Couple more incidents of pilots forgetting how to fly a plane...

[Evan]

I tend to agree that if the software on that tablet is relied upon to ensure a safe flight, then yes, it is deemed critical. Unless there is a reliable manual way to validate for potential mistakes, I would call it critical. If I can check in passengers with a piece of paper instead of a computer, it has no bearing on the process of flying the plane safely. Time pressure comes from all directions and would exist whether the computers were working or not. The systems that create smooth business operations are a business problem and they will allocate their investment as they deem appropriate to manage the risks of failure. Regulating business operations software would be a sure way to ensure all innovation dies and that no process improvements will ever be made.

What I mean is a set of standard requirements for RELIABLE redundancy in hardware provisions and software engineering standards including third party dependencies, so that if a server or hub or switch or individual power source goes down anywhere in the enterprise, it doesn't bring the entire system down with it; it doesn't cause massive groundings that last many hours and result in elevated scheduling pressures that might infect a pilot's better judgement, so it doesn't result in accidents.

There are many things that can introduce confusion in the flight deck and it is the the responsibility of the pilots to remember to fly the plane first.

Yes. However sometimes it doesn't really work that way. Confusion is a very good way to fumble that responsibility. Therefore it is the responsibility of industry oversight, within the realm of practicality, to thwart anything that might lead to such confusion. Also, vision and the ability see beyond one link of the casual chain never hurts.

[brianw999]

Wow. And that's a lot of debris to leave on the runway!

Naaah, that’s OK, that will polish out with some T Cut and a lick of touch up paint.

[Schwartz]

What I mean is a set of standard requirements for RELIABLE redundancy in hardware provisions and software engineering standards including third party dependencies, so that if a server or hub or switch or individual power source goes down anywhere in the enterprise, it doesn't bring the entire system down with it; it doesn't cause massive groundings that last many hours and result in elevated scheduling pressures that might infect a pilot's better judgement, so it doesn't result in accidents.



Yes. However sometimes it doesn't really work that way. Confusion is a very good way to fumble that responsibility. Therefore it is the responsibility of industry oversight, within the realm of practicality, to thwart anything that might lead to such confusion. Also, vision and the ability see beyond one link of the casual chain never hurts.

Speaking from experience, I guarantee they had all that in place for those other failures. Guaranteeing redundancy is just not possible, and in my experience every redundant system I've ever worked with in 30 years in computers still has a way of failing. You're better off just being able to restart the thing or move it quickly.