A flight recorder was found 28 hours after the accident, so it's been in the AAIB's possession for nearly 2 days, but not a word has been said about its physical condition. I recall in other recent accidents that the physical condition has been publicised quickly, perhaps not least to manage the public's hunger for information.

This leads me to suspect that it is in good condition, has already been downloaded, what happened (if not why) is clear, and it is embarrassing.

There are only three kinds of root cause:

- a design weakness or fault, but this will probably take some time and analysis to establish;
- a maintenance fault by a less-than-stellar organisation in an aircraft which has been used as a Christmas Tree;
- a personnel fault.

The last two directly embarrass AI. I think we're now in the political positioning and CYA phase.

This is assuming too much. There are legal and technical reasons why the CVR may not have been downloaded yet or, if it has, why no information has been released.

In an ICAO Annex 13 investigation, interested states and experts are invited to participate. In this investigation, investigators from the US and UK are participating (there may be others as well), and experts from Boeing (and possibly component manufacturers) have also been invited. The lead investigators will often hold off on recovering data from the CVR and FDR, or examining the data, until there's agreement between the interested parties on how to go about doing it. With the distances involved, travel logistics could easily delay examining the data by a couple of days.

In some cases, the board leading the investigation doesn't have the internal expertise to recover data from the CVR and FDR. In these cases, agreement needs to be reached on which technical experts are going to take the lead on recovery. I don't know if the Indian AAIB has internal expertise or if they need to find external expertise to recover the data.

In some cases the CVR and FDR can be damaged and technical experts need time to come up with a plan to repair the recorders and recover the data.

Annex 13, Article 5.12 prohibits the release of CVRs and analysis/opinions from FDRs, unless the investigators decide that the benefits of the disclosure would outweigh the adverse impact that release would have on future investigations. This rule exists to encourage full cooperation in investigations from participants in the industry, including pilots and operators. India's accident investigation rules are consistent with Annex 13: see section 17 of the Aircraft Investigation of Accidents and Incidents Rules, 2017. So even if the CVR and FDR have been analyzed, it may well be that information isn't being disclosed to the public because they haven't reached any definitive conclusions yet that they're confident will end up in the final report.

OK, technical questions for those who might know the answers: I think many on this thread are expecting the DFDR data to give clear answers to what happened to this flight, but due to the likely electrical power loss, what would actually have been recorded? AFAIK the 787 DFDRs have an internal battery but if the power is off to the rest of the aeroplane, what data, if any, is going to make its way to the units? Is anything recorded while on battery power? Does the RAT coming online after a pause restore some of the recording functionality?

You answered your own question. The units would have had power. One pilot made a radio call so electrical power was available at the very least at emergency level.

It has been established that the B787 utilises a EAFR, a combined CVR and DFDR.

from the little I know:
* only the forward EAFR (Enhanced airborne flight recorder) has an (external) RIPS (Recorder independent power supply)
* and this RIPS provides, in addition to the forward EAFR, the cockpit area microphone and the preamplifier for this microphone with 10 minutes of backup power

Pretty recently, Indian media was reporting that the forward "CVR" has been recovered. Pure speculation..but one might interpret the use of this incorrect terminology and the continued search for this second 'black box' while the rear EAFR should have been easily recovered from the mostly intact tail-section, as an indication that things aren't going as well as expected regarding FDR/CVR data-extracting (as in rear EAFR lacking the last x crucial seconds, and forward EAFR now essential for recovery of the last x-seconds of area-CVR recordings, thanks to RIPS)

My expectation is that at the very least the AAIB (India) would want to open / read in the presence of the AAIB (UK) and the manufacturer. They might even want to read it at a location with specialist equipment (NTSB in Washington, BEA in France, AAIB in UK). None of this would be unprecedented when looking at previous accident investigations particularly for those that took place in developing countries.

Quick example, EK 521 accident in DXB:

Pakistan A320 accident Karachi:

(1st June was 10 days after the accident).

You answered your own question. The units would have had power. One pilot made a radio call so electrical power was available at the very least at emergency level.

