Maybe 2120.

A single-pilot airliner will need to be FULLY autonomous because pilots do keel over dead sometimes.

You will need new technology and extra redundancy, which will cost a lot of money to design, build, and maintain. Every nuisance computer message at an outstation will be a no-go item...for real. What is currently a minor nuisance fault will be a divert-to-nearest-suitable.

You need to massively re-engineer ground, approach, and enroute systems. Who is going to get the gubmint off their butts to do THAT??? Who's going to pay for it? The FAA is having a hard time getting nextgen up and running by 2020 and that is nowhere near capable of handling autonomous airliners.

The FAA's standard for 121 safety is that any changes result in equivalent or improved safety. It would be VERY hard to get them to buy off on something like this. And that's to say nothing of public perception and congress.

The DoD has reaped some capability and financial benefits with UAV's because of improved endurance and hardware cost savings because they don't have to meet manned safety requirements. The USAF is in hot water because they lost nearly half of their predators to non-combat accidents in the last ten years. They have also discovered (the hard way) that certain phases of flight cannot safely be conducted via SATCOM due to control response latency...this means line-of-sight between controller and UAV is required for those phases.

Embrear may be considering it, but what airline is going to pay extra money for a capability that regulators will not allow them to use?