcement

Collapse
No announcement yet.

Air France 447 - On topic only!

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

  • The airlines are rolling in profit
    I'd love to know which airlines these would be Evan. The airline industry has been losing money as a whole for many years. Only a select few airlines ever make decent profits.

    Picking up the part that TeeVee has bolded:

    The question then is, what actually is meaningful experience to an airline pilot these days? Realistically, there is only a certain amount of training that can be done. Can there be more than there is currently? Sure. But how much is enough? Is doing more hours in a KingAir going to make the airline pilot any safer? Who has actually been at the controls of recent airline accidents?

    The thing is, we seem to find this type of "pilot error" accident hard to swallow. We don't like thinking that the human reacted incorrectly when the machine failed, seeing that is why he is there. But the whole point is that while we have an increase in the rate of these errors, the overall rate has decreased.

    We are safer having a highly automated machine that occasionally kills us, than we are having a less automated machine that will kill us more often, but in different ways. But we like it less, and find it something that needs to be fixed.

    I'm not sure where the balance is between allowing pilots regular hands on currency (with the commensurate slight decrease in safety) compared to the current scenario. The insurance costs just don't let new pilots get the experience in the aircraft they should. But at the same time, is it new pilots that are crashing aeroplanes?

    Comment


    • Originally posted by MCM View Post
      I'd love to know which airlines these would be Evan. The airline industry has been losing money as a whole for many years. Only a select few airlines ever make decent profits.

      Picking up the part that TeeVee has bolded:

      The question then is, what actually is meaningful experience to an airline pilot these days? Realistically, there is only a certain amount of training that can be done. Can there be more than there is currently? Sure. But how much is enough? Is doing more hours in a KingAir going to make the airline pilot any safer? Who has actually been at the controls of recent airline accidents?

      The thing is, we seem to find this type of "pilot error" accident hard to swallow. We don't like thinking that the human reacted incorrectly when the machine failed, seeing that is why he is there. But the whole point is that while we have an increase in the rate of these errors, the overall rate has decreased.

      We are safer having a highly automated machine that occasionally kills us, than we are having a less automated machine that will kill us more often, but in different ways. But we like it less, and find it something that needs to be fixed.

      I'm not sure where the balance is between allowing pilots regular hands on currency (with the commensurate slight decrease in safety) compared to the current scenario. The insurance costs just don't let new pilots get the experience in the aircraft they should. But at the same time, is it new pilots that are crashing aeroplanes?
      I wonder, MCM, do the real numbers prove that humans caused more crashes before automation than they do now? I mean, pre-autopilot, when pilots actually HAD to hand-fly, did they crash because of pilot error more often than they do now? not sure where these stats would be. I do doubt that the percentages of crashes caused by pilot error were higher back then though...just a hunch

      Comment


      • Originally posted by MCM View Post
        I'd love to know which airlines these would be Evan. The airline industry has been losing money as a whole for many years. Only a select few airlines ever make decent profits.
        Ok, maybe not rolling in profit like the oil industry, but they are raking it in now. This is from a recent Associated Press article:

        After a decade of multibillion-dollar losses, U.S. airlines appear to be on course to prosper for years to come for a simple reason: They are flying less. By grounding planes and eliminating flights, airlines have cut costs and pushed fares higher. As the global economy rebounds, travel demand is rising and planes are as full as they've been in years.
        Profit margins at big airlines are the highest in at least a decade, according to the government. The eight largest U.S. airlines are forecast to earn more than $5 billion this year and $5.6 billion in 2012.

        "The industry is in the best position — certainly in a decade — to post profitability," says Southwest Airlines CEO Gary Kelly. "The industry is much better prepared today than it was a decade ago."
        And you might have missed that tax-holiday when the Republicans shut down the FAA over a labor issue and the airlines kept the prices steady and pocketed the difference...

        Now, that all very good for the airlines. I expect to see some improvements to training standards i.e. investment in training programs, and regulators should no longer accept the excuse that this is not a reasonable or sustainable expense. Of course, my point remains that it should not be up to the operators to decide how much training is required, and it some areas it seems that this is the case.

        Comment


        • Yeah TeeVee, but look at the other variables. How many heavy storms did they fly into without the color and look down radar?

          The old SOP for heavy weather was gear down to dirty the airframe so the thermals didn't cause overspeed. Stuff like that.

          I think there were too many other things going on to get a true perspective.

          My bud trusted his new fancy (old) AP one night. It was too hazy for me and my comfort zone so I passed on my turn. Well his trusty AP put us dead over an air base which he swore was a civilian airport. I looked down at the C-130's and KC's on the ramp and asked him if they had them at "Fly-On-In" and one look down put him in a panic. This was pre 9-11 and late and they were nice.

          You need to know and learn when to, and when not to trust the equipment. I think they entered new territory that night.
          Live, from a grassy knoll somewhere near you.