It has been established that the B787 utilises a EAFR, a combined CVR and DFDR.

from the little I know:
* only the forward EAFR (Enhanced airborne flight recorder) has an (external) RIPS (Recorder independent power supply)
* and this RIPS provides, in addition to the forward EAFR, the cockpit area microphone and the preamplifier for this microphone with 10 minutes of backup power

Pretty recently, Indian media was reporting that the forward "CVR" has been recovered. Pure speculation..but one might interpret the use of this incorrect terminology and the continued search for this second 'black box' while the rear EAFR should have been easily recovered from the mostly intact tail-section, as an indication that things aren't going as well as expected regarding FDR/CVR data-extracting (as in rear EAFR lacking the last x crucial seconds, and forward EAFR now essential for recovery of the last x-seconds of area-CVR recordings, thanks to RIPS)

Something that occurred to me after I went to bed last night: My assumption that the FDR readouts would rapidly reveal the cause may be flawed.

Let me explain.

The consensus is that both engines quit shortly after liftoff (that assumes that the RAT did in fact deploy). At least one of the data recorders has battery backup, so it should have kept functioning when all aircraft power was lost.

However...

Over the years, I've looked at lots and lots of digital flight data recorder outputs when investigating some sort of incident or other engine anomaly, So I have become rather familiar with some of the interesting characteristics of DFDR data.

On the 767 and 747-400, when you shutdown an engine and the IDG goes offline, there is a momentary 'glitch' in the electrical power system as it reconfigures for the available power source - this is why you see the flight deck displays flicker and return, and the cabin lights momentarily flicker. As a result, most of the avionics boxes 'reset' - this is quick, but it's not instantaneous. This shows up in the FDR data - sometimes as 'no valid data' for a few seconds, or as garbage readings of zero or 'full scale'. Now, looking at the FDR data, it's easy to simply disregard the data, so normally no big deal.

Starting with the 777 (and on the 787 and 747-8), this electrical power glitch was 'fixed' - there is slight delay (~quarter of a second IIRC) before the fuel cutoff signal is sent to the engine - during which the electrical system reconfiguration takes place so no more 'glitch' during a normal engine shutdown...Except whatever happened to these engines wasn't 'normal'.

If there is a fuel cut at high power, the engine spools down incredibly rapidly - a second or two from max power to sub-idle. Assuming the fuel cut wasn't commanded by the flight deck fuel switches, the electrical system won't know it's coming, so it can't reconfigure until after the engine generators drop offline - and you're going to get that power glitch. Nearly every avionics box on the aircraft will reset due to this electrical glitch, and the FDR isn't going to get useful data for a few seconds (and then, only from the stuff that's on the battery bus).

Whatever happened, happened quickly - it's quite possible that whatever initiated the high-power fuel cut didn't get recorded.

True, I would bet that the voltage regulating architecture of the voice recorder at least will give useful information for a short time around the loss of power, as it did for the Lockerbie CVR using far less sophisticated recorder systems. That was sufficient to show the pressure pulse in the fuselage and to give a fairly good idea of where the explosion had emanated from. TWA was similar, a bigger bang though. Would be listening for any mechanical noises related to the fuel switches, and frankly I doubt that they existed, but they would be recorded.

If the cause is what I have suggested it will dificult to get direct evidence of that case, as it was for the QFA072 event as well. Like icing cases, a water ingress into the avionics is going to be a tough investigation, water would have been sprayed all over the wreckage in the aftermath. Dousing the E/E bay with 20 or 30 gallons of water will be an expensive investigative exercise to do in a real plane, with engines running. Would not want to be observing up close.

if not on the rear EAFR, then all hope is on the forward EAFR being downloadable

Thanks as always TDR for your excellent professional input. It is therefore so much more perplexing that even you cant logic our way out of this impasse. That is, the assumption that the aircraft experienced a double engine failure (supported by a reasonably convincing argument that the RAT deployed), and yet no plausible reason (that we can see) for such an event. So some then collectively slip into the tired and lazy theories of intentional or unintentional crew actions that 1. beggar belief (intentional), 2. defy physics (flaps instead of gear despite clear evidence to the contrary) and call into question the professionalism of a very experienced Captain and crew as well as the aircraft manufacturer (because...well its Boeing so it must be software ).

Yet, the answer must be simpler and staring us in the face since logic and experience (everything you have offered TDR), tell us that modern airliner engines generally do not just suddenly quit flying at the same time. In this regard we can recall several instances of double engine failure associated with bird strikes generally involving large birds or large flocks or both. But it seems we have discounted this theory very early in discussion. Why? Because we cant see any birds, or flocks of birds or engine flames/surges or puffs of smokes from the engines which would support this. Really?

