You are an airline pilot or something related if I remember correctly. That is what I was hoping for. I work in IT software development and I am an aviation buff.
One element of control the pilot would have is whether to turn it on at all. It is built to be a deterrent system and it would be unlikely that a hijacker would pick a time when heavy icing or similar conditions exist. If that happened, the pilot still has the ability to make a decision about whether to turn it on at all. In the end, it would be a judgment call up to the pilot and that is what we want right?
That said, there is a whole lot you can do with avionics software (or any software). You could set conditions when the system would disengage on its own (with warning. still without human intervention). This could be because any serious type of mechanical failure. If the terrorists blew up the ILS transmitters at the go-to airport then the GPS could locate another. If there was complete avionics failure due to human destruction then the system could disengage also. That would also work as deterrent because who know how to surgically disassemble the avionics of a 737 with a fire axe?
To those of you who say it is fighting a non-problem: Why do little old ladies get searched for weapons all day every day? My wife recently cleared a checkpoint and my three year old ran through and grabbed her. The whole family had to go through a hand search. It is now needless hysteria out there. This should be the time of effective and non-intrusive solutions.
Commercial airliners have a service life of over 20 years. A lot can happen in twenty years. In addition, people react emotionally to airliner crashes and hijackings like few other things. The dollar toll isn’t just with the lost plane, people and possessions. It is about public perception and the economic toll on the airline industry.
I am about the last person to propose feel-good but ineffective measures. However, I think it would be great if you told people “these planes are almost hijack-proof” and here is why.
Good, safe human-machine interfaces are not that hard to design. I propose an “arm” switch, a keypad with a four digit code, and then an “enter” key physically located a foot or so away in a non-obvious place. A trained person could complete the task in 5 seconds and an untrained person never.