Anybody thought that fatigue may have played a part in this incident, who knows what these guys where rostered in the preceding duties?

Reasonable ? Just think about it:

T0 GA instruction given
T0+??s AC crew pushes throttles to TOGA
T0+?s A320 gets a positive climb rate
T0+ 10s (presumably): AC759 roars up 100ft above UA001
T0+ 13s: astonished UA001 reports the close call

According to the AVH quoting the Canadian TSB: AC-759 had already overflown taxiway C by about 0.25nm when ATC instructed the aircraft to go around !

Based on a quick calculation: when ATC instructs to abort landing, AC759 and UA001 are horizontally 8-10 seconds from each other (600-800m ie. less than 0.5 NM) with seriously converging altitudes. Tiny margin left at that point :sad:

And a 789 taxiing towards you followed by more airliners at night is a arguably a pretty visible sight. How could AC759 be so severely confused to the point to having to wait the ATC in extremis instruction ?

Looking forward to the full report.

Ground track suggests that's exactly what they were doing, the FMS Bridge visual approach which provides lateral and vertical guidance to touchdown.

I can't agree with this.

Ages ago, I was on a visual night approach to IAD from the southeast, runway 1C I think. When I didn't report the runway in sight, the control tower asked for that information. It wasn't that I didn't know where the runway was, but I did not have a direct view of the runway lights - so I responded "No". Shortly after that, they asked the question again. I was less than 3 miles from the runway threshold, but still didn't have the view I needed - so I said "No".

They're response surprised me. They turned on the approach lights. It looked like a huge bright ball heading straight for the runway - repeating its path every second or two. I can't image what the electric bill for that was. It certainly left no doubt about where the runway was. Anyway, I reported "Yes", thanked them, and landed as instructed.

There are certainly many more challenges in landing a large jet compared to my little Cessna 172. Then again, won't they always give an ATC flight a straight in approach?

At SFO, 28R has: ALSF2: standard 2,400 foot high intensity approach lighting system with centerline sequenced flashers (category II or III)

Who is going to do it? NTSB as the incident happened in the US or TSB Canada as the state of registry?

Of course, I'm dismayed by the reaction of many around the internet that this incident is either "nothing" or "something".
Technically, it's "not nothing". I say that as a qualified ontologist.

Here's the points to take home:
The crew's bodies were in that magic 2-4 AM zone, fatigued or fresh, they were at their circadian low.
Their call was classic cognitive dissonance: what they were seeing was not what they expected to see, and so they sought confirmation of what they expected to see.
They then went around successfully.

Here's what cannot be said at the moment:
*Whether the crew initiated a go-around of its own accord
The unidentified voice from UA1 cuts in immediately after AC's readback. There are three possibilities:
1. During the readback, the PF decided to go around. The information they were given did not match what they were seeing out the window.
2. When UA1 calls "They're on the taxiway", they hit that TOGA button right quick.
3. They waited for the GA instruction from the tower.

I can see any of the three scenarios happening.

*How high they were when they flew over UA1
They had to go around fairly close in, since UA1 would need to be able to recognize that they were lined up on the taxiway. UA1 would spot it before the tower did. I think I posted a few years ago when another flight finished its bad day by lining up on 28R instead of 28L, causing the aircraft about to line up on 28L to bail out onto C. In that case too, the aircraft at the end of the 28s saw it before the tower, and that makes intuitive sense.
In any case, it's not clear how high they were or where they were when this happened. Simon over at AvHerald says "100 ft and past the 'taxiway threshold'." Of course, it's his "journalistic" policy not to name sources, and he's been wrong with this fine-grained data before, so, in this case, I'll wait for the NTSB animation (since nobody's gonna be filming that).

*Whether any other action was taken to avoid collision.
I wouldn't condemn those stuck behind UA1; they had limited visibility, and there's only one Tower frequency at KSFO. The press likes to cite Tenerife, but instrumental in Tenerife was a blocked transmission. I'll also note that thankfully, everyone on frequency was more or less speaking the same language. But there's no saying from the evidence so far whether anyone did anything. Would you risk blinding the crew with your landing lights, for example?

Two other notes:

I believe that for a very brief period about a decade ago, 28R was 28C, and the taxiway was 28R for lighter aircraft.