I have read all the 100's of posts (sadly) and while some very early posters tried to analyze the imagery, I suspect the very poor quality eventually discouraged most from seeing anything of interest. However, smattered throughout this discussion from the beginning to the end there have about four posts that describe seeing something where others have not. At least two of these were related to possible smoke but which were probably just the dust blown outwards by the wingtip vortices. Two others however have mentioned possible flames and puffs of smoke.

The video of course is very poor. There should be a special place in hell for people who subject us to looking at a video with continuous zooming in and out, inability to retain focus on the subject (it was just a CCTV monitor, not the actual aircraft they had to focus on) and constant camera shake. A video of a video, and then the resolution probably reduced for social media upload. This all results in a very unwatchable record of the aircrafts departure. The only immediate information gleaned seems to be some idea of how far down the runway the aircraft was at takeoff and the parabolic curve as it very clearly described the aircrafts flight path.




Air India Flight 171 on departure
But take a look at this frame. The right engine shows an artifact (pixelation if you like) that might represent a surge flame. I can almost see a puff of smoke just inboard of the aileron that may be associated with that too.

Am I just seeing distortion? Am I just seeing some smoke because that's where I would expect to see it?
We are all very used to seeing everything in 4K today but back in the day when everything was low res we used to join the dots. If pixels existed then something was there. If they didn't, it wasn't.
So if it's just pixels caused by distortion then they have coincidentally appeared in the tailpipe of an aircraft that crashed shortly after takeoff with a presumed double engine failure.

But surely we would see the birds? Well, not in this video. You cant even see the registration number on the side of the aircraft and that is much bigger than a bird. Haze, distortion, focus and low res, and each individual bird wouldn't even make up a pixel.

So make of this what you will, but this problem may have started on the ground. Birds strikes are very common according to Some AI pilots who interviewed for this following article but I have no idea of the authenticity of this report:

https://www.rediff.com/news/report/a...h/20250613.htm "The Air India pilots also added that Ahmedabad airport has long been known for bird activity near the runway, which could have contributed to the incident.

"This issue (of the excessive presence of birds) has been flagged multiple times," a third Air India pilot said, asking not to be named."
Of course, a single engine failure would not have brought this aircraft down, nor would it have deployed the RAT, but we can't see what happened on the left engine when the aircraft slipped behind the radio antenna building.

While these high bypass engines are designed and certified to keep running after experiencing certain types of bird strike, the effect on two engines concerns have been voiced about the contribution of certification to the mitigation of the risk hazardous bird strike in the two engine case.

This from Sky Library:
https://skybrary.aero/articles/aircr...nue%20to%20fly .

" A number of concerns have been quite widely voiced about the contribution of certification to the mitigation of the risk of hazardous bird strikes:

• The case of bird ingestion into more than one engine at the same time is not addressed directly and it is clearly extremely difficult to meaningfully estimate the probability of such an occurrence. However, it has been observed that, since some of the current standards only require that a damaged engine can be safely shut down, this circumstance should be more fully considered when determining the acceptable outcome of ingestion into single engines, especially for the twin engine case.
• It has been noted that the potential effects of bird strikes on modern electronic flight control systems and flight deck instrument displays have not yet been fully assessed.

Maybe someone can do some video enhancing of this image as others have done with the audio enhancement to give strong probability of RAT deployment.

If my suggestion can be corroborated at all, then the question of what happened next becomes somewhat easier to answer. Perhaps neither engine stopped running but they did so with limited thrust? If anything from the pilots mayday call can believed, it wasn't engines shut down..it was no thrust. So why did the RAT deploy? Cant answer that. And, I cant imagine it would be manually deployed if both engines were still running.
However, TDR did say.

"On the 767 and 747-400, when you shutdown an engine and the IDG goes offline, there is a momentary 'glitch' in the electrical power system as it reconfigures for the available power source - this is why you see the flight deck displays flicker and return, and the cabin lights momentarily flicker."

Startle factor that electrically systems were about to fail? Manually deploy RAT?

Edit: I might add, they would have found remains on the runway if this did indeed happen. But we have heard anything from anybody?

