Asiana B777-200 crash

[Monsta]

Αφού δε το βαλε κανείς, ορίστε. All evidence so far is pointing to rookie 777 pilot.

An Asiana Boeing 777-200, registration HL7742 performing flight OZ-214 from Seoul (South Korea) to San Francisco,CA (USA) with 291 passengers and 16 crew, touched down short of runway 28L impacting the edge separating the runway from the San Francisco Bay 115 meters/375 feet ahead of the runway threshold while landing on San Francisco's runway 28L at 11:27L (18:27Z), the tail plane, gear and engines separated, the aircraft came to a rest left of the runway about 490 meters/1600 feet past the runway threshold. Following first impact at high angle of pitch the aircraft lost gear and tail section, skidded along the runway, pitched up to about 45 degrees, the left wing entangled with ground sending the aircraft in a spin and separating the engines, the aircraft turned around counterclockwise by nearly 360 degrees with the nose coming down again and stopped, burst into flames and burned out, 305 occupants were able to evacuate the aircraft in time and are alive. 2 people are confirmed killed in the accident, 10 people are in critical condition, 38 more are in hospital care with injuries of lesser degrees, 82 occupants received minor injuries. The majority of survivors escaped without injuries.

Emergency services initially reported all occupants have been accounted for and are alive. Emergency services repeated ALL occupants have been accounted for in response to media reports that two people have been killed. A number of people were taken to hospitals with injuries of varying degrees. In a later press conference the fire chief of San Francisco said, not all people have been accounted for, two people were confirmed killed in the accident. In a second press conference Saturday evening (Jul 6th San Francisco local time) the fire chief reported, all passengers and crew have been accounted for, final numbers were 2 occupants killed, 10 in critical condition, 38 with serious injuries, 82 with minor injuries, 175 uninjured. The confusion about people being not accounted for was the result of survivors being taken to two different locations at the airport. The two fatalities were 16 year old Chinese girls travelling as part of a school outing.

Asiana confirmed their aircraft suffered an accident while landing in San Francisco, there were 291 passengers and 16 crew on board.

The General Hospital in San Francisco reported they received 10 passengers from the flight OZ-214, 8 adults and 2 children, all in critical condition. In the evening of Jul 7th PDT the hospital reported 6 passengers were still in critical condition.

The city of San Francisco and emergency services in a joint press conference reported that emergency services responded post landing, 48 people were transported to hospitals, 190 passengers were collected and taken to the terminal, 82 of which are probably going to be tranferred to hospitals, there are some people unaccounted for. The aircraft carried 291 passengers and 16 crew, total 307 people on board. There were two fatalities. In a second press conference Saturday evening the fire chief corrected the earlier statement now stating that all passengers and crew have been accounted for, the earlier confusion in Saturday afternoon's press conference was caused by two locations at the airport used to take survivors to. There were 305 survivors, 2 fatalities, of the 305 survivors there are 10 in critical condition, 38 received serious injuries, 82 minor injuries.

The airport was completely closed for about 5 hours, then runways 01L/19R and 01R/19L reopened, both runways 10/28 remain closed.

The NTSB reported the Boeing 777 of Asiana, flight 214, approached runway 28L when it suffered an accident, three investigators from the West Coast as well as a response team from Washington have been dispatched on scene. Korea's ARAIB have been invited to join the investigation.

ATC recordings show, the aircraft was on a normal approach and was cleared to land on runway 28L, no emergency services were lined up, all traffic was running normally. During a transmission of tower shouting in the back of the tower is heard, emergency services began to respond, all aircraft on approach were instructed to go around. The airport was closed. United flight 885, waiting for departure at the hold short line threshold 28L, reported people were walking around both runways, there were a number of people near the numbers of runway 28R, obviously survivors.

An observer on the ground reported that the approach of the aircraft looked normal at first, about 5 seconds prior to impact the aircraft began to look low and then impacted the sea wall ahead of the runway.

