NTSB update on Asiana 214
Join Date: Jan 2000
Location: FL410
Posts: 860
Likes: 0
Received 0 Likes
on
0 Posts
Improper use of automatics...
What I still do not understand, is why they didn't fly a LOC approach or VNAV as only G/S u/s if they were so concerned about a visual approach and its perceived problems.
But having set the Missed Approach Altitude to 3000', every Boeing pilot knows that pressing FLCH will cause the aircraft to climb up, not go down. Basic school boy error, like hitting TOGA instead of A/T disconnect as sometimes seen on another Boeing type due incorrect throttle positioning of the hands. Time and height permitting if can be corrected with visual references: continue, otherwise the best course of action is to continue that climb and try again.
What I still do not understand, is why they didn't fly a LOC approach or VNAV as only G/S u/s if they were so concerned about a visual approach and its perceived problems.
But having set the Missed Approach Altitude to 3000', every Boeing pilot knows that pressing FLCH will cause the aircraft to climb up, not go down. Basic school boy error, like hitting TOGA instead of A/T disconnect as sometimes seen on another Boeing type due incorrect throttle positioning of the hands. Time and height permitting if can be corrected with visual references: continue, otherwise the best course of action is to continue that climb and try again.
Join Date: Jul 2001
Location: Nashville
Posts: 72
Likes: 0
Received 0 Likes
on
0 Posts
why they didn't fly a LOC approach or VNAV as only G/S u/s if they were so concerned about a visual approach and its perceived problems.
Further unexplained is since the 777 uses back-driven thrust levers and they were not moving -- sitting at the idle detent (where the pilot put them) why the surprise over low thrust?
It's ironic this accident is commonly attributed to "automation dependence", when in fact the pilot did not understand how the automation worked. He wasn't over-dependent on automation which failed, rather he didn't understand the automation, plus failed to use available automation modes, plus couldn't hand fly a visual approach. That was not a good combination.
Join Date: Apr 2002
Location: MNL ex CCR ex CLE
Age: 65
Posts: 201
Likes: 0
Received 0 Likes
on
0 Posts
" Cory Lidle's widow was slapped with $80,000 judgement to repay Cirrus' court fees after losing the battle against them for $50 mln. Her lawyer called it a "double injustice". Of course the $80 K is a relative chump change compared to what was in stake in this trial and what lawyers' fees were. "
....and even though Cirrus 'won'....how much did defending their company ultimately cost them - and the consumers? Because sure as God made little green apples, there ain't ANY company that doesn't pass along costs to US. The sad fact is that most Americans consider the civil 'justice' system to be the same as playing the lottery. Most often, the plaintiff pays NOTHING and the 'lawyers' get paid from whatever judgement is won. Fortunately, some states have begun changing - such as Texas - where, at least in medical malpractice cases - it's now 'loser pays'. Prior to this change, many so called high risk medical practicioners left Texas - obstetricians, anethesiologists, neuro-surgeons, etc. AFTER the change, many have returned and many more have moved to Texas. A little know fact, is that aLOT of medical costs are because of 'defensive medicine' practices, which is just CYA for physicians and hospitals - NOT because it's clinically needed. Unfortunately, as long as the American trial layers association keeps pumping money into political campaigns, it's very unlikely this will change elsewhere anytime soon.
....and even though Cirrus 'won'....how much did defending their company ultimately cost them - and the consumers? Because sure as God made little green apples, there ain't ANY company that doesn't pass along costs to US. The sad fact is that most Americans consider the civil 'justice' system to be the same as playing the lottery. Most often, the plaintiff pays NOTHING and the 'lawyers' get paid from whatever judgement is won. Fortunately, some states have begun changing - such as Texas - where, at least in medical malpractice cases - it's now 'loser pays'. Prior to this change, many so called high risk medical practicioners left Texas - obstetricians, anethesiologists, neuro-surgeons, etc. AFTER the change, many have returned and many more have moved to Texas. A little know fact, is that aLOT of medical costs are because of 'defensive medicine' practices, which is just CYA for physicians and hospitals - NOT because it's clinically needed. Unfortunately, as long as the American trial layers association keeps pumping money into political campaigns, it's very unlikely this will change elsewhere anytime soon.
