Air France 447 - Interesting article
#51
Gets Weekends Off
Joined APC: Dec 2005
Posts: 8,883
A video does not show sensations and the report, while good, missed many aspects that explain why they did what they did. I realize it makes us feel good to think we would not have done that, but sadly if you believe that, we have done nothing to prevent a repeat aside from install better pitots. Sad.
I'm sorry to be harsh. I've been flying bit airplanes for over 30 years, as well as investigating accidents. I have participated in evaluating aircraft performance, human performance, operations and numerous other factors. I have been a check airman and instructor on several types of large transport airplanes, worked in both management and in ALPA positions. I have spent the past 20 years also working on handling quality issues and researching based on the most recent understandings of human factors, including control feedback theory and cognitive factors.
I used to be like many of the respondents in this thread. I have learned better. We are all too quick to blame our fellow pilots as being weak. We also, in conjunction, believe that humans are the "weak point" in aviation safety, and also like to tell laypeople that "flying is easy". Then we are shocked when people tell us we're overpaid and not necessary.
The evidence, on the contrary, is overwhelming, that pilots are what keep flying safe, but we are now encountering areas for which we are not training pilots. That needs to be fixed. I have presented several examples here, and there are a lot more. We have not fixed exposure to microbursts really (although we are somewhat protected by MIT's LL algorithm at 67 U.S. airports and HKG), as a quick additional example.
I used to be like many of the respondents in this thread. I have learned better. We are all too quick to blame our fellow pilots as being weak. We also, in conjunction, believe that humans are the "weak point" in aviation safety, and also like to tell laypeople that "flying is easy". Then we are shocked when people tell us we're overpaid and not necessary.
The evidence, on the contrary, is overwhelming, that pilots are what keep flying safe, but we are now encountering areas for which we are not training pilots. That needs to be fixed. I have presented several examples here, and there are a lot more. We have not fixed exposure to microbursts really (although we are somewhat protected by MIT's LL algorithm at 67 U.S. airports and HKG), as a quick additional example.
I agree with you that our training regime needs to be changed to reflect the reality of today's environment of high automation, all glass cockpits, with very little hand flying in reality. That should require a full revamp of the tired old Proficiency Check standards. Also, we should have taught stalls from day 1 to be immediately nose down to reduce the AOA and add power. Instead of the BS I learned in 2007 at a regional (to be fair, all airlines back then) to power out of a stall by adding power, spoilers in, and ride the stick shaker. In the landing configuration that required slight back pressure. It was ridiculous. Now this has been fixed. Sims are also slowly being fully certified to represent a true deep stall in the high Flight Levels.
#52
I know you are very experienced and an author. But on that same token of 30+ yrs experience and accident investigation work, keep in mind that experts in their own fields wrote that report which you say missed the mark (even if good). Those are multiple groups that investigated thoroughly AF447 and came up with their report. I'll be honest that I did not read the entire report, but I read many pages of it and I thought it was a great report.
I agree with you that our training regime needs to be changed to reflect the reality of today's environment of high automation, all glass cockpits, with very little hand flying in reality. That should require a full revamp of the tired old Proficiency Check standards. Also, we should have taught stalls from day 1 to be immediately nose down to reduce the AOA and add power. Instead of the BS I learned in 2007 at a regional (to be fair, all airlines back then) to power out of a stall by adding power, spoilers in, and ride the stick shaker. In the landing configuration that required slight back pressure. It was ridiculous. Now this has been fixed. Sims are also slowly being fully certified to represent a true deep stall in the high Flight Levels.
I agree with you that our training regime needs to be changed to reflect the reality of today's environment of high automation, all glass cockpits, with very little hand flying in reality. That should require a full revamp of the tired old Proficiency Check standards. Also, we should have taught stalls from day 1 to be immediately nose down to reduce the AOA and add power. Instead of the BS I learned in 2007 at a regional (to be fair, all airlines back then) to power out of a stall by adding power, spoilers in, and ride the stick shaker. In the landing configuration that required slight back pressure. It was ridiculous. Now this has been fixed. Sims are also slowly being fully certified to represent a true deep stall in the high Flight Levels.
Last edited by ptarmigan; 09-05-2017 at 05:23 PM.
#53
Gets Weekends Off
Joined APC: Apr 2011
Posts: 1,433
The truth remains, they were weak. Or at the very least, unprepared. A component supplying critical fight data failed. With that went the aircraft's ability to adequately monitor & control itself. The flight crew, unable to "sort things out," rode her down 38,000 feet (16 degrees nose up) into the abyss. Aeroperu 603, hauntingly similar, something we've chewed over more than 20 years. Asiana 214, right into a sea wall, day VFR, due to, what do you even call that? These machines are, highly automated. They are not, fail safe. If you're saying we need to better incorporate training for descending the ladder of automation , fine, I agree. Fly the airplane, hard to argue with that.