On Jul 7th the NTSB reported in a press conference at San Francisco Airport, the crew was cleared for a visual approach to runway 28L, the crew acknowledged, flaps were set at 30 degrees, gear was down, Vapp was 137 knots, a normal approach commenced, no anomalies or concerns were raised within the cockpit, 7 seconds prior to impact a crew member called for speed, 4 seconds prior to impact the stick shaker activated, a call to go-around happened 1.5 seconds prior to impact, this data based on a first read out of the cockpit voice recorder. According to flight data recorder the throttles were at idle, the speed significantly decayed below target of 137 knots - the exact value not yet determined -, the thrust levers were advanced and the engines appeared to respond normally. The NTSB confirmed the PAPIs runway 28L were available to the approaching aircraft before the accident, however were damaged in the accident and thus went out of service again. The localizer was available, the glideslope was out of service, according NOTAMs were in effect. There were no reports of windshear and no adverse weather conditions. The air traffic controller was operating normal, no anomaly was effective, until the controller noticed the aircraft had hit the sea wall. The controller declared emergency for the aircraft and initiated emergency response. ARAIB and Asiana personnel have arrived on scene and have joined the investigation. The Mayor of San Francisco reported runway 10L/28R was cleared for service.

On Jul 8th 2013 the NTSB reported the pilots' flight bags and charts were located, the proper (approach) charts for San Francisco Airport were in place at the cockpit. There were 4 pilots on board of the aircraft, they are being interviewed on Jul 8th, which will be determine who was pilot flying and who was in command at the time of the approach. The cockpit was documented and the switch positions identified. Both engines were delivering power at time of impact consistent with the flight data recordings, the right hand engine found adjacent to the fuselage showed evidence of high rotation at impact, the left hand engine liberated from the aircraft also showed high rotation at impact. The aircraft joined a 17nm final, the crew reported the runway in sight before being handed off to tower. The autopilot was disconnected at 1600 feet 82 seconds prior to impact, the aircraft descended through 1400 feet at 170 KIAS 73 seconds prior to impact, descended through 1000 feet at 149 KIAS 54 seconds, 500 feet at 134 KIAS 34 seconds, 200 feet at 118 KIAS 16 seconds prior to impact. At 125 feet and 112 KIAS the thrust levers were advanced and the engines began to spool up 8 seconds prior to impact, the aircraft reached a minimum speed of 103 KIAS 3 seconds prior to impact, the engines were accelerating through 50% engine power at that point, and accelerated to 106 knots. The vertical profile needs to be assessed first. There was debris from the sea wall thrown several hundred feet towards the runway, part of the tailcone is in the sea wall, a significant portion of the tail is ahead of the sea wall in the water.

On Jul 8th 2013 South Korea's Ministry of Transport reported the captain (43, ATPL, 9,793 hours total) of the ill-fated flight was still under supervision doing his first landing into San Francisco on a Boeing 777, although he had 29 landings into San Francisco on other aircraft types before. He was supervised by a training captain with 3,220 hours on the Boeing 777, all responsibilities are with the training captain.

Relevant NOTAMS:
07/051 (A1331/13) - NAV ILS RWY 28L LLZ/DME U/S. 07 JUL 17:00 2013 UNTIL UFN. CREATED: 07 JUL 03:08 2013

07/048 - RWY 10L/28R CLSD. 06 JUL 23:10 2013 UNTIL UFN. CREATED: 06 JUL 23:10 2013

07/047 - RWY 10R/28L CLSD. 06 JUL 23:09 2013 UNTIL UFN. CREATED: 06 JUL 23:09 2013

07/046 (A1326/13) - RWY 28L PAPI U/S. 06 JUL 22:19 2013 UNTIL UFN. CREATED: 06 JUL 22:19 2013

07/045 (A1324/13) - AD AIRPORT CLSD. 06 JUL 20:10 2013 UNTIL UFN. CREATED: 06 JUL 20:10 2013 (cancelled at 23:09Z)

