Concentration on the electronics must not divert us away from other possibilities that can cause very bad results.
That's just it isn't it?
Computerization is all well and good and electronics have come a long way beyond what was state of the art just 2 decades ago, but both of these advances in technology have native limitations in understanding via sensors and algorithms when things mechanical, electrical, and/or aeronautical go wrong.
It's not at all down to a science just yet no matter what the sales team may come up with. The sooner we honestly come to grips with these limits of the state of these arts, the better we all shall be - particularly with said systems and their component parts now reaching 'an age'.
A mitigation technique? Keep the PF in the loop at all times.
This last suggestion applies all across the board from shore to shore, manufacturer to manufacturer, and lastly engineering to final "out of beta" product.