News Asiana 777-200 Crash on landing at SFO

Something distract the pilots.
All pilots is trained to always watch their PFD at all times.

Not when landing.

When landing the PF will focus on what is outside and hand flying. The PNF should be focused on the instruments. Yes, they should have seen the speed decay, the box around the IAS will flash when the speed is too low. They should have felt it - why didn't they?
 
And it bring me to this question.
Should a HUD like in the B 787 not be helpful in such situations?
I also noticed that their was a B 747 taxiing to runway 28L/R at the moment of the crash.
The pilot's in the B747 surely do see it coming.
Luckily the B 777 did not veer in their direction.

Here is a typical landing to SFO runway 28L without ILS.
 
Last edited:
It's ironic that a lot of people don't trust machines to fly planes / drive vehicles and prefer the human factor. Especially when most tasks in an airplane are already somewhat automated and could be automated fully with some more R&D and work.
How do we know that wasn't a contributory cause? The pilots might have just thought "oh, we're on autoland, we can take a few seconds to discuss training rather than focusing on the instruments." Not saying they did, but I'm skittish about giving the final say over to a machine in such a safety-critical application.
 
Hey, not only the 787s have HUDs: I have seen multiple Delta 737-800s with what appeared to be a F-15E-derived wide-angle HUD on the left side of the cockpit.
 
How do we know that wasn't a contributory cause? The pilots might have just thought "oh, we're on autoland, we can take a few seconds to discuss training rather than focusing on the instruments." Not saying they did, but I'm skittish about giving the final say over to a machine in such a safety-critical application.

Below 10,000ft they should be focused on the landing and talking only about the landing (sterile cockpit). Automation is fine but I'd prefer humans in the loop although I'd prefer my humans to be better trained with more hours and more capability to handle a simple ILS out manual landing.
 
What is a bit shocking is that the presence of a pilot-instructor didn't changed anything. Or he was not listened, or he didn't noticed anything too... I just hope he wasn't sleeping in the pilots/crew rest seats, in which case it would be dereliction of duty...
 
What is a bit shocking is that the presence of a pilot-instructor didn't changed anything.

Maybe......

Or he was not listened, or he didn't noticed anything too...

He shouldn't have said ANYTHING unless it was serious, below 10,000feet it's sterile cockpit - shut the hell up unless it's relevant to landing. Any training should be left behind - above 10,000ft.

But it raises several questions:

1. Who was in charge? The captain with few 777 hours?
2. The FO with more 777 hours than the captain?
3. The training captain?
 
They have survived and the investigation teams has the FRs, so we will know. Maybe it will end in pilots rejecting fault on each other or the plane's systems, but there are enough elements to draw a "more than plausible" conclusion. Now, I guess it will take months, listening and analyzing the stuff plus interrogating the pilots is time-expensive.
 
What is a bit shocking is that the presence of a pilot-instructor didn't changed anything. Or he was not listened, or he didn't noticed anything too... I just hope he wasn't sleeping in the pilots/crew rest seats, in which case it would be dereliction of duty...


There was no pilot-instructor listed in the crew, there had been just two pairs of pilots, flying the aircraft in shifts.
 
There was no pilot-instructor listed in the crew, there had been just two pairs of pilots, flying the aircraft in shifts.

Not according to this:

Airline spokeswoman Lee Hyomin said Monday that Lee Gang Guk, who was at the controls of Saturday's nearly 10 1/2 hour flight from Seoul as it arrived at SFO, was a veteran pilot with nearly 10,000 hours flying other planes. But he had only 43 in the 777, a jet she said he still was getting used to flying.


Lee had flown Boeing 747 jets into San Francisco International previously and was assisted on this flight by deputy pilot Lee Jeong Min. The deputy had about 12,390 hours of flying experience, including 3,220 on the 777.
National Transportation Safety Board Chairman Deborah Hersman said Lee Gang Guk was flying with a supervisory training captain, another captain and a first officer. It was his ninth training flight on a 777, she said.

Emphasis mine and source: http://www.usatoday.com/story/travel/news/2013/07/08/asiana-airlines-san-francisco-crash/2497769/

So we have the Captain, the 'deputy' (first officer), another captain and first officer plus the supervisory training captain.

Just who was actually in charge?
 
Interestingly in the Turkish Airlines 737 crash that I posted earlier the situation was a bit similar - a first officer that has recently joined the company and was being assessed by the captain, who was one of the most experienced pilots in the airline. A third "deputy" co-pilot was assisting from the jump sear on the short flight from Istanbul. Yet none of the three has found any problems with their airspeed or that the plane was automatically throttling down in response to the erroneous radio altimeter reading when landing in Amsterdam.

