We currently have ground controlled aircraft flying in the middle east, with pilots sitting at consoles in the US. You hear them mentioned on the news as "drones" or "pilotless drones". These drones autopilot to the area of operations then remote pilots take control for the required task (recon, bombing, whatever), then autopilot back to base, while the pilot goes home to have dinner with his/her family.
When airlines go pilotless, they will follow the pattern of military drones, with the slight modification that it will autopilot all the way unless human intervention is called for, then control will be given to a remote pilot sitting in a control center, and there will be some established ratio planes in the air to pilots on the ground, who when they are not actively flying a plane can be practicing emergency situations. Really, a pilot on the ground in such a situation would be far more practiced at emergencies than current pilots in the air.
Of course, CARGO planes will go pilotless long before passenger airlines do. Most likely, nearly all cargo planes will be pilotless and have a long demonstrated safety record before passenger airlines try.
Even when it's technologically safe and sound to go pilotless, I'm not sure it will ever happen, even for cargo:
1) The cost of 1 or 2 pilots in the scheme of things won't lower ticket prices, particularly compared to
2) Our human perception of safety. I doubt people will be thrilled about pilotless 7x7 size jets flying overhead, let alone getting on one. No matter how safe, it's never 100%. Even now, there are still couples that don't fly together just in case.
3) When a pilotless plane crashes, it will be difficult to convince people that an on scene pilot wouldn't have made just that little bit of difference.