Join Date: Jun 2011
Location: france
Posts: 760
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by joema
It's ironic this accident is commonly attributed to "automation dependence", when in fact the pilot did not understand how theautomation worked. He wasn't over-dependent on automation which failed, rather he didn't understand theautomation, plus failed to use available automation modes, plus couldn't hand fly a visual approach
Join Date: Apr 2009
Location: Oakland, CA
Age: 72
Posts: 427
Likes: 0
Received 0 Likes
on
0 Posts
He wasn't over-dependent on automation
Join Date: Nov 2009
Location: Somewhere
Posts: 73
Likes: 0
Received 0 Likes
on
0 Posts
I would argue this was still the case of rather chronic automation dependency. Who said that you can't be automation dependent while at the same time poorly understand the same automation.
Join Date: Jan 2008
Location: United Kingdom
Posts: 274
Likes: 0
Received 0 Likes
on
0 Posts
If anyone wonders why 777 pilots' basic hand flying skills are not to a high standard consider the following when manually flying the 777.
1) When you put on power there is no pitch up couple, this is taken out automatically
2) The autothrottle is always engaged and pilots are not allowed to disconnect it. This destroys the basic instinct of taking off or putting on power when levelling off after a climb or descent.
3) If there is an engine failure rudder is applied automatically.
4) Pilots are encouraged to use a flight path vector which makes flying manually a bit easier.
5) There is no pitch change when gear or flap is selected.
6) If you overbank then the automatics chip in and roll the wings back towards level.
7) When you roll into a turn there is no need to apply back elevator pressure.
The only time 777 pilots genuinely hand fly the aircraft is when it is in "Direct Mode". This takes out all the aids mentioned above apart from the autothrottle. However this is only ever practiced in a simulator.
As 777 sectors are typically 8 hours long or more, and are frequently overnight, hand flying in to a busy airport is not recommended. The fatigue factor which the Asiana pilots undoubtedly experienced has not been discussed much and I believe it would have been a significant factor. Although I frequently hand fly the 777 I strenuously avoid it after overnight flights apart from disconnecting the autopilot below 1000 feet and hand flying the remainder of the approach to landing.
In no way am I saying that what happened in SFO was anything other than pilot error, but I am just explaining to those who have not flown a 777 how it has been made relatively easy to hand fly using artificial means, but consequently basic flying skills can be eroded by those same artificial means.
Does anyone remember the SAS DC10 accident at JFK about 1984 when it overran into the water? I seem to remember over reliance on the autothrottle system was cited as a cause.
1) When you put on power there is no pitch up couple, this is taken out automatically
2) The autothrottle is always engaged and pilots are not allowed to disconnect it. This destroys the basic instinct of taking off or putting on power when levelling off after a climb or descent.
3) If there is an engine failure rudder is applied automatically.
4) Pilots are encouraged to use a flight path vector which makes flying manually a bit easier.
5) There is no pitch change when gear or flap is selected.
6) If you overbank then the automatics chip in and roll the wings back towards level.
7) When you roll into a turn there is no need to apply back elevator pressure.
The only time 777 pilots genuinely hand fly the aircraft is when it is in "Direct Mode". This takes out all the aids mentioned above apart from the autothrottle. However this is only ever practiced in a simulator.
As 777 sectors are typically 8 hours long or more, and are frequently overnight, hand flying in to a busy airport is not recommended. The fatigue factor which the Asiana pilots undoubtedly experienced has not been discussed much and I believe it would have been a significant factor. Although I frequently hand fly the 777 I strenuously avoid it after overnight flights apart from disconnecting the autopilot below 1000 feet and hand flying the remainder of the approach to landing.
