It appears that the 737 MAX is not the only case of Boeing’s over-reliance on only one of two available sensors and not providing clear warnings and complete (or any) documentation of operational details. Today’s Seattle Times (reprint from NYT) article titled 2009 accident foreshadowed handling of MAX crashes goes into the details of the crash of a Turkish Airlines 737 NG airplane outside of Amsterdam. (The online version of the article is more harshly titled — How Boeing’s responsibility in a deadly crash ‘got buried’)
In this case the cause was — reliance on only one of two available altitude sensors aggravated by an apparently distracted cockpit crew. The sensor went bad, and that caused the autothrottle to assume the plane was about to land when it was still on approach. The autothrottle then slowed the aircraft below the speed the crew had set — which they apparently did not realize until the “stick shaker” warnings went off. The captain took manual control back (more on this a bit further on), but by then it was too late to recover — 9 people died.
Sidney Dekker, an aviation-safety expert who was commissioned by the Dutch Safety Board to analyze the crash called it:
A sentinel event that was never taken seriously.
But…
...at the behest of the American team, the Dutch board deleted or amended findings in the accident report about issues with the plane.
The “American team” included Boeing representatives as well as “federal safety officials” (NTSB, FAA? article doesn’t mention). Basically the blame was put fully on the crew because, like the recent MAX crashes:
Boeing assumed before 2009 that if the altitude sensor malfunctioned, the crew would quickly recognize the problem and prevent the plane from stalling.
...Or not. And Boeing carried those assumptions through to the MAX.
Another familiar sounding problem was the autothrottle continuing to reassert control after manual override and immediately dropping speed again. The captain did eventually recognize the problem and shut off autothrottle, but too late.
No aural alert, BTW. Boeing added that after the crash. Boeing had also corrected the autothrottle single sensor problem for many NG models in 2006 (but only as an optional update — sound familiar again?), but the Turkish 737 NG was using a different model autothrottle that was incompatible with the fix. Boeing created a compatible fix — after the 2009 crash.
Interestingly enough, in 2014 Boeing cited the altitude sensor fix as part of their justification to the FAA for relaxing alerting requirements for the MAX, despite the fact that they had reproduced a VERY similar problem with a different single (of two) sensor reliance in the MAX that was NOT corrected until a lot more people died.
Of course, no report on Boeing/FAA/NTSB being dickish would be complete without a spokesman (Joe Sedor — NTSB — who led the American team investigating the 2009 crash, and is now “overseeing the NTSB’s work on the MAX crashes”) acknowledging that a single sensor failure contributed to both (MAX?, or MAX and NG?) crashes “but cautioned against focusing on it”.
Because that would hurt their fee-fees or something.
I’ll risk one final quote from David Woods, Dekker’s doctoral thesis advisor in reference to both the 2009 crash and the two MAX crashes:
That this situation has continued on for so long without major action is not how engineering is supposed to work.
Such an innocent…
Add — reading more carefully, Boeing and NTSB reps were definitely on the “American team” — not clear if FAA had any representation.