Rachel Chitra
India’s deadliest aviation disaster, Air India flight 171 that killed 260 people, likely suffered a systems‑driven failure, as per new simulator data and the India’s Aircraft Accident Investigation Bureau’s (AAIB’s) own timings.
Fresh 787 Level‑D simulator tests commissioned by the Federation of Indian Pilots (FIP), which represents 6,000 pilots, show that the aircraft’s emergency power – a small generator called the ram air turbine (RAT) – would have needed around 10-14 seconds after engine failure to begin supplying hydraulic power.
And yet the AAIB preliminary report, released on July 12, records RAT hydraulic power as available by 08:08:47 UTC – just eight seconds after liftoff at 08:08:39 UTC. That implies the RAT deployed in response to an earlier electrical disturbance, not as a consequence of the engines flaming out after a fuel cutoff, suggesting a cascading systems problem was already in progress before thrust was lost.
What the findings show
Captain Randhawa, president, FIP, told The Wire, “In plain terms: if RAT hydraulic power was already available at 08:08:47, then the RAT must have been triggered before the engines fully failed. That suggests a prior electrical or systems disturbance was already underway, serious enough to trigger emergency power, well before the main engines dropped through idle.”
He added that this undermines the simple picture of “engines fail → RAT comes out”, and instead supports a sequence where degraded electrics and core network issues destabilised the aircraft, and the loss of thrust followed. The core network was marked a medium risk on June 9, 2025, three days before the crash, as per AAIB report.
The aircraft’s Core Network System was one of four Category C Minimum Equipment List (MEL) items active on the aircraft as of June 12, 2025. According to the AAIB preliminary report, the MEL was invoked on June 9 and remained valid until June 19, meaning the aircraft was legally permitted to continue operating while maintenance action was deferred within the prescribed rectification period.
An MEL is a regulator-approved document that specifies which equipment may be inoperative and for how long an aircraft may continue flying before repairs become mandatory. Category C MEL items generally allow rectification within a specified number of calendar days, reflecting an assessment that the aircraft remains airworthy despite the deferred defect.
The Core Network System is the Boeing 787’s digital communications backbone, linking numerous aircraft systems and enabling the exchange of data between flight-deck, maintenance and aircraft-control functions. The FIP, in a submission to the Supreme Court, has argued that regulators should re-examine whether defects involving the Core Network System should continue to be treated as deferrable MEL items, given its role as a primary communications pathway connecting multiple flight-critical and non-flight-critical systems across the aircraft.
Captain Randhawa said the pilots’ union commissioned the simulator tests after repeatedly calling on the AAIB, since July 2025, to conduct similar evaluations as part of the Air India 171 investigation. According to the pilots’ body, those requests were not acted upon.
“Our objective was to examine whether the sequence of events described in the AAIB’s preliminary report could be replicated in a real-world simulator environment, in the interests of aviation safety, accident prevention and public confidence in the investigation process,” he told The Wire.
The Wire has asked the AAIB whether it received these requests and considered holding such simulator tests, or whether such tests were held and what their findings were. This article will be updated once a response is received.
Why the AAIB’s narrative doesn’t work
Foundation for Aviation Safety executive director Ed Pierson, who has submitted key evidence on both the Max 737 crashes and AI 171 to the US Senate and AAIB, told The Wire, “Crash flight VT-ANB [the crashed AI 171’s registration number] had a history of dangerous electrical and electronics defects. We made multiple attempts to share documentation with India’s AAIB that proves this fact, but they were completely unresponsive. Fortunately the UK’s AAIB passed the documents to the India AAIB Investigator-in-Charge on the foundation’s behalf.”
He added, “The US NTSB (National Transportation Safety Board) showed little to no interest in the documentation. This was very concerning and we felt they were stalling, so we bypassed them after the UK AAIB offered to do it.”
Against that backdrop, the engine story also now looks very different to the popular Western media narrative of a pilot cutting off fuel switches.
The AAIB’s report says there was a fuel cutoff, followed by both engine cores, measured as N2, passing below minimum idle at about 08:08:47 UTC, eight seconds after liftoff. It then describes a relight, where “Engine 1’s core deceleration stopped, reversed and started to progress to recovery” and how Engine 2 “was able to relight”. The report implies this relight happened after the fuel control switches were cycled back to RUN by the crew at 08:08:52 and 08:08:56 UTC; the 13th and 17th second after liftoff.
In essence, the AAIB report (without giving timestamps for relight) implies a neat causal chain: pilot moves switches to CUTOFF, engines wind down, pilot moves them back to RUN, engines start to recover. But engine maker General Electric’s engine manuals and the FIP’s new simulator runs show that this chain is backwards.
On modern turbofans like the GEnx — the Boeing 787’s main engine — fitted to VT‑ANB, once N2 or engine core speed falls below idle, these large engines cannot simply “come back” on their own. They need a massive amount of energy – from an auxiliary power unit (APU). On AI 171, the APU was not online.
