The prelim report on the Air India 171 crash, deliberately or not, has left the finger pointing at the pilots. In this article I want to explore if blaming the pilots is logical given the information and statistics available.

Recap

A quick recap of the prelim report of Air India 171.

-Conversation in the cabin with one pilot asking the other if they cutoff.

-Data recorder says fuel control switches were put to cutoff 4 seconds after takeoff.

With this information the world seems to be leaning towards blaming the pilots but is this logical? Is this the most likely situation?

What Are We Saying?

When we blame the pilots what exactly are we saying?

I think it’s fair to say we are saying the following

Statistically speaking, pilots on airliners are more likely to crash the plane than a plane has a technical issue that causes plane to crash.

But is this correct?

How Many?

How many incidents are there of a commercial flight being brought down by a suicidal pilot?

And how many incidents are there of a plane encountering a software issue that brings down a plane?

ChatGPT

Asking ChatGPT we get the following for pilots deliberately crashing plane.

Year	Flight	Location	Fatalities	Status
2013	LAM Mozambique TM‑470	Namibia/Angola	~33	Confirmed
2015	Germanwings 9525	French Alps	150	Confirmed
2022	China Eastern 5735	China	~132 (est.)	Strongly Suspected

ChatGPT

And for planes with software issues directly causing crash, ChatGPT says this

Flight	Year	Fatal?	Cause

Lion Air Flight 610 (737 MAX)

2018

Yes

MCAS software triggered by erroneous AoA

Ethiopian Flight 302 (737 MAX)

2019

Yes

Same system/software failure

Qantas Flight 72 (A330)

2008

No

Software design + hardware anomaly

ChatGPT

And if we ask ChatGPT how many software issues have led indirectly to fatal or serious incidents we get this

Incident	Year	Fatal Crash?	Software Role	Human or Hardware Primary Cause
Air France 447	2009	Yes	Misleading display behavior	Pilot confusion after automation change
Asiana Airlines 214	2013	Yes (fatal injuries)	Complexity of automation modes	Pilot mode/manual mismanagement
Sriwijaya Air 182	2021	Yes	Autothrottle malfunction	Hardware/software interface issue
TUI 737 near-runway disaster	2024	No	Autothrottle glitch	Software failure

ChatGPT

Analysis

So in the last twenty years we have two confirmed crashes caused by suicide. And one suspicious.

And with regards to software malfunction we have two confirmed crashes caused by software & three crashes caused by software behaving in unexpected way (software not directly the cause) and two near crashes.

Conclusion – Air India 171, Pilots vs Software

Looking at this data it would seem twice as likely that Air India 171 encountered a software issue versus a pilot suicide.

People are saying that systems have backups and the chances of anything going wrong are almost inconceivable.

And this is correct.

But the data also tells us that the chances of a plane being crashed by a suicidal pilot is even less.

So I ask again, is it more logical to state “it must have been a pilot” rather than saying “it must have been a software glitch”?

If we are going to jump to a conclusion, which conclusion is the most logical to jump to?

Explore StrikeEngine

StrikeEngine TV Highlights