If you look at webtrack they turned left, turned right (abeam the Montebello label on the map, turned left and then right.

This basically lines up with the various instructions as best I can tell.

I probably got the idea from some of the FR24 forum posts like this one from 'FlightAware Staff':

MLAT - This flight is restricted from public view: ADS-B Flight Tracking

This was discussing an mlat plot of maneuvering fighters but it does seem to explain some of the FR24 plotting behavior as the ADS-B coverage gets spotty.

Well.......apart from the fundamental 'heading 180 degrees' instruction !

As pilots, our primary response to any ATC heading instruction is the heading given. If we are told to turn onto a new heading, then that heading is the number one priority, and that 'number' is what we set with the heading knob. If the turn direction is not given or seems illogical or wrong, we will query that turn instruction. BUT we will not ignore the heading given, (as long as we are not being turned directly towards a hazard of course).

EVA015 were given a heading of 180 degrees. The actual turn direction might have been wrong or odd or whatever, but they :

a) failed to query the turn direction, and;
b) failed to turn onto the heading they were given.

They actually maintained a heading which was about 160 degrees opposite to what they were initially told (and read back) - despite repeated instructions to turn - and it could have got them killed. Twice. First with AC788 and again with Mount Wilson.

I still wonder if they had a compass/heading malfunction, because I find it hard to believe that a longhaul B777 crew would ignore such basic instructions otherwise.

I doubt it. It looks to me more likely that they were about a third of the way through the original left turn to 180 deg when they were told to reverse the direction of turn, which they began, only to be told to reverse the direction yet again (and head 270 deg), which they began, and then told to head southbound to which they tried to query the controller about the direction of turn and got no answer. It is only during that last part that they maintained their present heading (which happened to be north-ish as a result of the previous maneuvers) while they tried to obtain clarification about the direction of turn. After all, it was the controller getting her left and right mixed up that started the whole mess, so they figured direction of turn was important. Had they maintained heading and queried the controller for that first turn, all of this would be averted, so they were just a bit late in doing the right thing.

Anyway, it wasn't their compass, but rather the timing of the controller's instructions that got them going north.

You have been repeating the same of line of argument in this thread. FAA spokesman has said the controller meant to say "right" but said "left."

We could hear the EVA crew read back the first "left turn" instruction, and the controller did not correct. So what do you think the EVA pilot should do at this point? Steer left or right or keep chatting on the radio?

Additional tracking data have shown that EVA had initiated turns as instructed, but appeared to have repeatedly been given new directions to turn the opposite way before the current turn was completed.

Towards the end, the controller stopped giving turn directions all together. But, like others have pointed out, after having been given repeated, emphatic instructions for turn direction, the EVA crew could assume by this point that somehow for reasons beyond them that the turn direction was important. Hence the delay to commence the turn to southbound.

If the compass really malfunctioned, how did the flight complete the trip to TPE later?

Uplinker:

The 777 doesn't have a compass. It is a highly redundant inertial system (three inertial platforms) that also has mag var tables to convert true heading into magnetic heading when in domestic airspace. These inertial platforms also provide the attitude platform for the airplane. On rare occasions one inertial reference unit (IRU) might fail. In that rare event the airplane only loses triple redundancy.

These highly redundant inertial systems are not foolproof and if not aligned properly can display incorrect headings. eg:

AirAsia Navigation Error

It passed pretty much over the top of my apartment. I was there but don't claim to have heard it.

Here's the LAX Terminal Chart (1:250,000). I have pretty much stopped flying now so it is from 2014 but they had not built any new towers the last time I looked out of my window.

Perhaps some photoshop whizz with a bit of time could superimpose the track to see how close to the towers it was but it looks pretty near to me by eyeball. Maybe a fraction to the South. In the class G for sure.

Bleve:
Gross pilot error combined with a company too cheap to buy the mod from AB that would have prevented that.

EVA obviously did not have that problem as represented by their track to TIA once they got it pointed in the correct direction.

Uplinker, I think you are missing the point.

The aircraft started various turns in response to instructions from Atc.

Give someone a turn through 270 degrees to turn to 180 and the aircraft will, at some point turn through north.

When it is pointing in the northerly direction give it another instruction to turn right and it will head generally north while the angle of bank comes off and the right turn commenced.

Issue another instruction and the right turn comes off and a left turn commenced, all the while the aircraft heads generally northish.

Would an native English speaking crew have handled it differently, possibly, but they weren't put in that position. It was a crew with English as a second (third or fourth) language.

On another point I am not sure what the panic was re air Canada. The webtrack (I don't know how accurate this is) on face value has plenty of vertical separation between the two, despite the various level off instructions given by the controller. The webtrack may not give the complete (or accurate) story but by the time the two aircraft got close there were thousands of feet vertical separation according to the webtrack labels. Davereiduk's screen shot shows that there is close to 6000ft vertically between the two aircraft.

