3 ms·

I think 5 years isn't necessarily a bad estimate for full self-driving cars, but I'm very skeptical that is in all traffic conditions, operating with mostly hum

I think 5 years isn't necessarily a bad estimate for full self-driving cars, but I'm very skeptical that is in all traffic conditions, operating with mostly human drivers, in all road "structures", and especially in all weather conditions. Does anyone here have some experience with Tesla's Auto Pilot on snowy roads? Do Teslas have camera that try to read speed limit signs? If so has there been any problems when the signs were mostly or completely covered with snow? How does Auto Pilot do on ice?

Someone 7y ago
That’s quite a loose definition of “full”. I think we’ll make from making small, but significant progress, but not get too self-driving soon. For example, a convoy of two trucks of which the second simple follows the first at short distance would decrease personnel costs and fuel economy, and increase road capacity. A step up would be self-forming convoys of many cars on highways. Then, drivers in the cars would have to trust the driver of the first car, or that car would have to have good automatic lane-keeping. That problem can be simplified by changing the road design. The result wouldn’t be autonomous and wouldn’t work everywhere, but who cares if it makes things safer and more economical?