In no way am I saying that what happened in SFO was anything other than pilot error, but I am just explaining to those who have not flown a 777 how it has been made relatively easy to hand fly using artificial means, but consequently basic flying skills can be eroded by those same artificial means.
Does anyone remember the SAS DC10 accident at JFK about 1984 when it overran into the water? I seem to remember over reliance on the autothrottle system was cited as a cause.
Per Ardua ad Astraeus
Join Date: Mar 2000
Location: UK
Posts: 18,579
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by sime
The fatigue factor which the Asiana pilots undoubtedly experienced
Join Date: Aug 2013
Location: chicago
Posts: 359
Likes: 0
Received 0 Likes
on
0 Posts
suninmyeyes
are you thinking of the world airways DC10 at Boston? Autothrottles were part of it,even the engine kept running after it went into the water.
here is the report: http://www.airdisaster.com/reports/ntsb/AAR85-06.pdf
while I understand everything you have written about the 777, I also know that if you only take the elevator between floors on a skyscraper (for british folk, THE LIFT on a tall building), then if the elevator fails you will not be in shape to take the stairs.
dear pilots, take the stairs!
are you thinking of the world airways DC10 at Boston? Autothrottles were part of it,even the engine kept running after it went into the water.
here is the report: http://www.airdisaster.com/reports/ntsb/AAR85-06.pdf
while I understand everything you have written about the 777, I also know that if you only take the elevator between floors on a skyscraper (for british folk, THE LIFT on a tall building), then if the elevator fails you will not be in shape to take the stairs.
dear pilots, take the stairs!
2) The autothrottle is always engaged and pilots are not allowed to disconnect it. This destroys the basic instinct of taking off or putting on power when levelling off after a climb or descent.
Join Date: Jan 2008
Location: United Kingdom
Posts: 274
Likes: 0
Received 0 Likes
on
0 Posts
Suninmyeys, there is a A/T "DISARM" switch on the forward panel. Flip that switch to 'disarm' and the A/T is taken completely out of the loop. Does your airline prohibit using that switch?
BOAC fair point, tired then.
Join Date: Jan 2006
Location: US
Posts: 2,205
Likes: 0
Received 0 Likes
on
0 Posts
Then that's a COMPANY problem and not an AIRPLANE issue.
IMO you don't become a better pilot on the 777. You retain your previous skill set. IMO putting new airline pilots into long haul fleets will restrict, or even prevent, the development of decent piloting skills.
IMO you don't become a better pilot on the 777. You retain your previous skill set. IMO putting new airline pilots into long haul fleets will restrict, or even prevent, the development of decent piloting skills.
Join Date: Jul 2001
Location: Nashville
Posts: 72
Likes: 0
Received 0 Likes
on
0 Posts
Can't agree. I would argue this was still the case of rather chronic automation dependency
Rather, he didn't understand the automation behavior. Had he been a stick-and-rudder guru, he could have still crashed by not understanding that.
The A320 pilot at the Habsheim crash was a highly experienced test pilot with excellent hand flying skills. Yet a misunderstanding about automation combined with poor decision making caused a crash.
Likewise, further improving the hand flying skills of the Asiana pilot -- by itself -- might not have prevented the problem if he *still* misunderstood aspects of the automation. This implies the overall problem and solution is broader than automation dependency.
joe, this takes one back to a point raised during the prolonged discussion on AF447: how well do you know your aircraft, and how well do you know your aircraft systems and sub systems?
Join Date: Apr 2005
Location: Southeast USA
Posts: 801
Likes: 0
Received 0 Likes
on
0 Posts
Hello stator vane:
My compliments on one of the most cogent, accurate, and professional posts on this thread. I think your comment …”so that is why I am always right there with knowing what I want it to do and making sure the airplane does what I want” ... should be tattooed on the inside of every pilot’s eyelids – so that is what they see every time they blink!
I posted a comment earlier on this thread where I described a comment I’ve often made to students … but I’m not sure that your version of it isn’t more professional … What I posted says … “I’ve often told relatively new pilots that if you remove the A/P Engage switch-light cover, and look on the back of it, it DOES NOT SAY – Pilot Disengage.