The AAIB’s own timeline has the APU inlet door only opening at 08:08:54. And engineers say it would take another 50-60 seconds before APU would start generating power. In a flight that lasted just 32 seconds from brake release to impact, that makes the APU a bystander. “Its contribution [is] zilch,” said an Air India engineer.
The RAT, as noted earlier, can help with hydraulics but cannot drive the heavy core of these large turbofan engines back up from idle. GE manuals and engineers say that the RAT can have a role in a relight if the plane was flying fast enough at 250-300 knots, so that the engine blades are spinning fast enough that a small spark provided by RAT is enough to get the engines back on. AI 171, unfortunately, was at 180 knots at 08:08:42 UTC, three seconds after liftoff, and fast decelerating.
The simulator runs, commissioned by FIP to mirror the AAIB timeline, back up what the GE manuals show. Across nine documented tests, the N2 values at the moments tagged “1 Eng Start” and “2 Eng Start” – aligned with the 08:08:52 and 08:08:56 marks – sit in the 32-45% band, well below typical self‑sustaining idle of the GEnx engines.
Fuel flow at those points, commanded by the engine computer i.e. the Full Authority Digital Engine Control (FADEC), ranges from 0.0 to about 0.6 – essentially no sustained combustion. FIP’s simulator data shows that the engines’ cores are too slow or they are not rotating fast enough and the fuel is effectively off. There is no genuine relight.
Taken together, this creates a serious conflict between the behaviour implied in the AAIB’s report and what both GE documents and simulator tests show the engines can actually do. If a crew‑commanded restart at 08:08:52 and 08:08:56 is physically impossible, then the fuel‑switch‑led causal chain collapses.
Engineers argue that, instead, the AAIB’s described “recovery” behaviour is more likely to have occurred much earlier, in a far narrower window — the first few seconds after a fuel cutoff around 08:08:44 UTC, when the cores were still spinning fast and the engine‑driven permanent magnet alternator (PMA) was providing full voltage. In that brief interval – roughly 08:08:45 to 08:08:50 UTC – FADEC could plausibly attempt an automatic relight in response to a power disruption, before N2 sank below idle.
A systems failure, not a pilot error
There is an important implication here. FADEC can only attempt an automatic relight if the cockpit fuel switches remain in RUN. So that means the fuel switches have to be in RUN for the first 11 seconds of the flight, say engineers. Which would mean neither Captain Sumeet Sabharwal nor First Officer Clive Kunder had their hands anywhere near the switches until the engines flamed out.
Engineers say the new data plus GE documents point to a relight sequence like this:
Power disruption → FADEC cuts fuel under its logic → FADEC attempts relight while cores are still spinning → engines eventually decay through minimum idle at 08:08:47 UTC and die → pilots attempt a fuel‑switch recycle within 5-6 seconds after loss of thrust to save the aircraft.
“Very early on, I kept saying that engine computer FADEC can command a cutoff under a protection logic called thrust control malfunction accommodation (TCMA). I kept stressing on it and yet people were so ready to believe pilot error, with no proof whatsoever,” said Captain Randhawa.
For the FIP, and for engineers who have tracked VT‑ANB’s long history of electrical issues, this is the heart of the new story. The RAT’s timing and the simulator’s N2 (engine core) behaviour and fuel‑flow traces do not just chip away at the AAIB’s preliminary narrative; they point decisively towards systems failure as the prime mover and away pilot error.
Simon Hradecky, electronics and aviation engineer and editor of Aviation Herald, who has given technical analysis and evidence to the AAIB as part of the Air India crash case in the Supreme Court, said that the AAIB’s – and Western media’s – implied story does not add up technically or psychologically.
He has asked the AAIB, “Did the investigation ever consider the scenario, if assuming suicide, that the pilot determined to kill himself and everybody on board (and thus being a murderer destroying his reputation forever) would put the switches again to CUTOFF after the other pilot moved them back to RUN?”
He then asked, “If one of the pilots attempted suicide on takeoff, would it not have been easier to just push the aircraft’s nose down to immediately impact ground, particularly if the proclaimed method of suicide by engine shut down seemed to be thwarted with the fuel control levers back to RUN and the engines attempting to relight?”
These were among many other questions raised. The Supreme Court on November 13, 2025, directed the AAIB to respond by December 14, 2025. However, the AAIB has not responded to the questions asked by petitioners – which include Captain Sumeet Sabharwal’s father Pushkaraj, the FIP and Captain Amit Singh from Safety Matters Foundation – till date.
The Wire has reached out to the AAIB and the Directorate General of Civil Aviation to ask for its response on the FIP’s simulator test results and whether this has changed their outlook on the timeline of events before the tragic AI 171 crash. This article will be updated when a response is received.
Also Read: India’s International Day of Yoga Machine Is Losing Altitude https://www.vibesofindia.com/india-international-day-of-yoga-machine-is-losing-altitude/
