So is the expected model that people won't own their own AV but there'll be a massive network of AV taxis constantly ferrying people around and being shared?
If you look at the public statements of executives at AV leaders like Waymo, GM Cruise, Ford Mobility, etc, they all plan a commercial ride hailing service first and well before the concept of AV ownership by individuals is contemplated. For now, an AV is expected to be a very costly vehicle---I have read estimates that a single Waymo AV which in essence is a modified Fiat Chrysler Pacifica SUV with LIDAR, sensors, radar and cameras and software and hardware costs $ 250,000 per vehicle to deliver.
As my previous link to efforts in the US Congress provides, even if proposed legislation passes exempting Level 4/5 AVs passes, it will be with severe AV number limitations that make it difficult to achieve true economies of scale and hence huge cost reductions. That is why Waymo decided to scrap the idea of producing its own vehicle back in 2015---remember the Firefly? Instead, the vehicle that Fiat Chrysler produces for them is a fully compliant, road worthy vehicle that has been modified by Waymo engineers and hence why it still has a steering wheel and brakes despite it never being used by the "driver". For example, the Fiat Chrysler Pacifica SUV driven by Waymo software and hardware has room for 7 passengers but because the front seat must remain empty to prevent any passenger from driving the vehicle, it is limited to 5 passengers.
Next think Return on Investment. The average driver today in a manually controlled vehicle uses the vehicle perhaps 4 or 5% of the time and the vehicle remains idle for 95-96% of the time (parked in a driveway, on the street or in a car park). An EV costing $ 250,000 to deliver would simply be uneconomic for the average human driver to own. Waymo's commercially used vehicle will probably see 60% usage with the remainder of its life spent being recharged or repaired. With these economics it should be possible for Waymo to turn its commercial intentions into a very profitable business. How so?
I believe that Waymo can build a fleet of commercially owned vehicles into hundreds of thousands and potentially millions of vehicles that operate first in the US and eventually globally by a combination of owned and operated vehicles and through partnerships with today's ride hailing services like Uber and Lyft. Lyft (recently valued at $ 11 billion in the private market) already has such a partnership with Waymo and others. Uber has stated publicly that they would like to enter into such a partnership with Waymo. Waymo already has other partnerships with Avis (car rentals) and Autonation (car parts supplier) to include servicing and repairing their fleets. Waymo also has entered a partnership with an insurance company to provide passenger insurance for each ride.
The economics: if Waymo can grow its business to represent 1% of global miles driven by 2030 (based on a fleet of 3 million cars driving 65,000 miles each per year) and that Waymo can generate an average of $ 1.25 in revenue per mile driven, that equates to a valuation of $ 75 billion today using discounted cash flow to provide EV (economic value). So a large, publicly traded AV company like Waymo (which I believe will be spun out of Alphabet) could be a huge player in the commercial AV ride hailing business. It could also be a huge purchaser of vehicles from traditional auto manufacturers in its own right.
That is the economic model I see developing first. Such a model could also be applied to trucks and to other types of vehicles such as delivery vans (with robotic assistance at the curb).
