Say within 10 or 15 years?
I am not a pilot and don’t know what kinds of things that they deal with that might be too difficult to program a computer to do. They can certainly take off, fly and land autonomously right now.
I also don’t know what the regulatory adoption process will be like. (I do think that it ought to include tiger team testing).
How will it affect the economics of commercial air travel as well as air freight?
Thanks,
Rob
I’m not a pilot either, nor am I involved in any way with the airline industry.
That said, I don’t think we’ll have pilotless airliners any time soon. Not in our lifetimes. Not for a few generations.
I have two reasons for saying this. First, the air-traveling public will not have confidence in pilotless aircraft, and simply will not fly in them. Even if airliners were able to take off, navigate to their destinations, and land safely (and I don’t doubt that that will be possible fairly soon), the public will want a qualified person in the cockpit just in case. And that’s a rational concern – after all, the programmers can’t think of every contingency, right? Under perfect circumstances, a pilotless aircraft might do just fine, but once in a while, someone might actually have to fly the plane.
Second, I can’t imagine (although I’m open to correction) that the cost to the airlines of putting pilots in the cockpit is all that significant a fraction of the cost of a flight. So why would they risk alienating the ticket-buying public by not doing it?
As I understand it planes already largely fly themselves. Humans are there for the weird shit.
For instance, the Google self-driving car has put on a few hundred thousand miles with no problem. However, it is apparently completely thwarted in heavy snow and parking lots.
I imagine planes would be similar. The computer can do just fine in a variety of situations but the outliers need a human to cope with. It will be a long while yet before computers can manage the unexpected.
Total hijack but this really qualifies for a sequential posters thread. 
Not with me in them. Although I will note they are far closer to being practical, with more immediate benefits, than “self-driving cars.”
This pretty much nails it, really. No matter the logic behind it, if ABC Air is the first to use pilotless planes, ABC Air will be the first airline to see sales crater.
I don’t see a lot of benefit in the short term anyway. As has been pointed out, airplanes largely fly with minimal input most of the time anyway; the pilots are there to handle the weird things and tricky stuff. A pilot is not an especially large part of the cost of flying the plane, in the grand scheme of things; airlines try to screw them because they try to screw everyone, but the guy flying the plane is probably not one percent of the gross revenue or two percent of the cost of the flight (those are offhand guesses.) Even if the pilots job is solely to sit there and jump in only if Otto loses his shit, people will want the pilot there, and for quite some time the airlines will want the pilot there.
Driverless trucks are by comparison a much bigger savings.
Airlines will be flown by computers, two pilots, and a Rottweiler. The pilots will be there to watch the computer, and the Rottweiler will be there to bite the pilots if they try to touch anything.
Emphasis added.
Two words: Drones.
My WAG is not until the cars go driverless, and then maybe not ever. I reason if people can let the AI drive little Timmy to soccer practice, then a robot plane is not so scary. On the other hand, planes tend to be huge and carry loads of fuel and fly really fast and by necessity near large population centers. The first AI incident will scare a lot people, and may shelve many plans to go pilot-less. Its possible by that stage though there will still be a human in the ‘cockpit’ but they won’t be called ‘pilots’.
I think the better question is when will the militaries of the world stop employing fighter pilots, but that might need a different thread. I see a lot more advantages for a warbird rather than airliner to go completely pilot-less.
Passengers wouldn’t want to fly on an airplane they knew had no human pilot on board. They need that sense of psychological reassurance and comfort.
Didn’t they originally design the 747 with no cockpit windows, but changed it quickly?
I think it could happen much faster than people imagine. Autopilot planes could be trialed on cargo flights first. It’s a much easier sell since there’s no risk to human life. Once they demonstrate a proven track record, some budget airline like Easyjet or Spirit airline could offer a $20 human pilot charge and you watch how quickly people will realize they don’t really need the reassurance of a person at the front of the plane to save $20.
Well except for the people living in the houses that the plane crashes into.
Computers do great when the rules are clearly laid out and the physics are not complicated. For example, freeways and roads. You have finite physical limits, traffic heading overwhemingly in controlled directions, etc. Computers don’t do so hot when things get fuzzy. Such as parking lots, where there ARE no rules 
There’s a lot more fuzziness in the air.
Are they? It takes serious money to provide the training so they can do this and there is a debate going on that in many places the airlines don’t want to spend the money. There have been two major crashes in Asia recently where there is a good possibility pilot error may have been the cause: in Taiwan where an airplane lost an engine–and the pilots shut down the good engine and in Indonesia where the pilots may not have been able to handle a stall after going into a bad weather situation.
No driving on a road is more fuzzy–roads are not controlled access–so there are conflicts with other vehicles, pedestrians–and there is very little reaction time. Except for weather issues and mechanical problems I think flying could fairly easily be programmed.
The problem with the airplane is that there is no equivalent to pulling off to the side of the road and stopping if there is a problem.
Pilot error has caused many crashes, that’s hardly a new phenomenon.
I’m not convinced people will be too scared to fly a pilotless plane.
I agree that for large passenger jets, having a pilot is a small price for a perception of increased safety. If going pilotless made tickets 10% cheaper, it would have a chance because passengers tend to be very price sensitive but if it makes a difference of a few cents at most, it won’t be perceived to be worth the saving.
What about small planes and helicopters, though? Say, planes that only transport half a dozen to a dozen people or helicopters that transport 2-4 people. There, having a pilot (actually piloting and paying him to be available) represents a greater proportion of the cost of flying than it does in jumbo jets.
If so, it might mean that having your own private aircraft for travel will be the preserve of people with 7 digits in their bank accounts instead of 9. It might mean that lots more organizations like SMEs will be able to afford timesharing an aircraft. Business-class flyers might find it worthwhile to be able to go from point to point and skip the 2 hour airport wait.
Far-flung places which are chiefly reachable by aircraft might be more readily accessible too.
ETA: Sorry, I see that OP specified “airliner” which excludes the eventually I outlined.
What’s the second word? 