News:

Precision Simulator update 10.173 (24 February 2024) is now available.
Navburo update 13 (23 November 2022) is now available.
NG FMC and More is released.

Main Menu

Asiana 777 crashes in San Francisco

Started by Phil Bunch, Sat, 6 Jul 2013 23:28

Jeroen Hoppenbrouwers

As far as I know, there are two major certification organisations worldwide, FAA and EASA the European equivalent. They do recognize each other's approvals with minimal paperwork and it depends on who wants to certify what, where you go first. If you certify with another organsation it may be more work to get the big guys to accept your material, but it will always help.


Hoppie

IefCooreman

#81
Took a while, because pretty long :-)

Farrokh747:
The misconception results from people thinking the autothrottle wakeup is the only stall protection on a 777. In a descent, in FLCH and VNAV SPD modes where the autothrottle has changed to HOLD, the autothrottle wakeup will not activate because the philosophy of these modes is that pitch governs the speed. The 777 however also has a "stall protection" in pitch. The fly-by-wire disables the speed trim system once you really slow down into the yellow band. So the pilot is forced into pulling the stick if he really wants the nose high attitude, or the nose will drop down (so, yes, the Asiana pilot was pulling the stick in the end, fixating outside on the runway). The mistake they made is you shouldn't be in FLCH that close to the ground.

To go back to the answers of Will and SwissCharles. I agree up to a certain level (although to me it becomes a question of attitude and thrust, you don't need a speed indicator) however statistics show that the idea that the well trained pilot should be the ultimate last protection is "incomplete" and solely does not allow for safe aviation. The reason is simple: there is hardly any backup for you. And this is where the focus of training has changed. You don't allow anymore for such situations. Murphy never shows up with 1 single problem... And there is one certainty: we are human, we all make mistakes.

Old simulator training habits (at least from my experience): failing as much as possible until the pilot was left with only himself. If he landed, job wel done. Modern training has completely abandoned this approach. In modern training you could call "threat" a keyword. From the moment you enter the crewroom, you look for threats, and you pre-build defences. This allows for mistakes or problems to happen, and you can deal with them in a safe -unrushed- way. The airplanes we fly are very automated, which results in VERY safe normal flights. But there are many many layers in all the systems. Our job becomes one of checking and working with different layers, assessing and reviewing the possibilities if one layer becomes inoperative. Use of automation is one of these layers, because monitoring is much easier than executing (overview vs focussed). Modern simulator checks are filled with "calm" compared to old school checks. The emphasis is about gathering information, showing you know your aircraft capabilities, your own capabilities, creating safety, creating safe options, choosing, reviewing if anything changes,...

Downgraded situations do not make the flight unsafe, it only forces us to review the possibilities the aircraft-crew combination is giving the crew. Is ETOPS still safe? MNPS? RVSM? High altitude destination still an option? A tiny downgrade can have big consequences if you don't consider it. Even in time-critical situations (smoke, fire, low fuel) you need to choose the path of the "least threats for the available time".

This does not stop you from flying manual. But you choose the right circumstances; situations where there are no extra threats. Boeing allows you to cut out automation and continue to fly manual, when you consider the situation acceptable. Boeing officially also recommends standard use of automation for visual approaches as long as feasible. The big thing here is: it is YOU as a pilot who decides. It should NOT be ATC. If you are tired or stressed (threat), Boeing would rather see you create a visual reference point on final in the FMC, and using LNAV/VNAV for the approach. On the long haul, you don't have the chance to do a lot of visuals. When an ILS is inop, the experience of both pilots will be discussed. When one has done a visual recently and feels up for it, a visual is an option. If there is no experience, a visual simply becomes a threat. This way of thinking is more important than just being able to do it. I might as well quote one of my first bosses: "anyone who elects to fly a manual circle-to-land after an engine failure and successfully lands, I will invite personally to my office and it will be the last visit he makes"

Now you could say, more training is required then. Every three months extra simulator sessions with the focus on very basic flying. I agree completely. However, this is a money thing... for the moment regulations define minimum training requirements, and we deal with it. On the other hand, as a pilot I already feel "overly checked", simulator every six months, theory courses, line check, medical,... And I think for the moment we have come a long way in aviation, flying has become extremely safe. And training still evolves as well. We still make mistakes, but we catch them a lot more, earlier, and create options that are much easier to deal with.

The discussion today: crashes like the AF out of Rio, the Asiana in SFO and the Turkish in Amsterdam, show that despite having an extra layer of automation, pilots have difficulty in stressfull situations of interpreting what is happening exactly. The philosophy of the "automation user-interface" today is still one where information is not always clear to the pilots. The pilot of the AF kept pulling the sidestick back, because he was trained to do so for stall recoveries and the real important messages of the aircraft got lost in the overflow of other systems failing as well. The Asiana was in training and unaware of his autothrottle functionality, hence he was not as protected as he probably thought he'd be, and so was the Turkish crew fooled by a failing radio altimeter. And all of them reacted in the end, but too late.

Jeroen:
My "problem" with American ATC (although I admit, it's pretty much "approach" frequencies) is that they have no consideration at all for your situation as a pilot. You should be able to do what they tell you. They are sitting behind a desk, I can accept that at times they have big problems to solve (everybody needs to get in, everybody needs to get out), and therefore they expect you to do exactly as told. But it's a "no errors allowed" environment, so from my viewpoint: unsafe. And they will use the "visual" rules to make their own life easy, but it makes our life harder – more threats. They haven't evolved. There are still too many "chaos" situations you hardly ever find on other major airports. And it's chaos because humans try to do everything too much "anticipating", not "prepared". On the long haul you can offer flexibility, but you require more time to prepare. New York never has time. But that's only my feeling.

Anyway, these are only my own thoughts on the subject(s). Time to go out and face my jetlag now :-)

farrokh747


Jeroen Hoppenbrouwers

I can attest that in those couple of cases I found myself up front in a light jet aircraft in the US, ATC has been "interesting" during the final 5-10 minutes. I thought they just liked to drive the sports car hard, but apparently it is more like don't-care? Directing you straight towards the field until 5 nm out at 4000 ft and then declaring you cleared for the visual. It's that we expected it, but else ... barn doors out ... The other fun one was being about 30 nm from Memphis still at FL450 to keep us on top of the heavies and then having us shoot the bull's eye. My ears still hurt.


Hoppie

Richard McDonald Woods

Very interesting how we (don't) learn from all the combined experiences of human-human and human-machine interactions.
Cheers, Richard

Will

I talked to an approach controller who told me once that she loved turboprops, because they could scream in at 250 kts until short final and then slow down to approach speed in something like two miles, whereas the heavies had to make more subtle changes. (Oh, so that's why the ride was always so thrilling...)
Will /Chicago /USA

Jeroen D


IefCooreman

Almost out of jetlag... think I need a character counter or limiter in the reply box...

farrokh747


Jeroen Hoppenbrouwers


jb747

That'll keep several folks employed for many years to come...

Jeroen Hoppenbrouwers

... to say the least ... what a list!