This is a troubling article, especially the description of the A330 controls. That the sticks of the two pilots don't register the same control inputs is unfathomable. One reason for having two pilots, rather than just one, flying an airliner is so that they are aware of what each is doing and they can correct each other if they see something amiss. Automated sticks that don't talk to each other for aileron and elevator control inputs...well, what were the designers of this cockpit thinking when they came up with that one? And then there's the automated trimming. There's a long history of the development of realistic force-feedback systems to simulate the resistant forces a pilot would feel if he/she were flying a plane with manual controls. It should feel hard to tug back on the stick with the vigor that the pilot-in-command apparently used. The automated pitch trim seems to have taken a lot of this feel away...and defeated decades of realistic force-feedback development. That was precisely the problem in the early days of fly-by-wire systems: there was no force feedback, and it was too easy to put in a command without feeling the customary resistance one would feel performing such an extreme manoeuver. An extreme rudder input into an A300 caused the vertical stabilizer to rip off a plane bound out of JFK Airport in 2001. The pilot put in excessive rudder inputs, partially because the A300 had poor force feedback in the rudder actuation system. It was too easy to push the rudder pedals to their stops, unlike it would be flying a plane with mechanical controls. So it seems that Airbus doesn't yet have it right.
I also like the idea suggested in this article of a simple backup autopilot that might buy the pilots some time to sort things out in a complex and confusing situation. Seems like that would be easy to implement with a little programming.