KSFO has inaugurated a new tower since the Asiana incident, so they probably won't be operating the same old camera.

Did they not put a kink in the taxiway after the LGW incident, so as you could see it was not a straight runway?

Don't know about the TSB but the NTSB is certainly going to take a look:

https://twitter.com/NTSB_Newsroom/st...21284513959936

This would not happen if the crew use ILS or VNAV-PATH for cross reference to back up a visual at night.
My question is what strength was the Approach and RWY lights on. If low, and focused on the taxiway, I can see this developing. But to take it as far as they did is pure madness.
My guess is that AC needs to fire the chief pilot and the head of training and get some competent people that knows how modern aircraft works.
Or are they going to blame this one on Airbus as well?

I wonder were MOT is in all this and if they have AC under control.
Me thinks not.

Anyway , it ended well and we shall all take notes, Eh!

Circadian low
 

I would put my money on their circadian low being a factor in this incident.
Perhaps we should start adding "circadian" to our callsign if we are in it. Just like "heavy" or "super".

Something like this: "Air Canada 123 circadian, established on final runway 28R."

Google "Juliet Kink" (though it has since been straightened out).

Air Can(ada) We Blame Someone Else Please
 

Air Canada: isn't that the outfit that gave us the Gimli (I can't do simple arithmetic) Glider and the Let's Fly below the threshold in Halifax who-cares boyz and now Hope SFO ATC know where I am because I've not a clue. Long past time that TSB stopped covering up for their buddies.

Here's a couple of possibly pertinent SFO Notams in the traditional cryptic format:

From the Mercury News article "’11 seconds to impact’: Expert calculates how close SFO near-miss was to disaster" :eek::

SFO near-miss: Air Canada pulled up with 11 seconds to spare

Not a bad idea. Seriously. :cool:

I asked before but was ignored.

1/ was the 28R ILS switched on? If not why not?
2/ where the HIAL's and RTZL's switched on? If not why not?
3/ did the crew bother to tune the 28R ILS and check it.?

This should not happen, if a runway is equipped with an ILS and it's not U/S then it should be switched on by ATC and all landing A/C should tune it to use as a backup for just this very reason.
Especially at Airports like SFO with close spaced runways and taxiways.

It's should be common sense and airmanship.

1/. If the ILS was serviceable, yes, it was on. They don't just turn it off for no good reason.
2/. Approach lighting is set to an appropriate level for the prevailing conditions. If they are set to high and the vis is good, it can be distracting.
3/. In modern aircraft we don't "tune the ILS." That happens automatically. If they were going to land on 28R and the localizer was operating, I have no doubt that it was displayed in the cockpit. Being able to include that information in your scan when you are landing at 140 kias visually at night at 3 am against a confusing background is another thing altogether.

Good job to all those involved for averting what could have been a disaster.

if they were doing the Quiet Bridge Visual, just crossing the San Mateo bridge, the chart depicts joining the localizer...possibly there had visual contact and decided to continue visually..is this procedure in their FMGS database is an interesting question


update....if the Quiet Bridge is selected from the FMGS database and activated, the localizer is not tuned, and no ILS data will be displayed...

If you had seen the NOTAM several hours beforehand that the Approach Lighting System was out of service, even though it was for a runway one character different, would that be in your mind as you line up visually for what looks like the right hand of two parallel lit strips, knowing SFO has two parallel runways (it seems all the lighting for 28L had been turned off for works, not just the approach lights).

I also wonder why they were given this curved visual nighttime approach at a time of low traffic to a runway fully equipped with all the aids. I'll bet when they came back round again that was not done.

And you have to wonder how the Flight Radar 24 traces shown above, which is something put together by a bunch of enthusiast amateurs, can show quite clearly, based on real time information, that they were lined up on the taxiway, and not like the later parallel trace shows for the runway, yet with all the millions of dollars of kit the ATC staff in the tower didn't get to be aware of this.

This attached chart is dated, and not for use for navigation.

"I also wonder why they were given this curved visual nighttime approach at a time of low traffic to a runway fully equipped with all the aids. I'll bet when they came back round again that was not done."


This approach is used at this time of night because of the extreme noise sensitivity of the area. The folks on both sides of the bay have telephones and they use them.