My compliments on one of the most cogent, accurate, and professional posts on this thread. I think your comment …”so that is why I am always right there with knowing what I want it to do and making sure the airplane does what I want” ... should be tattooed on the inside of every pilot’s eyelids – so that is what they see every time they blink!
I posted a comment earlier on this thread where I described a comment I’ve often made to students … but I’m not sure that your version of it isn’t more professional … What I posted says … “I’ve often told relatively new pilots that if you remove the A/P Engage switch-light cover, and look on the back of it, it DOES NOT SAY – Pilot Disengage.
Join Date: Feb 2008
Location: Europe
Posts: 80
Likes: 0
Received 0 Likes
on
0 Posts
Sounds to me like the similar "automation dependence" on electronic calculators which crept into the classroom.
I found that some classmates were quite adept at bashing calculator keys parrot-fashion and coming up with the right answers, but were stumped if they had to solve the same problem by hand.
In other words, they didn't understand the reason mathematics worked. They didn't know what the calculator was doing, and why - only that it worked, as if by magic.
It's not enough to teach pilots the procedures of accessing modes on aircraft. You have to teach them why you're accessing those modes. And you have to teach them not just how it responds, but why.
I found that some classmates were quite adept at bashing calculator keys parrot-fashion and coming up with the right answers, but were stumped if they had to solve the same problem by hand.
In other words, they didn't understand the reason mathematics worked. They didn't know what the calculator was doing, and why - only that it worked, as if by magic.
It's not enough to teach pilots the procedures of accessing modes on aircraft. You have to teach them why you're accessing those modes. And you have to teach them not just how it responds, but why.
Join Date: Apr 2003
Location: Germany
Age: 76
Posts: 1,561
Likes: 0
Received 0 Likes
on
0 Posts
This is a growing problem...
I came up the usual "self-improver" way. I did a year flying out of Miami to some rather far-off islands using light twins without autopilots. You wouldn't want to fly into a busy terminal area doing that, so that a lot of the necessary skills for flying the line are not learned, but you certainly do master hand-flying and keeping your scan going.
Years later, I ended up flying a Dornier 328, having to learn modern systems I had only read about. (The 328 has Honeywell Primus 2000 glass, but no autothrottle.) There was a period of "What's it doing now?" but at least that was underpinned by the basic ability to revert to primary cues and fly the aircraft when the automation had been told to do something I did not want it to, such as taking us to the wrong waypoint. or ignoring the localizer because I had not armed APP.
The insidious thing is that hand-flying skills deteriorate if they are not used; I would take any good chance to practice, when the young FOs would groan when they heard that "cavalry charge" warning tone. Of course, later I could watch some of them really struggle in a sim session when one of the first systems to fail was the autopilot; they had obviously poor hand-flying skills. Some had probably never learned to do that very well, having been carried along through one of our infamous FAA-approved training mills, while others had simply not kept that skill up.
When it came to other things such as following a SID or a STAR then they were often superior to me; I was still essentially just a bush pilot in some ways, but which skill was primary? You need both skill sets, of course, but only one, being able to hand-fly without undue effort, is going to keep that thing dependably shiny side up!
A related problem, addressed on other threads, is the way that stall training has been perverted into an exercise in not losing any altitude. If you go in at 5 thousand feet, then you have to come out at that, even if that means sort of mushing along instead of making a very positive nose-down recovery from the stall. Well, usually not even a stall, of course, just the approach to a stall. It's easy to see how a real, full-breaking stall would come as a very nasty shock to someone only used to getting a bit of buffet or the stick shaker, and thinking that's the same as a stall flown to the break.
All of this seems like swaddling the flight crew in fluffy cotton, keeping them well away from the sharp edges implicit in flying any aircraft at the edge of its envelope, whether that's a critically low speed on final, or an upset in turbulence, or a stall, perhaps one encountered in icing at a speed taken to be above stall speed. I think that problem is spread across much of aviation now; I don't think it's specific to the culture of Korea. What happens as trouble approaches, whether good CRM is practiced, that might be culturally specific, but this lack of hand-flying skill is widespread, I think.
