I think the next step of mobility is a mobile robot. The cars of the future should be a central hub for move people between places, but able to do a lot more through physical manipulation, instead of becoming a enlarged mobile information terminal.
Thus, developing cars as robots are going to be more useful in that future.
I personally will invest time in building products that are mobility robots, not cars with giant screen.
Self-driving done right could save me from having two cars. It could take me to work, drive back home, pickup the wife, drive her to work. Go back and pick me up in the afternoon, drop me of at home and the drive back to get the wife.
Firstly we need to spread out peoples work day. We can't all be at the office a 8:00, and go home at 16:00. it wouldn't make economical sense in have a fleet large enough to handle peak hours, but have the cars do nothing most of the day.
There's also details like: I would like a car to pick me up at 15:45, and drive me to the kindergarten, and that car MUST have a seat for a three year old (but not a seat of brand X, Y, or Z, because they're to small). Or I need a car with a trailer, or a truck, because I'm moving 500kg of dirt to the recycling station.
Most people shop around for cars and pick very specific cars that fit their needs. So renting for a fleet of cars also need a rather large variety and flexibility in allowed usage.
I believe that many would still want to buy their own cars, even if it's self-driving, to ensure it's available when they need it, and that it fulfil their specific needs.
If we didn't have all these specific needs, most of us could just take the bus, which in turn would make it feasible to have all the bus routes people would need.
But that’s the current state of the world.
Even if you have enough cars to meet peak demand you will have an efficiency saving (at least where I live you see a lot of cars in driveways during rush hour traffic).
That said I agree that for a lot of people their car is an extension of their home, they’ll keep the golf clubs and a coat in the boot etc. But for the 80% case of just carrying a rucksack to and from work there is a massive efficiency improvement to be had.
True, but the difference is who's carrying the financial risk of the buying the car. Right now it's the individual consumers, who are mostly forced to do so, because better alternatives are not available.
Leasing companies would never buy a fleet of cars that would sit mostly unused. Their cars needs to be on the road most of the time.
This puts us back in Uber vs. taxis. Uber works because (depending on area and regulation) there cannot be sufficient taxis to handle peak demands, that wouldn't be financially sound for the taxi companies. Uber tries to fix this, by taping into idle cars in the driveways. A company that maintains a fleet of cars that you can just order on-demand, would always have to few cars for rush hour.
car-sharing already exists on a large scale in German capitals. you can park and pickup cars from e.g. Sixt or Share Now everywhere within a marked area. the business model for owning and maintaining a big fleet is already there.
It is extremely common for markets to face bursts in demand at peak times, and yet somehow we always find a sustainable equilibrium.
Thinking through what it costs me to own a car, I'm going to carefully do the math if they price my trip into the office at more than about $5 (it's a short drive). That's probably the bigger problem.
Supply and demand with dynamic pricing will sort that out.
Consider also that autonomous vehicles could make much more efficient use of existing road capacity.
> If we didn't have all these specific needs, most of us could just take the bus
Buses do not take you directly to your preferred location, without stopping, with your own private space, etc.
I find it confusing that one reason the tech crowd cheered on Uber's success was because taxis were often filthy and poorly maintained. But autonomous car fleets won't be?
Cars in that mode could be constrained to a less dangerous speed to limit damage in an accident. I don't want to drive to work at 20km/h, but once I'm out of the car, I don't really care how slowly it goes.
Of course, it wouldn't necessarily work to combine too broadly differing speeds on the same roads. But it would remove a bit of the "I'm not putting on my VR goggles to do work and trusting this car to drive to the office" fear.
If there was a common communication standard, the vehicles could also park much more tightly than is done in a normal parking lot. You don't need to keep empty space behind each car when you can coordinate the parked cars to move and clear a path. My guess is that you could park almost twice as many cars in the same space.
The combination of reducing parking lot size and freeing people from walking to/from the lot would open up many interesting options for increasing density and walkability in commercial areas.
The other people on the road trying to get to work probably do.
"Of course, it wouldn't necessarily work to combine too broadly differing speeds on the same roads."