          Comment


          • TeeVee,

            Reporting definitions have changed a lot since those days, because "pilot error" has taken on different meanings - but you can check out the safety stats through ICAO, IATA and Flight Safety Organisation if you're interested.

            We are talking about "handling errors" rather than "pilot error", which can be so many different facets. I'd say there were far more crashes caused by handling errors 'back in the day', although that is of course caused by the fact they were doing far more hand flying! Looking at like for like, the fact current pilots don't hand fly often would no-doubt lead to a higher accident rate if overnight all autopilots were banned and it was back to hand flying.

            We've definitely moved into a world where hand flying is becoming unacceptable for a few reasons. For commercial reasons, we now operate aircraft far closer to the edge of their envelope, which requires far more precision... precision that is not practical for a human to perform at for long periods. We fly at higher altitudes, where hand flying is more difficult.

            Even in the terminal area it becomes harder - we have very noise sensitive airports that require precision tracking with large fines for small deviations - and the autopilot can fly that far more accurately than I can. Similar to operating to parallel runways and to 1000ft spacing in terminal areas - the autopilot gives more protection, and "frees up" the pilot's brainspace and eyes to be looking out and thinking about whats happening next.

            We all know the role of the pilot has changed - there are far more operational and "management" style decisions that need to be made than demonstrating hand-flying ability. But it is clear that complacency and lack of preparedness do need to be addressed... and I do think that is happening.

            Comment


            • Originally posted by MCM View Post
              ... we now operate aircraft far closer to the edge of their envelope, ... We fly at higher altitudes, where hand flying is more difficult.
              I don't agree with that little bit.

              I don't think that say an A330 or a B-777 operates today any closer to the edge of its envelope than the early say 747s that had a ceiling of FL410 that was used very often in long flights after a number of step climbs (not to mention the Concorde that reached I think some 60,000ft).

              Of course those were flown almost exclusively on AP while at cruise.
              When you are flying at nearly the speed of sound, just the workload of holding the altitude takes most of the pilot's available workforce. The reason is simple: At about 500kts a deviation in pitch (and hence in trajectory) of just one half of one degree up or down means a vertical speed of nearly 500 fpm (again not to mention the Concorde that flew more than twice that fast). And we are talking 70's here.

              I've been once in the cockpit of a 737 at cruise that had no working A/P (yes, it had been released with all AP MELled, I think that it was the FD what was inop and that rendered the AP unusable) and the FO was with both hands in the yoke and the eyes fixed in the attitude indicator, altimeter and VSI. The captain even had to call the heading corrections because it was too much for the FO to handle. I was in the cockpit just a few minutes, but I'm sure that after the 2 hours of manual flight like that that pilot must have been physically and mentally extenuated. (an age before 9/11, RVSM and CTR displays).

              --- 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


              • so you expect me to believe that if the ap goes inop midway across the pacific ocean, the pax are in huge danger?

                Comment


                • Originally posted by TeeVee View Post
                  so you expect me to believe that if the ap goes inop midway across the pacific ocean, the pax are in huge danger?
                  No. Not huge, not little. Ok, maybe a bit more in danger than if the AP is ok and running. But nothing to believe you are going to die. The pilots won't enjoy it (maybe they will for the first minutes), the ride will not be as smooth for the passengers, and It might be wise to advise ATC that you are not longer RVMS compliant (even if you legally were).

                  Now, if you are in an A330 (in normal or alternate law), things would be much easier for the pilots than in a 747. Unless it was some of those pilots...

                  --- 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


                  • Gabriel,

                    I was comparing the ops to before the age of extensive autopilot usage. The 747/Concorde are "modern" aircraft in that regard. That said, there is a lot to be said for autopilots that don't always get it quite right. They do the hard work, but the pilot needs to stay alert to catch when it decides to do something a little bit wrong. Not sure it would be politically correct to say we should build in deliberate errors to autopilots though .

                    If the autopilot goes inop half way across the pacific, no, you are not in huge danger. But speak to anyone who has had to hand-fly a sector in a 'modern' aircraft, and they'll tell you just how much hard work it is. So much hard work that the pilot who is flying the aircraft is pretty much having to concentrate on that, and not much else. It very much restricts the available brain capacity to handle other issues.

                    The point is though, that if we routinely did that kind of activity, there would be an incident rate higher than what we have with autopilot usage.

                    Comment


                    • Originally posted by Gabriel View Post
                      No. Not huge, not little. Ok, maybe a bit more in danger than if the AP is ok and running. But nothing to believe you are going to die. The pilots won't enjoy it (maybe they will for the first minutes), the ride will not be as smooth for the passengers, and It might be wise to advise ATC that you are not longer RVMS compliant (even if you legally were).