06/005 (A1056/13) - NAV ILS RWY 28L GP U/S. 01 JUN 14:00 2013 UNTIL 22 AUG 23:59 2013. CREATED: 01 JUN 13:40 2013

Metars:
KSFO 061956Z 23004KT 10SM FEW016 19/10 A2981 RMK AO2 SLP095 T01890100
KSFO 061856Z 21007KT 170V240 10SM FEW016 18/10 A2982 RMK AO2 SLP098 T01830100
KSFO 061756Z 21006KT 10SM FEW016 18/10 A2982 RMK AO2 SLP097 T01780100 10183 20128 51005
KSFO 061656Z VRB06KT 10SM FEW013 SCT018 17/10 A2982 RMK AO2 SLP096 T01670100
KSFO 061556Z 02003KT 10SM FEW012 SCT018 16/11 A2982 RMK AO2 SLP096 T01610106
KSFO 061456Z VRB03KT 10SM FEW010 SCT015 14/10 A2980 RMK AO2 SLP092 T01440100 51006

Video by ground observer Fred Hayes filming the entire landing sequence at: https://edition.cnn.com/video/?/vide...esy-fred-hayes

Video of the aftermath by a ground observer (Video: SecretStoryExclu):

[zizu]

Δεν διέταξαν άμεση εκκένωση οι πιλότοι του αεροσκάφους της Asiana...

https://www.naftemporiki.gr/story/674232

[zizu]

Πολύ ενδιαφέρουσες απόψεις για το ατύχημα από το flightradar24

https://www.facebook.com/flightradar24?ref=ts&fref=ts

[pilot]

Άκρως ενδιαφέρον...

https://www.bloomberg.com/news/2013-...iners-say.html

[Monsta]

.....

[pilot]

άρμα μάχης πάντως το Triple Seven! κιχ δεν έκανε η άτρακτος!

[Monsta]

Here's the latest info. based on NTSB interviews, ATC recordings, flight paths and airport information at the time.

There is a significant amount of background info. involved, so will get that out of the way first to paint the picture.
The point of all this is for understanding and awareness, not to place blame or find fault.
Also, there is still some information missing such as complete interview transcripts, aircraft configuration at important points etc.
But there is enough information to show many of the factors involved.

Airport Environment: SFO has always had issues with arrival volume due to some unique aspects of its construction, location and weather.
-The main landing runways, 28L and 28R, are only 750 ft. apart.
That means normal simultaneous approaches in IMC or instrument meteorological conditions cannot be conducted.
Normal simultaneous ILS runway separation is 4300ft. min., and there have been plans
to move 28R out the required distance into the bay.
Legal and environmental factors are still preventing that from happening and no new date has been set for the relocation of 28R.
This can cause normal flight paths to be adjusted for aircraft separation and a need
for the controllers to try to obtain visual approach acceptance by pilots to expedite landings when flight conditions permit.
-The surrounding major airports around SFO, SJC and OAK, have conflicting traffic paths, namely OAK arrivals and SJC departures, during normal westerly operations, and SFO arrivals.
This can cause frequent cleared changes in speed and heading while also causing aircraft to become moderately high on the approach path to clear other conflicting aircraft paths.
-SFO commonly has IMC or instrument meteorological conditions which means aircraft separation must be increased and pilots cannot accept visual approaches.
This is what causes most of the delays at SFO since the closeness of the runways causes the two landing runway international airport to essentially become a single landing runway airport.
-SFO commonly does maintenance on the ILS or instrument landing system during VMC or visual meteorological conditions because of the frequency of IMC there ie. it's their only chance to do maintenance rather than close one of the runways that could close the airport.

Communications environment: SFO is one of the busiest international airports in the US.
That means that even though there are ELPR's or English language proficiency requirements for pilots, often those proficiency levels are suspect.
What this does is require multiple radio calls and repeated need for clarification from some international pilots and controllers.
This can reduce the time and increase the workload for pilots to do their jobs.

Cultural environment: It has been well documented in aviation training histories that some cultures have a tendency to automatically defer to authority and try to "save face" when a superior is recognized as making a mistake or underperforming.
It was sometimes present in the US back in the pre-'80's and after a major aviation accident, became one of the foundations for modern CRM or crew resource management.
While it seems that CRM has drastically improved safety with most airlines, there may still be some cultural differences with some countries that will cause pilots to still defer to authority or superiors while not pointing out discrepancies from objectives and safety.

