That's something which is bothering me. Why did they not continue on the STAR? What did they spot for them to decide to go hands on? And that fans the Juhu theory.

As are we and that's what the hubbub's about and it would be interesting to find out.

Fair enough. But we just want to know what they were heading for... if it wasn't Juhu, then what :ugh:

:)

The above image clearly shows that the a/c is heading for Juhu rwy 08 at 04:55:58 UTC! Whether they spotted Juhu or not is of course, debatable.

Reading this thread and analysis, I am losing confidence on avherald. They appear to jump to conclusion, always in favor of the airline involved before any analysis.

Did ATC ask to go around or crew did on their own?

And pilots choose to defend their kin. Perfectly fair. No offence.

Interesting question...

What if the crew were new to Mumbai and/or unfamiliar with rwy 09 since 27 is the one usually in operation...?

It's all very well that the crew realised that they were off course and performed a go around (or were advised to) but crew new to Mumbai could learn from knowing what happened in the first landing attempt. Isn't this what aviation safety is all about and the purpose of forums such as this...?

The OP's comment was made about Avherald, which isn't run by pilots, and not in respect of this forum.

And with some justification - Avherald is fine when it comes to reporting facts, but when he attempts to analyse the how/why rather than just the what, Simon frequently comes unstuck.

IC. I'm with you. Rather than focusing on the what they did, yes/no, it'll help us more to find why & how. This is quite a new a/c. If there is a glitch in the automatics, or the human interface, or a misunderstanding of the nav system then we can all learn from it; even the manufacturer. If it was a case of mistaken identity and a crew rushing, then we need to know. What strikes me is that JUHU is just over the beach, while their true target is much further inland. Clue 1. We have to assume that the turn off course was manual and because of a visual target; otherwise why would you? Once they were pointing at JUHU the vertical guidance on the MAP display must have been interesting. Given the displacement of JUHU off their target centreline this suggests a lack of confidence in the navigation set up for their initial approach. I'd assume that BEWARE of JUHU would be in an approach/airfield brief. Surely? And in this day & age of RNAV approaches who would expect to be so far off track on final approach? Surely that thought would ring bells and ask questions. Bad smoggy vis at Mumbai is well known, so diving off track at an early DME could be deemed a little precipitous.

RAT 5, I’m so glad you brought up the cliche about it being a new a/c, with regard to any glitches with equipment. To this I would add that Airbus a/c from the A320 onwards require a higher dependence on instruments than others and crew are trained accordingly.

The Jepp app charts for Mumbai have a caution clearly marked for rwy 09: “Do not mistake Mumbai (Juhu) for Mumbai (Chhatrapati Shivaji Intl.)” and so as you said, this would, we assume, be in their approach briefing. Then why the deviation to the left of their original app track?

There’s much to be desired in terms of what happened on the first approach attempt.

Coworkers have been frustrated at a FMC/RNAV approach that lines up with the edge, or slightly outside the edge, of the runway at one of our airports. A 16-17 degree turn, on a 3-4 mile final, is a completely different ballgame.

This event is not a normal everyday occurrence. But I’d guess that every professional pilot has picked up the wrong visual ques more than once in their career. Some catch it early enough and never bite at the wrong choice, some catch it shortly after they deviate from the desired flight path, and some catch it later. Debrief it and move on. Don’t be surprised if it gets discussed in recurrent training.

I defer to the experienced drivers of big iron, but ... do the data really show that the aircraft was "heading for Juhu"? Certainly it seems to intercept the extended centreline of Juhu 08 at 04:55:58 - but at a point only about a mile from the threshold. Up to that point it's been about 15° off runway heading and well south of the centreline. Is that probable on a visual approach in an A350? (For them to be "heading for Juhu" they'd have to be making a purely visual approach - even if there were any navaids at Juhu, any navigational system the crew were looking at would surely have shown them the way to their real destination.)

.....any navigational system the crew were looking at would surely have shown them the way to their real destination.

Hence the curiosity about the deviation. Much depends on crew techniques and SOP's. I've seen it where PF says, "can you see the runway yet?" The crew is visual with the ground and flying on automatics for the approach. Both pairs of eyes are then focused outside the window: one calls, "I think it's over there." PF confirms he has a runway in sight, disconnects and aims at the new target. Human behaviour, rather than system malfunction.
It has happened where a/c have landed at the wrong airport, even with a MAP. The real target runway showed large deviation from the a/c symbol while on finals for the wrong runway. Only picked up if someone is managing the office inside.
There is a Glideslope aural/visual warning. The lateral & vertical deviation is displayed on the FMC, silently. Someone has to monitor it. Are we going to introduce another aural warning to alert crews they are not doing their basic job?
I'm not saying any of this was the case here, hence our curiosity why a crew would disconnect an automatic approach so early on a bad vis approach without being certain of the visual reference. The geographical position is quite different, and I suspect the approach lighting is also, and the relevant DME from touchdown would have given a clue as well.
Remember the AC A320 crash in Halifax where there was no SOP to monitor DME inside the FAF. Similarity here? Situational awareness? It is such an odd action from a professional crew on a grand new a/c that the management needs to know the full story. It is not to be dismissed as a non-event. As with AC A320 event it could lead to altering SOP's.

