"I suspect that, by removing certain kinds of feedback, the experience of flying an Airbus is rewiring pilots' brains in dangerous ways."

I don't know whether that's true or not, but that seems something it might be possible for an aircraft design to do, and if so, something to avoid. Does anyone know to what if any extent aircraft designers try to avoid letting pilots become too detached from the experience of flying?

I worked on 757 flight controls design back in the early 1980's. I'm not a pilot and I'm not familiar with the Airbus designs.

The 757 flight controls were hydraulically operated, and were controlled by cables operated by the pilot that opened valves on the actuators (hydraulic rams) on the control surfaces. The pilot's controls did not directly feel the flow of air over the control surfaces feeding back into the stick, like a light plane would with the cables moving the control surfaces directly.

To correct this, Boeing designed a "feel computer" that would push back on the controls as if actual wind were pushing on the controls. The idea was that as close as possible the pilot's controls should "feel" like he was directly connected to the control surfaces.

The feel forces generated could be well over a hundred pounds, if I recall correctly, though Boeing had to reduce them in order to accommodate female pilots. Note that for a high performance aerobatic plane with direct controls, it is a physically exhausting task to control it.

My father flew B-17's, which were probably the biggest airplanes without hydraulically assisted controls. You had to be a strong man to fly it. It would fly with only one outboard engine running, but the force required on the rudder pedals meant the pilot and copilot had to constantly relieve each other.

There were actually bigger planes, like the Martin JRM Mars (a large WWII flying boat) with a gross takeoff weight in excess of 165,000 lbs. They use a mechanism called a "servo tab," which is sort of like a control surface for the control surface. In essence, it deflects in the opposite direction of the control surface, and usually in proportion to the control force (a system of levers and springs measures the amount of force being transmitted to the controls by the surface).

The primary disadvantage, aside from the complexity of the system, is that it tends to delay feedback, which results in PIO.

I don't mean that kind of feedback. I mean the feedback you get that teaches you to the limit of the vehicle's operational parameters.

For instance, when I started driving, I sometimes used the brakes too aggressively, and experienced a number of "micro-skids". These experiences quickly taught me how hard I could get away with pushing my vehicle. But it's not something I'm consciously aware of: my foot just automatically reaches for the brake pedal at the right moment. There's a subconscious "anti-skid system" built into my brain.

And that's the way it has to be. I couldn't possibly maintain conscious awareness of all aspects of vehicle operation. I'm not that smart!

The Airbus control scheme takes that feedback away. Pilots don't directly control the plane at all. They merely tell the plane what they want it to do, and the plane's control system decides whether it's a good idea or not. If a pilot does something that would stall the plane, the plane just doesn't respond.

I suspect that this causes the Asiana 214 pilot's subconscious "anti-stall system" to atrophy, to the point where he was no longer competent to manage a Boeing plane. The bit of his brain that automatically monitored flight speed and the red pips was no longer functioning. More seriously, his higher faculties were not aware that his subconscious systems could no longer be trusted (beyond a vague sense of unease).

Well, that's my theory, but note that I haven't explained why the instructor didn't spot what was going on.

Well, AFAIK a modern Boeing plane like the 777 / 787 will also prevent stalls just like the Airbus, the difference being that the pilot can override the airplane. So a 777 pilot would also assume that the airplane would normally prevent a stall.

In this accident, I think the main problem was not anything Boeing vs Airbus, but more a lack of experience with hands-on flying visual approaches, and a lack of monitoring, which would quickly have uncovered the fact that the auto-throttle wasn't operating. If the linked PPRune posting has any truth in it, I think we a have a substantial part of the explanation there.

I would say the Airbus comments are incorrect.

Prof. Missy Cummings (former F/A-18 pilot) at MIT is involved in research into cockpit automation. She was featured on at least one documentary into AF447.

What is interesting in the Airbus vs Boeing debate is that pilots like Capt. Sullenberger who has flown both Boeing 737s and Airbus A320s doesn't claim its an "Airbus" problem. The problem is that airline pilots are becoming overly reliant on autopilots and forgetting the basic skills of flying the aircraft. http://www.cbsnews.com/news/capt-sully-sullenberger-airplane...

You may be onto something there, but this explanation doesn't cover the instructor, who was experienced on Boeing, not Airbus.

The pilot in question has out and out stated that he was uncomfortable performing visual approaches. Now, while it does indeed boggle the mind as to how such a person could get through training and attain command of a 777, the fact is that he did so. I don't see how you can argue against negligence when the pilot has basically come out and admitted it (although he doesn't seem to think he has).

This reminds me of something (much less substantive and important) that I have been realizing after buying my first Apple computer recently (I held out for a long time). Things like the trackpad and cable wrapping system are so amazingly intuitive, once you figure them out. However, I am so accustomed to sub optimal design, that I had to be shown how to use them both. This may be a case where sub optimal design is a net win, but ensuring that the pilots must stay engaged to at least a minimal degree.