There are many, many obstacles to unmanned passenger planes and you have to be a systems engineer(or similar), and an experienced commercial pilot, and have a firm grasp of regulatory issues to fully understand this issue.
Can it be done? yes, absolutely. It could be done today.
But can it be done with an equivalent level of safety, reliability, convenience and at an economical cost? Not a chance in hell.
The USAF is in trouble because it has lost over 1/3 of it's predators to non-combat accidents...this is a fleet which barely existed ten years ago. Predators are useful because they provide CHEAP endurance...this means many shortcuts in redundancy and reliability. It gets the job done, but you sure as hell wouldn't want to ride on one...especially if it's controlled via SATCOM.
The DoD has no plans or interest in using UAV's to carry personnel, even combat troops.
Starting at the lowest level and working up...
- Artificial intelligence systems are far inferior to the human mind, and there are no indications that this will change soon.
- Since a computer cannot adapt to unforeseen circumstances like a human, unmanned aircraft would need to be far more redundant, reliable and automated than they are today....$$$$$$$$$$$$. You would also need a see-and-avoid system, unless you are willing to ban all manned aircraft. Anyone who has actually used TCAS knows that it is very unreliable and could never be relied on as the primary means of collision avoidance.
- Mx would have to be done in a nasa-style clean-room by guys in white suits $$$$$$$
- Ground handling and taxi must be automated. This affects not only the airplane, but the airport, ground personnel and systems, and ATC.
- Airports would need extra-redundant approach systems since a visual approach would not be an option.
- The national and global airspace and ATC systems would have to be totally re-engineered. Now we are getting into the political realm...$$$$$$$$$$$
Even if the cost of the airplane can be justified by firing the pilots, who is going to pay for the required infrastructure and regulatory changes?
The government? Not likely...why spends tens of billions just to put a few tens of thousands of pilots out of work.
The airlines? They can't see as far as the tips of their noses, and execs of near-bankrupt companies are not going to spend huge sums on a project which will not achieve payoffs until after they are long retired (or more likely dead).
Boeing and Airbus? Not going to invest huge sums in a product which the airlines can't afford and which cannot legally fly in the national/global airspace system.
Don't forget public perception...many, many folks simply would not fly on such an aircraft.
It's a chicken/egg scenario...you need to spends huge sums on technology, huge sums on infrastructure, and alter public perception. Any one of these tasks would very difficult to justify before the others are completed.
My guess? Not this century. I will start to worry when trucks, cars, and ships are automated. So far we have automated elevators and some small trains...long ways to go before we get to airplanes. People are still used as vehicle operators because we are cheaper than ultra-reliable, near-foolproof systems. Human pilots can make up for a lot of engineering shortcomings.