Pilot experience/training: Over the last few years it has become obvious that training and experience of pilots needs to be reexamined.
With the Colgan and Air France accidents lately, and others before them, a review of strategies to ensure pilot competency have come under the microscope.
Unfortunately, due to the costs associated with fuel, aircraft, insurance etc., and tightening economic times, pilot training and experience has become increasingly simulator based.
This has also caused more flight automation to be used to help "unload" pilot attention and awareness, especially since the sim environment and real aircraft operations are very different in the real world of real time multiple issues and events and the resulting stresses those create.

Pilot condition: Obviously a major issue to a pilot performing at their best is to be well rested, nourished and in a healthy state of mind.
It's also important to have distractions removed from the flight deck. Pilots are usually well versed in compartmentalizing emotional issues while flying, but they still can be a factor in simply distracting or reducing the efficiency of the pilot. Don't really have any information on this regarding Asiana 214, but it could have been a factor.

Boeing 777 factors: Although it's been reported the pilot in training only had ~40 hours in the 777, it is common for transitioning pilots to be very experienced in other aircraft, which he was, yet low time in the aircraft they're training on. The captain being trained in the 777 had previously come from flying the Airbus A320. An interesting point because the A/T system in the A320 is slightly different with different names and indications than the Boeings have.

Moving on to the specific events that day that led up to the accident, there are still pieces missing
such as configuration stages, complete CVR or cockpit voice recorder records and complete interviews from the pilots.
First a review of the facts and then some insight into how everything fits together and likely reasons for what happened based on the fair amount of information out now.
But just to be clear, this is one person's view with incomplete information and there may be additional information that comes out which could bias towards conclusions differing from the following.

Weather: KSFO 061756Z 21006KT 10SM FEW016 18/10 A2982 RMK AO2 SLP097 T01780100 10183 20128 51005

Approach: Visual 28L which means the pilots had the airport in sight, accepted responsibility for the approach and aircraft separation.

Aircraft: No emergencies declared, no deferred items affecting the aircraft's ability to conduct the approach and landing.
No mention of systems which failed enroute or on the approach.

Path and performance:

Altitude...........Alt. Dev..............Speed............Spd. Dev......Distance...........Comments

A. 6000-1800ft.......+500ft.............220-170kt.........+0..............12-4mi............fairly normal except approaching 1600-1800ft.

B. 1600.................+500.................170..... ..........+30.............3.5................fast at the FAF, disconnects A/P, turns off F/D

C. 1000.................+300.................140..... ..........+5...............2.5.................... correcting as expected, fairly normal

D. 500...................-50....................134...............+0........ .......1.5................starting to get low, on speed, PM says pull back for low path, see below

E. 200..................~-60..................118...........~-15-20...........~.8............getting low, very slow, instructor moves to push throttles up

F. 100.................~-70...................110...........~-25................~.5...........excessive sink, extremely slow, throttles up

G. 3 sec to impact..~-80..................103..........~-30+...............~.2...........leveling off above water, engines going to max thrust but excessive angle of attack

H. 1.5 sec-ground...........................106..........~-30..................0.............aircraft accelerates ~5kt.

I. Gear and tail strike sea wall, gear collapses, tail departs

Perspective: What jumps out is how normal the approach is until approx. 1800 ft.
Even though the acft. is high by 500ft. all the way to 3.5 miles, this can be fairly normal at SFO with the above factors of approach paths, visual rules and compensating for the other airports and aircraft.
Their high path correction makes the fact that the A/T's are possibly not engaged masked because he's trading altitude for airspeed.
Regardless of if the A/T's are engaged or not, the PF and PM are still responsible to manage airspeed manually with their hands on the throttles.
But if a pilot is used to the A/T controlling speed for him without scanning the speed, it can cause a false sense of security with speed control and the need to always scan it.
What happens then is interesting and fairly common actually.

