Asiana flight crash at San Francisco

[Offcut]

10002level,

On the 777 the auto throttles are normally left on while hand flying. You just need to ensure they are in "speed" mode. The danger is if you go from a non "speed" mode with the autopilot in and then disconnect. Now you are hand flying with the auto throttle turned on, but not operating. No drama if you know this and manually manipulate the thrust levers but this is not the normal state of affairs. If you expected the auto throttle to control your speed but it was in hold mode, and you were overwhelmed with manoeuvring the aircraft from a hot and high position, and you were new on type, and it was the end of a long haul flight.............

The holes all start to line up. Clearly an accident is not an acceptable outcome but the reasons for it are rarely as simple as "the crew stuffed up/are undertrained/incompetent ".

[wasthatit]

The NTSB have confirmed the PAPIs for 28L were operational at the time of the crash. They were subsequently damaged and hence NOTAM'd as inop which may have caused some confusion as to their status.

I am a little surprised that after such a catastrophic accident and a large charred airframe sitting at the side of the runway, that someone filed a NOTAM that the PAPI are OTS. After all they are for a runway that is going to stay closed for a significant period of time.

Presumably there was other infrastructure damage (GP antenna, met sensors, runway edge lights to say the least) but these are not NOTAM'd.

The only reason I can think of is that the PAPI form part of the overall visual picture that crews might expect to see when visually acquiring the airport?

[Raven1972]

"A senior Asiana official said the pilot was Lee Jeong-min, a veteran pilot who has spent his career with the airline. He was among four pilots on the plane who rotated on two-person shifts during the 10-hour flight, the official said"

"Lee Jung Min has flown a total of 12,387 hours, 3,220 on a 777, while Lee Kang Kuk has flown a total of 9,793 hours, 43 of which were on a 777, according to the statement.
Co-pilot Lee Kang Kuk was in the process of transitioning from flying Boeing 737 model to the 777, South Korean Transport Ministry Official Choi Jung Ho told reporters in Seoul today. The government is still investigating who had control of the plane during landing, he said."

[Absolutely Fabulous]

Yeah Right

If the guy in the right seat was PNF, theres no doubt he didnt dare speak up to the Captain.

To late to cover up loss of face now.

Yeah typical of asians affected by authority gradient

Just like the asian F/Os onboard the QF flight in BKK, the SW in Midway and the AA in Kingston, none of the spoke up at all

[413X3]

Those pointing the finger at Jakob Van Zanten miss the point.

After Tenerife, KLM went deeply into the circumstances of the accident, realised there was a problem with the martinet attitude of some of its Captains and came up with the prototype of today's CRM, the KHUFAC course.

So how many hull losses has KLM suffered since 1977? Zero.

That course and its successors in no small way account for the continuing improvement in world accident statistics and could be said to have saved many thousands of lives.

Every airline is one flight away from an accident. The only difference between KLM culture now and the Korean airlines seems to be the willingness to admit to the holes in their particular slice of Swiss cheese and fix them.

But the big difference is, after that crash, nobody said "white Europeans have a culture problem, and lack basic airmanship."

They said "this captain made mistakes" and stopped there.

When the Colgan Q400 went down, nobody said the American culture or lack of skills caused it, they said these particular pilots made mistakes, the end.

See the difference? When you start going into race and general "culture" of races is when people should rightly question you.

[BOAC]

That's confusing! I thought the 'pilot' had 43 hrs on type - ie Kuk, according to other statements. I had 'assumed' (I know....) he was LHS on a command type conversion in which case PF into SFO was a questionable choice.

As usual the use of the word 'pilot' is meaningless, but favoured by journos and others.

[funfly]

From the BBC website:

"San Francisco's coroner is currently trying to establish whether one of the two fatalities occurred after a passenger was run over by an emergency vehicle rushing to the scene of the crash."

[Exascot]

The government is still investigating who had control of the plane during landing, he said.

Interesting statement. There was no control. And, they call that a 'landing'!

I think they mean, 'who flew the approach'.

[Cows getting bigger]

(semi) automation

I read with great interest the inputs from professional pilots who obviously know the systems they operate to great detail. However, I get a little sceptical when it would appear that aircraft systems have developed to such high levels of complexity for something that, in my relative ignorance, seems to be a simple function.

The implication from some of the technical posts is that there are significant number of combinations which can result in various levels of automation and it would be relatively easy for a distracted, tired, confused crew to be on the 'wrong page', as it were. To one who is routinely faced with an autopilot on/off option and not much more, this seems to be a possible contributory factor. Do we really need so many options in order to satisfy the fundamental of Power + Attitude = Performance?

[trauha]

Slope graphic:

http://www.nytimes.com/interactive/2...nt.html?ref=us

[nitpicker330]

The two victims were only ejected ( an one possibly run over later ) from the Aircraft through a breach in the Tail after impact with the sea wall. Didn't happen in BA because they flopped onto smooth flat grass and the tail didn't get knocked off.

[Cows getting bigger]

(semi) automation

I read with great interest the inputs from professional pilots who obviously know the systems they operate to great detail. However, I get a little sceptical when it would appear that aircraft systems have developed to such high levels of complexity for something that, in my relative ignorance, seems to be a simple function.