Years later, I ended up flying a Dornier 328, having to learn modern systems I had only read about. (The 328 has Honeywell Primus 2000 glass, but no autothrottle.) There was a period of "What's it doing now?" but at least that was underpinned by the basic ability to revert to primary cues and fly the aircraft when the automation had been told to do something I did not want it to, such as taking us to the wrong waypoint. or ignoring the localizer because I had not armed APP.
The insidious thing is that hand-flying skills deteriorate if they are not used; I would take any good chance to practice, when the young FOs would groan when they heard that "cavalry charge" warning tone. Of course, later I could watch some of them really struggle in a sim session when one of the first systems to fail was the autopilot; they had obviously poor hand-flying skills. Some had probably never learned to do that very well, having been carried along through one of our infamous FAA-approved training mills, while others had simply not kept that skill up.
When it came to other things such as following a SID or a STAR then they were often superior to me; I was still essentially just a bush pilot in some ways, but which skill was primary? You need both skill sets, of course, but only one, being able to hand-fly without undue effort, is going to keep that thing dependably shiny side up!
A related problem, addressed on other threads, is the way that stall training has been perverted into an exercise in not losing any altitude. If you go in at 5 thousand feet, then you have to come out at that, even if that means sort of mushing along instead of making a very positive nose-down recovery from the stall. Well, usually not even a stall, of course, just the approach to a stall. It's easy to see how a real, full-breaking stall would come as a very nasty shock to someone only used to getting a bit of buffet or the stick shaker, and thinking that's the same as a stall flown to the break.
All of this seems like swaddling the flight crew in fluffy cotton, keeping them well away from the sharp edges implicit in flying any aircraft at the edge of its envelope, whether that's a critically low speed on final, or an upset in turbulence, or a stall, perhaps one encountered in icing at a speed taken to be above stall speed. I think that problem is spread across much of aviation now; I don't think it's specific to the culture of Korea. What happens as trouble approaches, whether good CRM is practiced, that might be culturally specific, but this lack of hand-flying skill is widespread, I think.
Education and training go together. Try telling that to a bean counter in any land.
Join Date: Aug 2013
Location: chicago
Posts: 359
Likes: 0
Received 0 Likes
on
0 Posts
someone said that there were only a couple of options available tothe pilot other than a visual, ie loc or rnav.
other options:
request ILS 28R
I think we have established that 28 right was ok.
Diversions could have been requested too...Oakland INTL , virtually abeam SFO, San Jose, maybe 45 miles away, the former Naval air station, MOFFET Field (now a federal airfield) even closer...all could handle a 777.
So many options.
Imagine a Dark and Stormy night over the pacific, engine quits and MIDWAY ISLAND is the only place open and in range, and further imagine the ILS is OTS.
IF the pilots could retard the thrust levers when the plane started to climb to the Missed apch alt, how could they not know how to advance the thrust levers when the speed got too slow?
And clutched throttles can be pushed forward even if the autothrottle sw is not used.
strength enough to hold 80 lbs of pressure on yoke, but tooo weak to push throttles forward?
other options:
request ILS 28R
I think we have established that 28 right was ok.
Diversions could have been requested too...Oakland INTL , virtually abeam SFO, San Jose, maybe 45 miles away, the former Naval air station, MOFFET Field (now a federal airfield) even closer...all could handle a 777.
So many options.
Imagine a Dark and Stormy night over the pacific, engine quits and MIDWAY ISLAND is the only place open and in range, and further imagine the ILS is OTS.
IF the pilots could retard the thrust levers when the plane started to climb to the Missed apch alt, how could they not know how to advance the thrust levers when the speed got too slow?
And clutched throttles can be pushed forward even if the autothrottle sw is not used.
strength enough to hold 80 lbs of pressure on yoke, but tooo weak to push throttles forward?