Air India Crash

[squerble]

No and they’re not just buttons. You don’t switch them off by mistake.

I hadn't even considered this, it's unthinkable that anyone would take this action. I tried multiple different things in the sim but didn't even bother with this because there could only be one outcome.

I called pilot suicide early in the thread as it was practically the only thing that made sense based on the video evidence.

You must have at least considered it given it was mentioned a few times here. I think I even replied to you and explicitly mentioned it.

 

[Mr Jack]

It has quickly also been reported that a Special Airworthiness Information Bulletin (SAIB) (NM-18-33) from 2018 highlighted a possible issue where the switches didn’t lock in position as they should, and they were used on many Boeing aircraft models.at on this aircraft.

That two unlocked switches could vibrate to the cutoff position if not fully up is not unfeasible and would explain the pilot’s reactions.

This seems the most plausible explanation so far. A failure of the system intended to prevent accidental release of these switches combined with either vibration or getting caught on a sleeve or something.

 

[nightwish]

"The switches were returned to their normal inflight position, triggering automatic engine relight. At the time of the crash, one engine was regaining thrust while the other had relit but had not yet recovered power."
https://www.bbc.co.uk/news/articles/cx2gy78gpnqo

That's heart-breaking. Just a touch more altitude and they might have made it.

 

[adam cool dude]

I called pilot suicide early in the thread as it was practically the only thing that made sense based on the video evidence.

You must have at least considered it given it was mentioned a few times here. I think I even replied to you and explicitly mentioned it.

Agreed. Everyone loves a good conspiracy but I too thought it was pilot error but didn't think of suicide. Would be grim if true.

The chances of both switches being faulty at the same time and being both switched off within a second must be 1 million to 1. I also presume there is a checklist to check the function of these locking collars before take off. Even in the little Cessna I fly part of my pre flight checks is to make sure that the fuel valve is switched in the on position and the fuel primer is in the locked position before take off.

We just need to find out why it was done now.

 

[SoundsGood]

I'd be checking all previous flights by those pilots for any similar behaviour, if that data is available.

 

[Jokester]

The chances of both switches being faulty at the same time and being both switched off within a second must be 1 million to 1.

I thought the same, but it could be possible that they both failed at the same time due to something like a big enough jolt, eg, maybe moving the flaps handle a position or whatever

 

[danlightbulb]

The problem with this is that there may well be emergencies scenarios during flight that may require these switches to be operable, so soft or hard locking the operation out entirely could potentially also cause problems and sometimes with conflicts like that you have to rely on some level of trust in the operators not to deliberately misuse the equipment.

This was discussed on Mentour Pilots live stream last night. It was suggested in the chat that the switches could be interlocked with the thrust lever position, and he agreed with the suggestion, not able to think of a reason why not.

 

[danlightbulb]

I thought the same, but it could be possible that they both failed at the same time due to something like a big enough jolt, eg, maybe moving the flaps handle a position or whatever

It appears from the report and analysis of it so far that there was nothing unusual. Flaps was in position 5 which was takeoff position and wasn't touched. Gear was down and wasn't touched.

"Possible brain fart of the century" Mentour Pilot said last night on his live stream.

 

[Freakbro]

So are we saying with the switches being turned off 1 sec apart that it couldn't be done with 1 hand, implying it would have to be done with 2 hands simultaneously?

Surely that action would be pretty obvious to the other pilot or could it have gone unnoticed?

 

[NickK]

But investigators are also zeroing in on what they describe is an interesting point in the report.

It says in December 2018, the US Federal Aviation Administration issued a Special Airworthiness Information Bulletin (SAIB) highlighting that some Boeing 737 fuel control switches were installed with the locking feature disengaged.

While the issue was noted, it wasn't deemed an unsafe condition requiring an Airworthiness Directive (AD) - a legally enforceable regulation to correct unsafe conditions in a product.

The same switch design is used in Boeing 787-8 aircraft, including Air India's VT-ANB which crashed. As the SAIB was advisory, Air India did not perform the recommended inspections.

Source: https://www.bbc.co.uk/news/articles/cx2gy78gpnqo

 

[Kenai]

So are we saying with the switches being turned off 1 sec apart that it couldn't be done with 1 hand, implying it would have to be done with 2 hands simultaneously?