The implication from some of the technical posts is that there are significant number of combinations which can result in various levels of automation and it would be relatively easy for a distracted/tired/confused/overloaded crew to be on the 'wrong page', as it were. To one who is routinely faced with an autopilot on/off option and not much more, this seems to be a possible contributory factor. Do we really need so many options in order to satisfy the fundamental of Power + Attitude = Performance?

[Mimpe]

Is there a case for standardised displaced thresholds at destinations where there are no safe undershoot options, or where there are difficulties assessing height visualy ( over water)? In this instance, if the aircraft had been 2-3 feet lower, most of passengers would have been lost.

The worst case scenario has (almost) happened.

[James7]

Girl run over

"San Francisco's coroner is currently trying to establish whether one of the two fatalities occurred after a passenger was run over by an emergency vehicle rushing to the scene of the crash."

Very sad indeed, poor girl survives being ejected from a ground spinning aircraft only to be run over by someone else not looking out the window.

Will be interesting to see how long it actually took for the 'responders' to find the deceased and whether immediate first aid could have saved them.

According to one of the eye witnesses the emergency services were pretty chaotic. I know it is a long way for them to travel and sense of time is pretty much lost.

[ohreally]

If the threshold has been moved, is this recent? Would the FMC nav database still have the previous threshold? This obviously should not cause landing short but would contribute to destabilising when low and in a low energy state.

Were the two relief pilots in the cockpit? Wonder what they were thinking?

In addition to thoughts for all on the 777, I feel for those on the aircraft at the hold point. Not nice to witness something so terrifying and so close. Been there.

[Bueno Hombre]

Spooled up and psyched up enough ?

Spooled up and psyched up enough to make a missed approach?
There, but for the grace of God...

[Non Zero]

Is there a case for standardised displaced thresholds at destinations where there are no safe undershoot options, or

where there are difficulties assessing height visualy ( over water)?

What exactly is that you are looking for? It was day VMC condition ...not the scenario for a black-hole approach.

The Black Hole Approach: Don't Get Sucked In!

where there are difficulties assessing height visualy ( over water)?

Even without a PAPI (and it was not the case) you can assess if you are on a correct glide path! I am expecting that for a 9000 hours pilot!



Excuses are acceptable only in case of fatigue!

In this instance, if the aircraft had been 2-3 feet lower, most of passengers would have been lost.

It was a training flight ... the safety margins should be increased ... 2-3 feet lower it would have been 2-3 feet well below the glide than they were already which it was already unacceptable especially if you do not have energy (speed in this case) to counterbalance you distraction.


But considering what kind of stressed stimulations the structure of the 777 went under ... wow she is a pretty safe airplane ...

[clayne]

Originally Posted by Oakape

Machines are not infallible. They sometimes behave in unexpected ways or fail, either partially or completely. Personally I keep my hand on the thrust levers until the flaps are up after take-off & on approach from the first flap selection, regardless of whether or not the autopilot and/or the autothrottle is engaged. I also have my hand on the thrust levers when I anticipate a major thrust change in flight e.g. enroute level change or top of descent. If the thrust levers don't move to where I expect them to be, I put them there. If the autothrottle fights me, I disengage it until I have time to sort the issue out.

This.

There are so many excuses being continually made around lack of G/S, automation traps, etc. as being "contributing" factors. As another poster mentioned some pages back - all they are are factors. At the end of the day, pilots should continuously be in a feedback loop of "trust but verify." This goes for both what the PF or PNF are doing *and* what the automation is doing.

Cultural issues combined with a new generation of magenta pilots completely dependent on automation is what some of the more experienced people have been hammering on as basic points contributing to a decline in aviation safety *when things aren't perfect*. Even when they are perfect, "the auto-pilot/auto-throttle did it!" are never valid excuses.

These points are not ignorable and neither is the lack of a go around much earlier.

Originally Posted by fsfaludi

I find it weird how the "new(er) generation" flight crews have got the basic flying skills so poorly learned and/or miserably handled at times.

Nothing weird about it (although I know you already know this). Talk to the bean-counters/corporate overlords who are hell-bent on turning pilots into human robots. The worst part is the punitive attitude taken towards pilots who *do* think outside the box combined with the policy-based reduction in hand-flying which could save people's lives someday. I doubt we'll be seeing many more Dennis Fitch types if things continue down the road they're on.

[Back at NH]

Oddly enough my guess is from everything I have read about what was just apparently a poorly executed approach and landing, is that it would have been LESS likely to happen in IFR conditions! The crew was likely much more lax in accessing and correcting for an unstabilized approach while visually flying it than had they been flying it on instruments.

I find it weird how the "new(er) generation" flight crews have got the basic flying skills so poorly learned and/or miserably handled at times.

Surely they were in IFR conditions, just not in IMC??

[bnt]

Someone earlier linked to this story, which points out the following:

"It was Lee Kang-kook's maiden flight to the airport with the jet... He was in training. Even a veteran gets training (for a new jet)," a spokeswoman for Asiana Airlines said on Monday.
...
"(Lee Kang-kook) has a lot of experience and previously flown to San Francisco on different planes including the B747... and he was assisted by another pilot who has more experience with the 777," the spokeswoman said.

So I'm less inclined to blame a lack of type experience, not when the same pilot has significant B747 experience.