Even self-driving coaches will have congestion issues. Even if there are dedicated travel lanes and all vehicles are connected, capacity will (probably) at most

Even self-driving coaches will have congestion issues. Even if there are dedicated travel lanes and all vehicles are connected, capacity will (probably) at most double (going down much further as speeds increase, since a safe braking distance is still needed). In that mode, the new techniques that are getting developed in rail signalling (ETCS level 3 moving block) might be transferred in some way, but capacity is still very much limited. If E-Coaches are shared, you'll have difficulties getting them to you at the point & time you want to travel (probably requiring at least ~15-30 min prior information in non-urban environments, where congestion doesn't limit adoption). They'll also have to drive in-and-out of multiple residential areas (if shared), which will increase travel time. I think self-driving cars (personal use)[1], autonomous electric busses (with many of the limitations as current busses, but lower cost & higher comfort as well as more flexible (smaller) capacity options) and rail transit will cover most demands, with on-demand transportation covering the night hours and uncommon travel routes. Some transit agencies are experimenting with services like ViaVan (providing most of the features of self-driving electric coaches, excluding the self-driving part) and adoption seems to be low, mostly working as a complement if no other attractive transit options exist, not as a displacement. [1]: I fear that this will be primary form of transportation, since costs aren't going to be much higher than current cars (or perhaps even lower). The possibility of doing something other than driving will make congestion more palatable and searching for a parking spot will probably not require a human on-board. There will still be a personal space owned by someone and stuff can be (permanently) left inside the car.