Surely that action would be pretty obvious to the other pilot or could it have gone unnoticed?

Unless the locking feature wasn't engaged as per potential from the bulletin mentioned earlier in the thread, then you could conceivably knock them both off pretty simultaneously if you brushed your arm past a bit too close imo

 

[Kill_Phil]

Unless the locking feature wasn't engaged as per potential from the bulletin mentioned earlier in the thread, then you could conceivably knock them both off pretty simultaneously if you brushed your arm past a bit too close imo

Yeah watching the video a couple pages back, it doesn't seem like a great idea to place them directly below the thruster

 

[Werewolf]

So are we saying with the switches being turned off 1 sec apart that it couldn't be done with 1 hand, implying it would have to be done with 2 hands simultaneously?

Surely that action would be pretty obvious to the other pilot or could it have gone unnoticed?

I was watching something last night where they said the switches were only polled for position once a second (I think for the flight recorders data),

So potentially they were moved anything from >0.1 second to 1.9 seconds apart. which means they could have changed position effectively at the same time.

 

[VaderDSL]

I’m sure it’s already been asked, but presumably the flight data recorder only sees voltages when switches are opened or closed? Could it be a freak electrical issue I wonder? Is there any scope for the switches to physically remain in the run position but the electrical signals to call for shutdown?

 

[NickK]

I was watching something last night where they said the switches were only polled for position once a second (I think for the flight recorders data),

So potentially they were moved anything from >0.1 second to 1.9 seconds apart. which means they could have changed position effectively at the same time.

That's a very good point and could still exonerate both pilots - it could point to a switch failure that occurred simultaneously but detected separately with spread.

Although the switches themselves may have been separately polled once per second, I suspect the recording of the fuel system diagnostics would be recorded on change or polled at a faster rate.

In theory the old analogue recordings of voltage levels was better in that it provided a non-polled recording which showed things like intermittents such as a connector coming loose. If a digital polling system only polls once per second it will miss these styles of events, and if the system is a digital recording of change of state event then it can overload the storage without averaging which is why a cost effective way is to record per second polls. I did some safety critical stuff at university around 1996 along with embedded control systems etc. I distinctly remember being told as part of the risk analysis module that the cost of losing a passenger on a flight was $37K, so a crashed plane with 120 fatalities would be far less than the cost of some features for the lifetime of the plane itself (including weight for fuel etc).

 

[Feek]

You must have at least considered it given it was mentioned a few times here. I think I even replied to you and explicitly mentioned it.

No, it was so far off my radar that I simply never thought someone would do that.

It's very sad.

 

[danlightbulb]

I’m sure it’s already been asked, but presumably the flight data recorder only sees voltages when switches are opened or closed? Could it be a freak electrical issue I wonder? Is there any scope for the switches to physically remain in the run position but the electrical signals to call for shutdown?

I doubt the switches are hardwired electrical relays direct to the engines. Surely they'll be linked to the software? So switch tells software, software does the action?

Or at least the interface with the data recorder would be via a software link? They can't hardwire a thousand switches direct to the FDR.

 

[adidan]

Black box also records the other pilot saying he didn't cut off the fuel.

"In the cockpit voice recording, one of the pilots is heard asking the other why did he cutoff. The other pilot responded that he did not do so."

BBC News - https://www.bbc.co.uk/news/live/cx20p2x9093t
Air India crash report: Fuel switches cut off before Air India crash, report says, as families ask if it was avoidable - live updates - BBC News


I wouldn't be so quick to write this off as human error or suicide.

 

[200sols]

Must have been more said in the cockpit between pilot saying 'I didn't' and pilot calling mayday 20 seconds later.

 

[ubersonic]

So are we saying with the switches being turned off 1 sec apart that it couldn't be done with 1 hand, implying it would have to be done with 2 hands simultaneously?

Surely that action would be pretty obvious to the other pilot or could it have gone unnoticed?

The theory that was being thrown about a lot yesterday was that because it required two hands it must have been done by the pilot monitoring (as the pilot flying had his hands full). With the switches being flipped back by the pilot flying (with a single free hand hence them moving four seconds apart).