Originally Posted by
MathFox
I would say that you were luckily not distracted by ATC and landing checklists. And lucky having sufficient altitude to correct the automation induced problem.
I fully agree - that was more or less the gist of my slightly sarcastic comment there. And even without those distractions my response was reactive, rather than pro-active, as I was clearly surprised by the system behaviour.
It just doesn't make sense that an autothrottle stubbornly continues to use erroneous sensor inputs when the autopilot still works fine on a correct set of data. And it sure is a sign of sloppy engineering that a flight control system does not degrade gracefully and unambiguously (e.g. by disengaging the autothrottle, coupled with the usual aural and visual cues) if an unacceptable discrepancy between two measurements is sensed (in our case IAS, in the Turkish case Radio Altitude and in the MAX case AoA). Well, actually that's the problem here: these systems didn't sense the discrepancy to begin with, because the system designers deliberatly chose to rely on one sensor only, even though multiple sensors were available.