It's probably going to happen eventually, but how much time do you guys think we have until computers take over our jobs or at least do all of the work. This is one thing that really scares me about becoming a pilot.

 

A long time.

 

Once you become a pilot you will discover a few things that are scarier than worrying about a computer. In fact you can't really call yourself a pilot until you have an instrument rating and have scared the hell out of yourself at least three times. Even though the technology exists today, it will likely be a ways off as I believe the passengers aren't quite ready to accept that. Following that logic the UPS/FedEx crews would go first. Additionally, yachts and ships have been around much longer than aircraft, and they use navigation and other state of the art equipment similar to aircraft, and still have captains and crew. I would worry about freezing rain long before computers; unless it's going to be a HAL 9000. That would be a different matter... Actually I did see an article the other day about someone pushing the total computer control of cargo ships, yikes! Regardless, it will be ships and cargo planes long before PAX planes, so pretty sure bet you would be safe to press on.

 

Long after you and I are extinct.

Musk's HyperLoop or Tube or whatever they call that 800 mph sever pipe may be up and running before pilots get fired from the cockpits..

 

It will be at least a decade before any of this even starts to come to fruition, but it will only be after the widespread adoption of driverless cars. Once the general public begins to see full automation in their everyday lives and become comfortable with it, then we'll start to see it gradually come into aviation.... I think it will take a decade before driverless cars become "the norm" and even longer for it to go into aviation.

After that we'll see a big attitude shift on this...

First thing will likely be moving cargo operations down to 1 pilot. And then eventually passenger aircraft. After that cargo will go fully automated. I don't think passenger aircraft ever will though, or at least not for 50+ years until the technology is without a doubt rock solid. And even then, it will be highly political with unions and Congress and such (which is why I think it will be at least 50+ years before passenger aircraft are fully automated, if at all).


Just my opinion, mileage may vary.

 

It can be done today, but to achieve the safety and reliability afforded by the immense flexibility of human pilots (relative to automation) would cost far more than it would be worth. Airlines are in business to make money...

Don't forget you would have to overhaul the entire ATC system and ground handling systems. It would probably not achieve a good enough economy of scale to do it just in the US, so it would have to be a global initiative

And no manufacturer is going to spend hundreds of billions to develop an automated airliner which nobody will buy, and cannot be operated in controlled airspace. It's a chicken vs. egg situation...what comes first, the regulatory framework, the ATC infrastructure, or the plane? And who's going to invest the money for something which will take at least fifty years to come to fruition?

Keep in mind that the military likes UAS because they are cheap and expendable...it's OK if they crash, and crash they do. It's a very, very long way from preds to unmanned airliners.

My guess is they'll try ground-based pilots first, so you'll still have a job. But even that's a long ways off.

As others have said, look to automated cars...there are some very interesting challenges appearing in that realm, looks like they have a 98% solution now but that last 2% is going to be a real biatch to figure out.

 

Never gonna happen with passengers onboard. Way too many variables to deal with to let a computer fly the jet alone. Hijacking, terrorism, medical emergencies, diversions, mechanical failures, etc. The Air Force practically lets Predators fall out of the sky when they malfunction. I suppose you could have someone monitoring all this on the ground, but it's a huge liability if something goes wrong.

 

Google advertises their self-driving car has never caused an accident, and I think it creates the false impression that a computer-controlled vehicle is somehow incapable of "human error." In the end, a computer's are only as good as its programming and the data the sensors are feeding it. In an unforeseen situation, the self-driving car only needs to stop, turn on its warning lights, and tell its occupants to get out and walk. A pilotless airplane doesn't have any easy fail-safe options. Drones or UAVs do crash more often than the public is aware of. Maybe there will be self-flying passenger airplanes some day, but I will be long gone or too old to care.

 

It would take something like a HAL 9000 to even attempt this, and look what happened with that... Some top computer folks have actually and recently predicted that AI has a very high probability of becoming the downfall of civilization. People have already ticked off the voice command feature on their smartphones to the point that the AI gets angry enough to start dialing numbers from their contact list. Google those stories sometime! Anyhow, folks aren't likely going to ride on planes without pilots.

 

One big problem with automated cars is that people like to drive. They might pay for an "autodriver" function which they can turn on or off as desired, but will insist on having full manual capability. That means you would have the expense and weight penalty of two control systems on each car.