Maybe the situation is similar here - none of the pilots in the cockpit has noticed that the airspeed is dropping to stall levels. By the time the hand shaker starts to jolt, it was too late (although I wonder why the warning came just 4 seconds before impact). Luckily we will probably know why the pilots responded like this, since all of them survived, unlike in the TK 737 crash.
 
It's ironic that a lot of people don't trust machines to fly planes / drive vehicles and prefer the human factor. Especially when most tasks in an airplane are already somewhat automated and could be automated fully with some more R&D and work.

It's likely that if the ILS was working, this wouldn't have happened...

That is no excuse.

If you cannot land an airplane manually in clear weather without stalling, you've got no business being in the cockpit. It's Pilot 101.
 
although I wonder why the warning came just 4 seconds before impact

Because that was the point that the plane entered the stall. They would have (should have) had numerous indications that they were flying to slow before hand. Just looking out of the window and seeing clear sky rather than a runway due to the nose pointing over 10degress into the air would have given it away as would a quick glance at the PFD. That means that no one was looking out of the window and no one was looking at the PFD. I would have also thought that the sound of the engines at flight idle would have been picked up to, normally on approach, an aircraft's engine pitch changes quite noticeably.

Not liking the sound of that FLCH trap at low altitude at all. I can imagine the autopilot design team saying "wow - never thought they would to THAT".

But that's a known issue, probably circulated in various bulletins and maybe even mentioned in training and who puts themselves into a FLCH trap? The crew do by letting the plane get ahead of them.

I'm no pilot but surely common sense says 'If you don't understand why something is behaving the way it is then you need to give yourself time, space and room to work it out' yet they pressed on with the landing.

As the flight bound from Incheon in South Korea approached San Francisco after its 11-hour journey across the Pacific Ocean, three out of four pilots aboard were in the cockpit.

Lee Kang-kuk, who was still completing his initial training on the Boeing 777 and had never before flown one into San Francisco, was at the controls, Ms Hersman said on Tuesday.

Beside him and in command of the aeroplane was an instructor pilot, flying in that capacity for the first time, Ms Hersman said.

In the jump seat behind the two pilots was a relief first officer who had flown to San Francisco five or six times as a monitoring pilot. A fourth crewman, serving as relief captain, was in the cabin as the plane landed, and was still being interviewed by investigators on Tuesday.

As the plane approached on a clear day, the pilot in control of the plane was cleared to land. About 34 seconds prior to impact, the plane was flying at 500ft and at about 134 knots (154mph; 248km/h), when the instructor pilot realised it was flying too low.

He told the pilot to pull back on the stick, and seconds later he realised that the automated throttle controls, which had been engaged, were not maintaining the correct speed of 137 knots. About eight seconds before impact, the pilot in control pushed the throttles forward to speed up.

Less than two seconds before the crash, the pilot tried to abort the landing, but it was too late. The plane came in much too shallow. The main landing gear struck a sea wall well short of the end of the runway, then the tail struck and was ripped off the rear of the aircraft.

The aeroplane then rotated left and went into 360-degree spin before coming to rest to the left of the runway.

The first officer was hospitalised with a cracked rib, and neither of the two pilots were seriously injured.
Source: http://www.bbc.co.uk/news/world-us-canada-23249012

_68593615_plane_crash_624.jpg
 
all three were blinded by something? If so surely one of them would have said something/uttered an exclamation or expletive which would be on the CVR?
 
all three were blinded by something? If so surely one of them would have said something/uttered an exclamation or expletive which would be on the CVR?

And blinded from where? Also blinding does not explain the systematic errors. The stick shaker would have been felt and heard even if Motörhead would have been playing in the cockpit.

Asiana_Flight_214_Approach_to_SFO.png


Also note that the cockpit switch configuration was everything but normal after the crash. Both autothrottle switches had been set to "armed" not "on" or "off", and one flight director was disengaged.
 
Last edited:
A lot of the reports and data points things starting to go to hell 34 seconds before impact. If it's true that they were blinded they should have performed a go-around then.

Even with the slowest reactions in the world they'd have been performing the go-around with at least 10 seconds left in the flight - just enough time to power up the engines and fly away.
 
Yes. Also the sun does not work out as cause of blinding, it was noon when the crash happened.
 
They are now saying it may have been a laser but that shouldn't have affected the PNF who should have been 'head down' watching the speed.
 
Back
Top