And the key here are 3 things related to vertical control and indications of the aircraft:
- Mode control panel (MCP) interaction with flight director indications (F/D)and flight management annunciations (FMA)
- Autothrottle (A/T) interaction with MCP and FMA
- How the pilot normally flies the automation, FMA scan and hand flying technique

The standard procedure on a visual approach is to set the final approach fix (FAF) or last altitude on the ILS in the altitude window on the MCP
as you're cleared for the visual approach.
This provides an altitude target and safety for being stabilized and configured.
Different airlines have different procedures at that point for altitude window procedure, either to set touchdown zone elevation or go around elevation.
The flight director pitch during the descent will be determined by the MCP vertical function selected.
But, and this is a big but, it changes the A/T function and safeguards depending on what is selected.
The A/T is always armed (if the switches (L&R) are on and they should be), and if the aircraft is slow in most vertical modes, near the speed tape hash limit marks, will automatically reengage and add power (also known as A/T wakeup),
unless in FLCH or level change mode..........
This is why most airlines make it against procedure to select FLCH inside of the FAF to retain the A/T wakeup mode.
So as they're descending from 6000ft., they get cleared for the visual approach, the non-flying pilot puts the FAF altitude in the MCP which is 1800ft.,
and they FLCH or V/S the vertical mode down to get on the assumed vertical slope.
All Boeing jets with FMC's can build a virtual glideslope indication even if the ground based ILS is out (as it was) and use that as a reference, also called a VNAV approach.
The problem as stated earlier with SFO is they often give changes on the visual approaches, that happen too fast for the computer to recompute and capture.
So most will use the VNAV vertical slope indication as a reference but hand fly the jet down due to the rapid changes that can happen on a visual approach.
Continuing, they are in FLCH or V/S and approaching the FAF at 1800ft., the MCP, FMA and F/D will capture the altitude, the vertical FMA will go to altitude hold, the A/T will come back on and add power to maintain the MCP speed and the aircraft will accelerate.
This is probably why at this point, the pilot flying the jet (PF) turns off the autopilot, disconnects the A/T and turns off his F/D, because he's high, fast and doesn't want the airplane getting farther off the vertical path by leveling and the power automatically coming up.
At this point the PM's F/D isn't giving good info., because the airplane thinks they wanted to level off at 1800ft., the FAF, and so the pilot monitoring (PM) in this case the instructor pilot, changes the altitude in the MCP and hits FLCH to get "his" F/D to give a descent indication. But remember, the FLCH function disables the A/T wakeup mode.
Now the FP believes he has the A/T because he never uses FLCH on an approach, yet it's not going to come on to correct his speed because the PM selected FLCH to get his F/D back.
He gets on path and speed around 500ft which is fine, starts raising the nose
to maintain path, but the aircraft is still sinking at the previous excessive rate that he had to be on to capture the vertical path, because remember he was high.
A 777 with idle thrust and decaying airspeed with full flaps and gear down will set up a significant sink rate that only a large thrust increase will arrest.
In fact all airlines would go around upon seeing that under 500ft.
But he's thinking the A/T's will come up and provide thrust, yet they're not. (he has to move the throttles with his hands to override the A/T and he's not)
At this point the airplane is sinking excessively, yet there seems to be a lack of any communication to resolve the issue. Training, overload, stress, scan assumption/failure, communication, culture?
At approx. 7 seconds to impact the instructor pilot notices the speed and moves to the throttles, calls go around at 2-3 secs., and the rest is as it happened.
Imho, it will come out that the PF and PM had a communication issue between what was desired, what was selected, and what was expected with the MCP, F/D and A/T in addition to major scanning issues
The PF, and in this case PM since he was responsible as the instructor, still are responsible to scan, cross check and fly the airplane regardless of automation assumptions.
In the interview the instructor pilot said, he thought the autothrottles were maintaining speed.
This goes back to the idea that real world training has suffered, although it satisfies a simulator and written test from computer training.
Another factor, and it will be discussed, is the "technique" for some to turn the A/P off, yet leave the A/T on.
So the pilot is hand flying the yoke yet the A/T is controlling the throttle.....I've watched this happen over and over,
the PF starts scanning everything "except" speed because he's gotten in the habit of the A/T controlling the speed for him, and
for whatever reason, the A/T disconnects, the FMA that it's disconnected is missed and the airplane starts slowing down.
Imho it should be A/P on for everything or off for everything, ie. it's you or the plane controlling inputs, not a combination.
Also, the procedure to not drug/alcohol test the pilots due to them being foreign needs to be changed imho.