Snakecharma has summed it up perfectly IMO.

Snakecharma,
I will make some inquires with a 787 pilot I know in Toronto, and see if he heard what the AC crew saw..re EVA below them. I'm sure they were looking for the other aircraft and had it on their TCAS!

Retired DC9 Driver:

The chances of a collision was near zero. The controller had been trained to think airplane-to-airplane separation and had only absorbed that into her "priority memory bank." Knowing the ATC facility I am sure they thought they had trained her better,

Alas, she was trapped by the terrain to the north of LAX, which may have been a cursory part of her training.

Damn, this is all pathetic without an NTSB incident investigation.

I wouldn't be surprised to learn that the 270 heading was meant for AC to provide additional separation but the controller was pre-occupied with EVA and gave the wrong aircraft the westerly heading.

HighSpeedAluminum:

The Ventura 7 SID requires radar vectors.

A question for non-native English speakers on PPRuNe ... what would be understood by the colloquialism "southbound"? Granted it contains the word "south", but if you were not familiar with the word, what would "bound" mean to you as an adjunct? And what did these guys take it to mean?

towards a Southern direction (as ATC instruction), towards the South, 180°

going / heading / travelling / flying / navigating towards and maintaining a Southern direction,
say, from 120° through 240° or from SE to SW roughly.


to me it is clear she wanted them to avoid flying Northbound where the mountains are.

"Southbound" is an adjective which describes the direction of a track, such as " southbound M6"

"South" can be an adverb which tells which way a specific motion goes.

So the correct English, whether in California, Cumbria or China is "Turn South" and no one will be confused.

Or even better, "turn heading south", but preferably with a "right" or "left" after "turn"!

The way he hesitates reading it back suggests that he's unfamiliar with the word.

Duncan, while the southbound instruction is a bid odd in my opinion it is decipherable, my point was more directed towards the initial instruction to turn left and the language and cultural differences that might have led to the instruction being followed without question, this leading to the problem in the first instance.

Perhaps a native English speaker should have been able to articulate their concern about the turn direction, whereas a crew with essentially procedural English would need to have a conversation on the flight deck.

The additional complication would be the cultural issues associated with questioning authority, if the culture of the crew is such that instructions are adhered to and only questioned under extreme circumstances then there is further room for problems to occur.

I am not suggesting the crew handled it perfectly or even well, but I don't think we can hang the crew given all the circumstances.

Here are 2D and 3D views of the path near Mt Wilson. These are reduced size, links to larger size images at bottom.
http://i.imgur.com/MsGwsWa.jpg

http://i.imgur.com/Mo6DGYA.jpg


http://imgur.com/vlBVLbh.jpg
http://imgur.com/SKUY8o8.jpg
http://m2ei.com/EVA_Flight_015/EVA_F...tal_height.kmz

Wow. Close indeed. :eek:

Thanks to you and DaveReidUK for the plots. :ok:

WebTrak may be showing the FAA data which is denser than the FR24 .kml file.

At any rate, the data from various sources that I've found is in close agreement, this incident was very nearly a major tragedy.

Aterpster,

That is factual but it doesn't address the point i was making.

After ATC corrects their initial error with "right turn, right turn heading 180" which EVA acknowledged "copy, right turn heading 180". Seconds later ATC instructs "expedite your right turn" which was responded with "roger, we are passing heading 010 continue right turn heading". In the next transmission to EVA, ATC instructs "stop your climb" and moments later "turn left, left turn heading 29..correction 270" and EVA responded with "left heading 270".

EVA crew were (albeit with apparent loss of SA) following each ATC instruction without query. The left turn heading 270 interrupted the "expedite right turn" instruction. This second ATC error IMO was meant for Air Canada to provide additional separation. Had that instruction been given to the proper aircraft EVA might likely not have been in the situation they found themselves in.

It appears that the crew failed not only to be aware initially of the terrain to the north, but also failed to respond to the GPWS warnings. Regardless of whatever ATC clearances they were given, they should have been responding much more aggressively to the GPWS. However, that aside, their confusion would have been caused and then compounded by all the contradictory instructions and non-standard phraseology. That was wholly incompetent controlling.

I have seen and read what is pictorially depicted above but unless you have the CVR/DFDR data, can anyone be so certain of this?

I didn't hear the EGPWS during any of their transmissions, although not to say it didn't happen.

Highspeedaluminium, if you look at the charts of their track and the terrain and obstacles, then the EGPWS would have been going nuts. Their terrain clearance was minimal judging by the data on those charts, and if accurate, then they must have merely followed ATC instructions and ignored the GPWS.

You are using information (in this thread) to suggest they narrowly missed antenna, terrain etc...and then conclude that they "IGNORED" GPWS warnings?
Case closed!