#54
The truth remains, they were weak. Or at the very least, unprepared. A component supplying critical fight data failed. With that went the aircraft's ability to adequately monitor & control itself. The flight crew, unable to "sort things out," rode her down 38,000 feet (16 degrees nose up) into the abyss. Aeroperu 603, hauntingly similar, something we've chewed over more than 20 years. Asiana 214, right into a sea wall, day VFR, due to, what do you even call that? These machines are, highly automated. They are not, fail safe. If you're saying we need to better incorporate training for descending the ladder of automation , fine, I agree. Fly the airplane, hard to argue with that.
#55
Gets Weekends Off
Joined APC: Apr 2011
Posts: 1,433
Will have to take your word for it. I can document numerous highly qualified & experienced professionals, on these pages, who don't share yours.
#56
I suggest people read Dekker's "A Field Guide to Human Error" for a start. The other part, though, is that people do not know what they don't know, and hindsight bias and fundamental attribution error reign supreme with many.
#57
Gets Weekends Off
Joined APC: Apr 2011
Posts: 1,433
I submit they did know. But for inability to shed a crippled FMS under extreme stress, couldn't get there soon enough. You worked in training, correct? Tell me you haven't observed crews in recurrent cling to flight director commands during stick shaker, unusual attitude or loss of situational awareness situations when they freely admit at de-brief they "knew" it couldn't help.
Came across a report from what I seem to recall was a Delta PI, many years ago. In it, he detailed using a raw data, hand flown, unfamiliar, non-precision IAP to minimums at a special airport (MGGT?) for both 757 & DC9 crews. As might be expected, while nearly all of the DC9 approaches resulted in a landing on the first attempt, a surprisingly high number of the 75 crews required another after going missed at some point. When Al Ueltschi traded his white cap at PAA for climbing the stairs to begin work at a characteristically Spartan hangar desk, making FlightSafety his legacy, he well understood the value in accurately & reliably prioritizing fundamentals of aircraft operations. Saved a lot of lives in the effort.
Came across a report from what I seem to recall was a Delta PI, many years ago. In it, he detailed using a raw data, hand flown, unfamiliar, non-precision IAP to minimums at a special airport (MGGT?) for both 757 & DC9 crews. As might be expected, while nearly all of the DC9 approaches resulted in a landing on the first attempt, a surprisingly high number of the 75 crews required another after going missed at some point. When Al Ueltschi traded his white cap at PAA for climbing the stairs to begin work at a characteristically Spartan hangar desk, making FlightSafety his legacy, he well understood the value in accurately & reliably prioritizing fundamentals of aircraft operations. Saved a lot of lives in the effort.
#58
Gets Weekends Off
Joined APC: Jul 2016
Posts: 463
The truth remains, they were weak. Or at the very least, unprepared. A component supplying critical fight data failed. With that went the aircraft's ability to adequately monitor & control itself. The flight crew, unable to "sort things out," rode her down 38,000 feet (16 degrees nose up) into the abyss. Aeroperu 603, hauntingly similar, something we've chewed over more than 20 years. Asiana 214, right into a sea wall, day VFR, due to, what do you even call that? These machines are, highly automated. They are not, fail safe. If you're saying we need to better incorporate training for descending the ladder of automation , fine, I agree. Fly the airplane, hard to argue with that.
The solution is somewhere in between: without forgoing the basic hand flying pitch/power (and rudder) skills, address stalls and unusual situations in automation/ glass cockpits. One of the best demos I had while training on the king air involved my instructor having me put the autopilot into a left hand hold, had me close my eyes and when I opened them the airspeed was right above stall. Was I to instinctively add power and put it in a spin? Or was I to just relax the wings and lower the nose. We need these outside the box training scenarios perhaps building on accidents from the past. Perhaps 777 training should involve correcting a low airspeed/below glidepath situation going into SFO for example?
#59
Gets Weekends Off
Joined APC: Apr 2011
Posts: 1,433
We need to do better.
One of the difficulties facing industry & accident investigators in particular, seems to me anyway, is pressure to produce big answers when the root cause of a big tragedy may reveal itself uncomfortably obvious or uncomplicated. As redundant & user friendly all the dreamy tech on new flight decks may be, it can and will fail somebody, somewhere, eventually.
#60
Gets Weekends Off
Joined APC: Sep 2007
Posts: 279
Wait, ok, so...airplane is plummeting towards the ground, radar power, in the bars, still plummeting. We need to have a symposium to decide whether or not, you know, maybe they should be mildly criticized for not at least trying out lowering the nose? Is this for real? Is Alan Fundt hiding in the bushes?
Thread
Thread Starter
Forum
Replies
Last Post
vagabond
Hangar Talk
0
02-20-2007 02:16 PM