I grew up a poor white boy-child in Santa Cruz, which means that there were three ways in which I was privileged, and only one of them was location. Growing up with a single mother with no car when surrounded by the affluent and mobile meant that those were the only ways in which I was privileged. Any place I did not reach via operation of my Chevrolegs had to be visited via the shame train, also known as "The Bus". And riding the bus incurs a stiff penalty.
When I began riding the Santa Cruz Metropolitan Transit Department's buses, they were 1970s GMC T6H and T8Hs, really beautiful sweeping things with glass everywhere but with all the ride comfort of a cinder block. With all that clear, uncoated glass they were total greenhouses and the AC was inadequate even to the demands of the central coast. You'd arrive at any destination more than fifteen minutes away cramped and sweaty in all but the coldest weather, when you'd only be cramped. Subsequent revisions to city buses have improved these particular faults, but have remedied none of the others.
The single greatest problem with the bus is road damage. When fully loaded with passengers, the pressure loading on the pavement is as bad as a heavy truck, or even worse. This is important because the pressure per square inch combined with a heavy load is what actually damages pavement. It is not an exaggeration to say that a bus does thousands of times as much damage to pavement as a normal passenger vehicle, like a van. In fact, ordinary vehicles can drive on properly prepared roads (with sound road beds and proper asphalt composition) essentially indefinitely without damaging them; weather does more damage to roads than essentially any amount of light vehicle traffic.
Another problem with buses is excess capacity. In order to provide adequate availability over portions of the route, it is necessary to provide excess availability over other portions. We pay for this excess capacity; buses pay a substantial fuel penalty for having to start and stop all that mass over and over again, even when it is not being used. That means they also pay a substantial emissions penalty, with higher energy cost and CO2 emissions per mile than passenger cars in many situations. Hybrid systems mitigate the mass problem by permitting recovery of some of the kinetic energy normally wasted during braking, but they exacerbate it by forcing buses to carry heavy battery systems, and the control systems needed to make electrified vehicles on that scale are complex and expensive. Both of these problems are even worse for other types of vehicles, like trains. With buses, it's at least possible to simple route traffic down other streets in many cases. But that kind of flexibility is even more present with other types of vehicle.
Simply having a vehicle as large as a bus mixing with automobiles on city streets causes problems of its own. Having a large vehicle continually pulling over and stopping, often within the roadway, endangers both the passengers and other motorists. These problems can even be magnified in rural areas, with narrow, winding roads which present a challenge for buses in even the best weather acting as an outright show-stopper in inclement weather. Making the vehicle smaller mitigates or even outright solves each of these problems.
What is the solution to our bus-related woes? Most automakers seem to be betting on the autonomous van or minivan as the mobility solution for the masses for the near future. They offer unbeatable packaging for their forward area in terms of usable cargo capacity, and their profile is an ideal match for electrification, whether full or partial. Waymo is deploying autonomous Pacificas and Ford is working up autonomous Transit vans as we speak, as well as Fusions. Both the Fusion and the Transit (as well as the Mustang!) are supposed to be available in electrified trim in 2020.
One thing that's exceptionally beneficial about the van shape is the ability to enable platooning, where self-driving vehicles ride just inches off one another's bumpers to minimize wind resistance. This is especially beneficial if you use a very boxy, square shape, as in Ford's driverless delivery van concept.
Besides solving many problems with buses, vans also provide other benefits. Uber and Lyft show that consumers would unsurprisingly prefer being picked up from their current location from having to visit a transit stop, and appreciate knowing precisely when their ride will arrive. A public transit system using automated vans could perform the same function, dispatching vehicles on-demand to locations where there are multiple riders, and crafting routes that deliver them all to their destinations as rapidly as possible and with a minimum of vehicle transfers.
In addition, because vehicles are mostly freed from following fixed routes, any particular vehicle might be dispatched across town repeatedly, or it might stay mostly on one side of town. Its activity could even be planned to match its current battery charge condition, so that a vehicle approaching a recharge event could be used for shorter trips with more regeneration, while a freshly-charged vehicle which has no free battery capacity for regen might be sent out on a longer trip. Passengers without access to mobile internet access might still have to travel to centralized locations, where they could use a kiosk to request transportation. Bus stops could feature micro-terminals, possibly as simple as a call button in neighborhoods in which more complex arrangements are stolen or vandalized.
The only reason we use buses today is that they permit a single human driver to transport many passengers. Everything else about them is inferior to smaller vehicles, and as autonomy is developed, the need for buses will decrease until they are entirely outmoded. While there are questions to be asked about what bus drivers will do in a post-driver world, the benefits of eliminating the bus from the transportation landscape far outweigh the drawbacks.