Here is an interesting animation/narration of the Turkish accident in 2009, that while different in lead up causes,
resulted in a similar confusion about the A/T and an almost identical outcome.

Και όπως λέει και ο τίτλος του άρθρου που έβαλε ο πάιλοτ...

.............

Once again back to the money and training idea that they don't have the experience for widely varying non-profile situations. (like SFO which is basic to us)
Even though FLCH isn't taught for approaches, it is used by everyone on Boeings to get their F/D back after an unwanted altitude capture at higher altitudes.
That habit on an approach will disengage the A/T and A/T wakeup feature, and we see that more as US pilots because of the frequency with which we see visual approaches,
unlike in foreign countries where they seldom do them. The other one I mentioned, TO/GA or takeoff/go around mode, will also disengage the auto wakeup feature,
but is rare to use when continuing an approach. What I have seen happen is the PF wants to disengage the A/T on the side of the throttles but mistakenly pushes the TO/GA switch below the throttle.
That could have also happened, and if they didn't reselect VNAV, V/S or APP at that point, they would also have a disengaged auto wakeup function.
What is disturbing about all of this is the idea of that being responsible. At any time, any automation function can fail/disengage for a number of reasons.
In fact I believe there are some 27 reasons the automation can give a fault on the approach for an autoland and require a go-around or hand flown approach.
To say that one of those failures, or simply not having the right mode engaged, is somehow responsible for an accident, incorrectly relieves the pilots of being responsible for the aircraft at all times.
Hopefully training and checkrides will be improved as a result of this if it turns out to be pilot error.

&

.............

SFO isn't really unique other than flight path changes happen quickly there due to the proximity of other airport paths. (see my write-up)
These changes happen to allow the spacing to be closer to allow the volume of traffic to land there without delays through visual approaches.
Unfortunately, many foreign carriers are very used to scripted ILS or autoland approach profiles as their primary ops.
When airports are clear in the US, we will almost always revert to visual approaches though to tighten separation
and get more aircraft on the ground, which can be a challenge for pilots not used to a non-ILS environment.

This has been a known issue with some international pilots, but they've gotten away with the lack of experience up until now.
Consider the context, most American pilots are doing visual approaches as teenagers, in some countries they may not see them until at an airline.
Then at that airline, their flying is almost all autopilot operated. You can't fly a tight-in visual approach on the autopilot, it doesn't react quickly enough.
We love it like driving a manual BMW or Porsche ie. you're driving the car. They hate it because they were raised on DCT/PDK and now have to do it manually, if you want an analogy.
It tends to overload those who aren't proficient at it and they get behind. But unlike a car, you can't put the brakes on, time and distance are moving and you have to get it right.
The right thing to do if that's happening to you and you're behind is abort the approach and go-around. That was their other major mistake is they waited too long to decide that.

They had to have seen the PAPI (precision approach path indicator) at least 4 miles out, with plenty of time to correct their path.
(they also had a VNAV profile on the FMC and altitude green arc on their EHSI)
They seemed to have waited to correct it, got behind, got overloaded/distracted, let the airspeed go, or even worse assumed it was automated and waited too long to go around.
This is a common set of mistakes students make when in a high speed, fast changing environment and visual with minimal experience.
At the professional level though, this shouldn't be happening and hopefully is a wakeup call to international training.
Keep in mind that's just what the info. and my experience says, and is not an official view yet.

&

.........

He had ~10,000 hours total time and had flown to SFO before. You have to look at the quality of their flight time.
If you turn the autopilot on at 500 ft on the takeoff, and off at 500 ft on the landing or do an autoland,
and do 15hr. flights, how much of that 10,000 hours is hand flown? Very little. And sadly, it isn't bizarre, there is a history of this.
SFO often does ILS's and autolands, he may have always done those before there. But that day was a visual with all the differences that go along with it.
That's my point, regardless of the time some international pilots have, they commonly are not high time manual/visual pilots.

[Monsta]

Drop them pdk's and get back to manuels !!!!

[ntinos]

Ελπιζω οι αυριανοι (2 πτησεις) των τριπλ-σεβεν να μην ειναι ρουκηδες και μας φανε οι ουρακοταγκοι....

[ntinos]

Λάθος... A340 το δεύτερο