If the crew had done nothing the plane, even with the autopilot off, would have overflown Mumbai airport. It took human intervention to offset it.

This https://aviation-safety.net/database...?id=20150304-0 is a nice testament of how accurate onboard FMS position is. (And of very opaque smoke screens and mirrors as well).

Given the newness of the aircraft it is possible that, as a contributory factor only, this was a line training flight as well, which would add a further dimension, namely the point at which the TC would intervene. Pure speculation on my part but a possibility nevertheless.


I've flown the B767 and B744 into Mumbai many dozen times and the 09 approach was always problematic, in my day the inbound track was on a VOR radial from the BBB and subject to radial instability, hopefully cured by now. Might help if we knew exactly which approach SQ were doing, particularly if it was a training flight.

Do you mean that the crew were being familiarised with the rwy 09 app (with a deliberate deviation to show them Juhu)? But then how would the TC know that rwy 09 would be operational on that day...?

No other a/c deviated from their app track that day so as to point to a probable radial instability from BBB VOR...

Pure speculation from me as well.

This seems to be an “Airbus specific” issue, very probable in the SIA incident. If this was the case, kudos to the crew for realising that the FMGS was giving them wrong information and taking evasive action.

A couple of thoughts, looking again at the detailed track of both the original approach, and the second, successful, one.
Both approaches seem to reach exactly the same point west of the shoreline, seemingly following a direction of travel very sympathetic with an approach to the international airport ( and most certainly not Juha).
The original approach turns left, and travels in that direction for some 35 seconds before adopting a course which is again sympathetic with the location of the international airport.
The reason for that left turn is not known to us, and I have no new theory to add to the speculation, but I would observe that if indeed the crew mistook Juha for the international airport based on the view out of the window, they would surely have been able to conclude immediately that a landing would not have been viable from their altitude. Had they nevertheless believed that it was, we would have expected to see more significant altitude loss during the 35 seconds than there was.
Also, and at the risk of sounding old fashioned, surely it it not at the crews discretion to make as significant a deviation from their course so close to an airport (s) without ATC clearance?
Finally, is it of interest that on the second approach, they still approached from left of the extended threshold?

That track is to be expected, it is the VOR approach after all.

The VOR sits south of the runway so a radial that intersects with the runway centerline at the proper distance from the runway (at minimums) is used.

You can even see when they saw the runway and corrected toward the runway centerline the second time.

Becoming visual with the runway is the only reason to deviate from the VOR radial. Therefore the assumption they were aiming for Juhu is very reasonable. Of course they should notice the runway they see is nowhere near straight ahead, but if as mentioned the map was not monitored it makes sense.

There is a tendency to go visual early on VOR approaches even if the track is often made as to make relatively late turn onto centerline possible. But most crews want to be done with the turn before being low so they need to first turn towards the centerline and then back onto the centerline instead of doing one turn when the radial intersects the centerline.

VOR22L at JFK is a good example and has good FR24 coverage if anyone wants to watch the wildly differing maneuvers on that approach (Mixed with the RNP Z 22L sometimes which aligns you with centerline earlier than the radial/centerline intersect on the VOR). This is anecdotal and i have not pulled the recordings at the time but i've heard two 747s make their final approach to 22R because they went visual very early. One went around after being instructed to go around while the other one the tower controller even cleared to land on 22R. He still made a go around which is the thing to do if you are approaching a runway you didn't brief for.

Overall this "incident" seems to be an error which was noticed very early and i'm sure does not deserve the press cover it got.


JFK VOR 22L approaches from several aircraft with runway centerline and VOR radial in black lines for illustration purposes :)

FMC aircraft don’t suffer from ‘radial instability’ on VOR approaches.

Want to see interest approaches? Look at JFK VOR 13L approaches. Nothing like seeing arriving aircraft climbing from 400’ at DMYHL on the TCAS when the minimums are 800’. Minor details I tell you, minor details.

Just a question from SLF.

What would the possibility be of the pilots using the wrong input dial for the go around height. ie 3000ft and instead using the heading selector 300 making the aircraft turn.

Is this at all possible or would the auto-land function over ride any manual inputs.