Using the FlightRadar24 .kml file (which is all we will ever see about this incident) and plotting it on 1:24,000 topo maps I see a distinct possibility they were paralleling Mt. Wilson flying east, and above the terrain off to their left, but possibly level with the main array of antennas. If their EGPWS did not have "peaks and obstacles" it is possible their EGPWS never went off. But, EVA isn't talking and the FAA's precise radar tracks will never be released to the public.

When faced with an unfamiliar term, the natural tendency is to directly translate each word. It is easy to translate word for word, but meaning can be lost in translation.

Not made any easier with beeping cockpit warnings and an ATC lady shouting at you in Americanglish.

Highspeedaluminium +1
This tread is going nowhere and I am surprised it still goes on. Trial and mobbing by internet.
For those waiting for a CVR transcript , as the flight continued to destination I doubt you will ever see it ..but that said. today with smartphones , everything is possible...
Rest assured that Eva Air will investigate , and probably send the FAA a nice letter. The guys running the airline OPS are very good and quite sharp.

As to the " southbound" calls , this would have probably been queered and understood in a normal no-stress situation. But here the lady first confused the crew by mixing up instructions, then lost it when spotting the potential conflict with ACA, , then shouted , words like " What are you doing? " and her repeated " turn Southbound" calls. confusing even more the crew.
A simple call " EVA turn right heading 180" at any time during the event would have solved the problem.
That is the lesson to be learnt here , used standard phraseology, the more so when in an emergency situation . She could have even added " expedite" . nothing else.

I bet you the lady will go back to the sim with a few reminders , something will be learnt out of that one for everybody else in San Diego and it will be a bit safer for everybody as a result.

Honestly, where do you expect the thread to go? This is typical of every other similar thread, pilots (and often times, non pilots) minus the relevent data to do so attempt to play judge, jury and executioner.

DuncanF

With all the usual disclaimers about me being a non-pilot, although very familiar with the common spoken language etc, I would absolutely imagine myself being confused by this word in a moderately stressed situation where directions are of the essence.

A native Swedish speaker, I would have no problems with the "south" part of the expression, but I can easily see that the part "bound" would have me reaching for the more dwindling parts of my brain. "Bound" is part of the verb "to bind" which would make me hesitate for those extra seconds about the meaning of this instruction. While I have heard the word "southbound" before it is always in the context of "he is going southbound on the M4" that is, the person is already travelling in the south direction.

"Southwards" would be easier for me as a Swede to associate with an instruction to turn south because I am currently heading in the wrong (that is, north) direction. The use of "wards" is the same as in "towards" and thus is easier to understand for me.

These chaps were Asian and I do not know how much this might have affected their thinking. Usual disclaimers about Indo-European (esp. Germanic) vs Asian languages apply.

I agree but....... To use standard phraseology when stress levels are high you have to have been using it regularly during normal ops, so the lesson is to use standard phraseology all the time without the verbiage.

southbound is a confusing word, when you dont know its meaning, despite the fact that part of the word is recognizable. If a person does not know what the whole word means, disecting it might not help. I remember when I first time encountered it it was the name of the song and I had to look it up. Even though I suspected that it meant direction of some sort. And I wasnt preoccupied with some afterTO routine.
Every new word even among other familiar onces might throw you off because you realize that i might mean something completely opposite to what you think it would mean.

MrSnuggles

Except that in this case, it isn’t. The “bound” in “the train bound for somewhere”, or “homeward bound” or “south-bound” is an independent adjective that originally means ready (i.e. in a state of readiness), and has an etymology which has nothing to do with the verb bind.

Which I think only reinforces your point that it could cause even an advanced English speaker to hesitate if they weren’t familiar with the phrase!

framer Good point . Absolutely agree.
I am curious if the FAA as a whole , and not only in San Diego, will use this incident to change their ATC training syllabus, especially in their refresher courses.

ATC Watcher:

The more likely conclusion is that this particular controller was not performing to present FAA standards. A lady friend of mine who is a former center controller listened to the tape. Her comment was "lazy controlling." She said she worked alongside a couple of those, which drove her up the wall. She added they were both "white boys" so the fact the controller at issue is apparently black is irrelevant.

Yes as native English speakers it would not be confusing although its meaning is contextual as bound has several meanings from how you run, tied up with rope or travel plus others.

I have always found when teaching people with English as a second language always be direct stick to a standard set of words for understanding, even if bad English, no hidden meanings. So agree standard phraseology is the best and safest.

ATC is not there to pass the time of day. Clear consistent English to a known set standard that's why takeoff and departure are used now as 500+ died due to the mix up over takeoff.

The question do we expect the USA to change ? Don't expect so they have a very poor concept of what they call Aliens. Their cabin crew and airports are terrible when dealing with international travellers, I know of people who travel the opposite way round the world to avoid passing through a US airport, so don't expect ATC to change they just don't get what the problem is.