(Opinion) The final accident report on the crash of Air France flight 447 in 2009 was released this week, and no new ‘shockers’ were presented as part of the official conclusion. Yes, the pitot tubes were faulted which was highly probable (as those in particular were the error prone Thales AA type). And yes, the pilots were faulted, as we assumed they would be. What wasn’t faulted was a probable third factor that might further implicate the manufacturers flawed philosophy on automated controls.
Low Face Value
In my previous article, I first discussed Clive Irving’s opinion on TheDailyBeast.com, and how I believe his view might be an oversimplification of the crash. His current post on the final report is more critical, but he still takes the report mostly at face value (his latest here), and his primary concern (as with other commentators) is the continued use by Airbus of the known fallible pitot tubes.
While of course it is highly plausible that the airspeed measurement error caused a problem (and there’s hard data to prove it), which then forced the substitute pilots who were at the controls but not trained adequately, to mis-correct for this problem. However nearly every aviation commentator on the Internet who has reviewed the current data seems skeptical of the report in some way, and justifiably so because not everything adds up.
Pitot’s Bad, Pilot’s Bad, Yes. But…
Yes, the pitot tubes played a role. I don’t dispute that (and again, there’s data). However in my (unqualified) assessment, to cause the complete loss of plane and life, there were only two other conditions in the realm of possibility which could have turned this fixable error into a tragedy.
Either: 1) modern airliners are no longer aerodynamic because they can (apparently) stall at 38k feet and fall to the ocean tail first without being able to recover, or 2) the apparently child-like ineptitude of the pilots who were purportedly at the controls, were not simply untrained, they were vastly unqualified and not competent to fly this plane.
Neither scenario seems like it could be true, but what led to the crash seems impossible otherwise. So the simple question remains: how in the heck could the pilots, even as bad as these supposedly were, have maintained the extreme nose-up attitude after the stall (presumably doing everything wrong) yet have plenty of altitude and time to recover but not be able to?
A better (more critical) summary of the final report was written by Wil Hylton on the Huffington Post. (His article here.) He gives a much more prickly perspective, and at the end summarizes his view by saying:
One hopes that the BEA’s emphasis on pilot error does not distract from other important questions about the faulty parts on Flight 447. -Wil Hylton, The Sniff of Politics on Flight 447
Hylton is also referring to the Thales AA pitot, but I still believe that some other force was at least partially responsible for the crash of Air France #447. And I believe the most likely culprit was the autopilot.
I’m Afraid I Can’t Allow You to Do That
Dave Marc (Dubois, Captain)
It seems only logical, that the illogical behavior of the airplane had to have some other cause. The only reasonable explanation I can come up with is that the French-built autopilot which is known for it’s built-in superiority complex, may have played a significant factor in the crash. As revealed wonderfully by Pat Flannigan here, the design philosophy surrounding the Airbus autopilot is one of arrogance and over-confidence. Here’s an excerpt from Flannigan’s report:
…(we wanted to) build a plane that would actively protect the passengers from pilots. (From Pat Flannigan’s article: Does Airbus Fly-by-Wire Technology “Protect Passengers From Pilots?”
The article above is worth a read, and if true, then why this probable cause wasn’t more fully investigated is tough to know. There were in fact many other related incidents (noted here) that were the result of pitot tube or other sensor problems, which led to a flight control interruption and then a drop in altitude.
If the investigation was truly unbiased, one may guess that perhaps autopilot code is terribly difficult to troubleshoot, but another likelihood is that the autopilot code design might be used to more easily implicate the manufacturer.
Regardless, to echo Wil Hylton above, hopefully politics in the competitive and cut-throat airliner market had nothing to do with it. We may never know.