←back to thread

Preliminary report into Air India crash released

(www.bbc.co.uk)
346 points cjr | 2 comments | 11 Jul 25 20:23 UTC | HN request time: 0.458s | source
Show context
nottorp ◴[12 Jul 25 07:42 UTC] No.44540123[source]▶
Almost 400 comments and no avherald link for reference?

https://avherald.com/h?article=528f27ec

replies(1): >>44540922 #
potamic ◴[12 Jul 25 10:23 UTC] No.44540922[source]▶
> On Jul 12th 2025 (UTC) India's media report that the investigation is NOT focussing on a human action causing the fuel switches to appear in the CUTOFF position, but on a system failure. Service Bulletins by Boeing issued in year 2018 recommending to upgrade the fuel switches to locked versions to prevent inadvertent flip of the switches, as well as the FAA/GE issued Service Bulletin FAA-2021-0273-0013 Attachment 2 relating to loss of control issue (also see above) were NOT implemented by Air India. The stated MN4 computer with faulty soldering, that might weaken and lose contact due to the thermal stress after a number of cycles, interprets data and commands fuel metering valves - with the lost contact attaching the MN4 processor to the EEC intermittent electrical contact, loss of signal processing and engine control faults can occur. The SB writes under conditions for the SB: "An LOTC (Loss Of Thrust Control) event has occurred due to an EEC MN4 microprocessor solder ball failure." According to discussions in the industry it may be possible with the number of cycles VT-ANB had already completed, the solder balls were weakened sufficiently to detach the MN4 from the EEC momentarily due to loads during the takeoff rotation leading to the loss of control of thrust and shut down of the engines.

Still quite early in the investigation, and so many things to consider. I don't know why online communities have been so quick to gravitate towards the murder/suicide theory. I thought aviation enthusiasts of all people would want to keep an open mind until every other possibility is ruled out, however minuscule it might seem.

replies(4): >>44541015 #>>44541663 #>>44542123 #>>44542484 #
1. labcomputer ◴[12 Jul 25 13:58 UTC] No.44542123[source]▶
> I don't know why online communities have been so quick to gravitate towards the murder/suicide theory.

Because the hardware failure theories seem preposterously far-fetched and require an unnecessary multiplication of deities.

Your ghost in the machine needs to be “just so” so that it can cause both switches to be read in “cutoff” nearly simultaneously. Then, 10 seconds later one of the switches needs to be read in “run”, then 4 seconds after that the second one needs to read “run”. You also need to explain why there have been zero single engine failures of this type before this double failure.

The ghost also needs to explain why one pilot asked the other “why did you cutoff?” instead of something like “what happened to the engines?” (which is the more natural response, unless you already know the switches are in cutoff).

replies(1): >>44544214 #
2. potamic ◴[12 Jul 25 19:00 UTC] No.44544214[source]▶
There's also maintenance lapses, faulty repairs, defective parts, and as far as software goes I can think of n number reasons how a ghost can manifest itself inside program logic. This is a new gen plane that relies more on software than any other before, and has in fact seen a couple of incidents with loss of thrust, both related to software. I think it's more prudent to be asking hard questions around these than to outright dismiss it as an open and shut case. Besides, the murder/suicide angle is the least interesting outcome. Because there's nothing you can do after that, other than to just move on.