The flight recorder on the 787 is called "enhanced airborne flight recorder" (EAFR), and it's a combination unit with CVR and FDR functionality. There are two on the aircraft, the one in the front has battery backup.
Since it has been reported that the "FDR" was found, they also have the "CVR".

B787 Skipper.... No longer able to sit on my hands!

The joker that published one or other of the fictional accident reports before the dead are even identified needs stringing up by the thumbs.... Well done mods. OK, some of these theories, just bearing in mind the the flight recorder (there is only one FDR + CVR combined) was in the intact section of the undamaged tail and has been with the authorities for almost five days......

- All you guys who are rushing down the TCMA rabbit hole: If it was established that a software error drove both engines to idle without warning, after rotate.... Don't you think the worldwide fleet of B787s would have been grounded by now? Such a glaring failure would be absolutely inescapable on the FDR.

Whatever was wrong with this aircraft was present at rotate, unbeknown to the crew. The fact that no ADs or notices to operators have been issued usually means that the cause is known and the aircraft was serviceable. The statement from a prominent Indian Captain about the skill and tenacity of the crew, right up to the last minute is absolutely laudable. However, the cynic in me says that the way is being paved for some bad news and by that I mean news that will do Air India reputational damage. Expect more management of expectation in the coming days.

I'm still going with
a) Incorrect derate + low Vspeeds or
b) Low altitude capture

Do you stand by your previous assertion :-

a) 🤷‍♂️
b) Wouldn\x92t it capture the Alt rather than descending thru it?

Maybe you should have kept sitting on your hands, theory a) in particular is about as risable as the last one you posted. Do you still think engine failure takes 30 seconds for the plane to react?! Thanks to rkenyon for alerting us not to take your post seriously. You still don't believe the RAT was out despite the in depth audio analysis done by an expert in the field. You still think its a motorbike?!

Given they took off at a perfectly normal point, at a perfectly normal speed, I don't see how you come to incorrect derate and low v speeds, or how on earth that would lead to a normal takeoff followed by sinking?

And they do have a new lab, as of this past April:

AAIB sets up flight recorders laboratory

More at linked site. The lab opening is also featured in the slide show on the top page of the AAIB website, following a series of slides for International Yoga Day.

Given the number of times I reviewed DFDR data supplied by an operator after some sort of event/incident, I think most major operators have access to the equipment needed to download a healthy data recorder. So I'd be a bit surprised if Air India does not have this capability. OldnGrounded has also posted that the Indian AAIB also has that ability.

Usually when I hear of data recorders going back to the US NTSB or the recorder manufacturer, it's because the crash damage is such that specialized equipment is needed to download the data. The recorder in the tail would likely have little damage.
While the AAIB may have held off on downloading the recorders until all the major players are present, it's been several days - I'd expect everyone who matters is already there. So I think it is reasonable to believe that the investigators have done a download and have had at least a preliminary look at the data. If there is a smoking gun, they probably already know (and the longer we don't hear something regarding the rest of the 787 fleet, or at least the GEnx powered fleet, the less likely it is that they suspect a systemic problem with the aircraft and/or engine). However the proviso that I posted earlier about potential data loss/corruption due to a sudden shutdown still applies - so maybe the data simply isn't on the recorder.

As has already been posted, EMI is highly unlikely - the current cert requirements for HIRF are quite high, and due to the composite airframe construction of the 787, the lighting requirements are much higher than for conventional aluminum aircraft (the higher resistance of the composite airframe results is higher lightning induced currents).

FDR has suggested a large slug of water hitting critical aircraft electronics at rotation - it is possible that resultant electrical short circuits could falsely signal the engines that the switches are in cutoff. Highly unlikely that it would do that to both engines, but possible.
Then again, all the other plausible explanations are highly unlikely, so...

BTW, I do have a life outside PPRuNe - and I'm going to be traveling the next several days, with limited to non-existent internet access. So don't be surprised if I'm not responding posts or PMs.

Boeing's patent application says: My understanding is that "EEC" is Boeing for "FADEC".

But I have to reiterate, and as others have noted, that we don't know what actual TCMA systems, if any, were fitted where and how it was wired into the accident aircraft's engines, or whether the specific state of the TCMA systems, if any, were monitored by the flight recorders.