What about the long-predicted, never achieved, flying car?
Moller ( http://www.moller.com/ ) has been developing his Skycar concept for the last 25 or 30 years and has so far produced just flying tethered test beds.
Now NASA is getting into the act with a one-man, all electric, VTOL entry. Only theoretical and simulations at this point, but maybe it will stimulate the field.
http://www.scientificamerican.com/artic ... alth-plane
Of course, if everyone had one, then we'd have flying traffic jams and flying traffic accidents. Then we'd have debris dropping from the sky everywhere. So we would need a real, accurate, and automatic traffic-avoidance system. Ground-based like our current ATC system? on-board, peer-to-peer in a distributed learning system? I don't know, both would have advantages and disadvantages.
For a common vehicle, say 50% of the current on-road load, it would have to be simple enough for 50% of the driving public to be able to operate it safely. Either cross-linked controls (so you just turn a wheel, rather than wheel-and-pedals) or fully automatic controls. I would imagine no less than current requirements to get a license, say 25 hours of training, plus another 25 hours of operation with a licensed pilot on board. Also, I imagine the FAA would continue to oversee training and licensing as it does now for aircraft, rather than having 50 different sets of regulations and training requirements as is currently the case for gound vehicles.