                      Now, if you are in an A330 (in normal or alternate law), things would be much easier for the pilots than in a 747. Unless it was some of those pilots...
                      I think you are all missing one vital point here: it is the transition from AP to manual flight that seems to be a catalyst in these accidents, or the situations where manual flight is being combined with automated flight, such as manual flight with autothrottle. Gabriel, your anedote about the 737 pilot is a good example. The pilot is very focused because he has been in manual flight all along. But when transitioning, especially when it is unexpected, it seems that there is the occasional moment of confusion, disorientation, lack of situational awareness and stealth factor that lead to pilot error. I believe that this is why there are memory items for unexpected AP failure. Is this issue being discussed enough in flight training?

                      Comment


                      • Originally posted by MCM View Post
                        Gabriel,

                        I was comparing the ops to before the age of extensive autopilot usage. The 747/Concorde are "modern" aircraft in that regard. That said, there is a lot to be said for autopilots that don't always get it quite right. They do the hard work, but the pilot needs to stay alert to catch when it decides to do something a little bit wrong. Not sure it would be politically correct to say we should build in deliberate errors to autopilots though .

                        If the autopilot goes inop half way across the pacific, no, you are not in huge danger. But speak to anyone who has had to hand-fly a sector in a 'modern' aircraft, and they'll tell you just how much hard work it is. So much hard work that the pilot who is flying the aircraft is pretty much having to concentrate on that, and not much else. It very much restricts the available brain capacity to handle other issues.

                        The point is though, that if we routinely did that kind of activity, there would be an incident rate higher than what we have with autopilot usage.
                        so the question then is, how much time is too much time to avoid fatigue. clearly, at least in my mind, pilots need more time with stick and rudder. so on a 2 hour flight is 1/2 hour too much? on a 5 hour flight, is 1 hour too much?

                        evan's last comment on the transition being the problem may have some validity, but so too does the proposition that dealing with the variances of even normal flight at cruise is not something pilots are used to handling, so their reactions may be delayed or exaggerated.

                        my friend was bitching because he had to fly his challenger 300 for two hours by hand after ap failure immediately following takeoff. he had no choice since the plane would have to have been flown in for service anyway.

                        Comment


                        • I perform reliability analyses for aerospace systems. But recently I started looking into the feasability of including human reliability in my analyses. Here is a graph from Dhillon's book "Human Reliability."

                          It shows that, in general, for high stress situations human efficiency is degraded. I know from my human factors training that humans have to regulary practice at high stress levels to improve eficiency (e.g. race car driver response time, police target selection and shooting situations, athletes under public pressure to win, etc).

                          I have not studied this topic sufficiently to comment further; but I though this was of interest in regard to the AF447 situation.

                          Comment


                          • Originally posted by Highkeas View Post
                            I perform reliability analyses for aerospace systems. But recently I started looking into the feasability of including human reliability in my analyses. Here is a graph from Dhillon's book "Human Reliability."

                            It shows that, in general, for high stress situations human efficiency is degraded. I know from my human factors training that humans have to regulary practice at high stress levels to improve eficiency (e.g. race car driver response time, police target selection and shooting situations, athletes under public pressure to win, etc).

                            I have not studied this topic sufficiently to comment further; but I though this was of interest in regard to the AF447 situation.
                            interesting...it is no secret that police are very inaccurate in their shooting in real life situations. while they may be master marksmen at the range, when they shoot a real bad guys in real situations, their accuracy is crap. an old friend of mine was a NYC cop with over ten years on the job. at the range he was dead-on-the-balls accurate. one day, while on foot patrol, he startled a drug dealer coming out of a building. without provocation, the drug dealer opens up fire on him hitting him twice. my friend of course returns fire, emptying his 16 round clip and his backup 38 and never touched the guy. only after the fight ended did my friend realize he had been shot.

                            the problem with police training, though i admit they cannot train in the real world, is that it is all staged--stationary targets for 99% of their training and a few "real-life" type ranges where they still shoot at paper targets, albeit ones that move.

                            of course, pilots should not train in real life emergencies in a real aircraft and of course the sim will have to suffice for that. however, i still believe pilots should have more stick-n-rudder time than they currently have.

                            Comment


                            • Just seen on, er... bloomberg.com:
                              Air France Crash Probe Shows Crew Errors



                              Air France Flight 447’s crew reacted badly to an autopilot shutdown and misread instruments showing the plane’s rapid descent before it plunged into the Atlantic, killing all 228 people aboard, a report shows.
                              “I’ve lost VSI,” the junior co-pilot said of the Airbus’s vertical-speed indicator, according to a recording detailed in the report from court-appointed experts. In fact, the instrument was functioning normally, its analog needle immobilized at the lower limit because the plane was hurtling toward the ocean at 15,000 feet a minute, the document seen by Bloomberg News shows.

                              Comment


                              • Shocking new report out with the recording of CVR

                                The exchange is from the Cockpit Voice Recorder (CVR) on Flight 447, which went down in a tropical storm with the loss of 228 lives while flying from Rio de Janeiro in June 2009.

                                Comment

                                Working...
                                X