BMW to crowdsource the design of their megacity-centric future car
extremetech.com
extremetech.com
http://simpsons.wikia.com/wiki/The_Homer
Seriously though, I'm not sure if the future megacities will really have many cars, at least not the personal vehicle variety.
I'm sure BMW is looking for ideas, but IMO the solution is outside of the box they're willing to work within (i.e., private car ownership that is substantially similar to what we have today).
With Google's self-driving cars, we would have driverless taxis that can take you somewhere, then go pick up someone else that called them with an app. Simple, convenient.
The main reason to have your own car is for image and largesse. In a world of scarce resources, it might make more sense to share the cars. Might even have taxi carpooling.
Point to point would be great, but it's really, really expensive and doesn't change the economics of land use.
Look at what cars cost. Most US cities have ~1 car per adult, and to average expensive cars and paid off cars, let's assume ~$5000 in payments per car per year, plus about that much for gas, maintenance, insurance, etc. So ~$10,000/adult per year.
Now assume with automated p2p 1 car can do the work of 10 private cars at the same vehicle cost. But barring some efficiency changes, operational costs would be the same as 10 cars, so each car would cost ~$55,000/yr. This means $5,500/adult, about half the cost of a private car - big savings!
For a city of 1M, assume 50K cars, and you need almost $3B/yr just for the cars, the whole organization (gov or private) would probably cost another $1B. $4B/yr would be a big savings over private vehicles and free up all the space used by parking.
Big caveats:
1) For reference, Nashville (~650K city, ~1.5M metro) had a mass transit budget of $60M, or about 1.5% of this back of the envelope budget. The entire municipal budget was $1.5B, or about 37% of this budget. This demonstrates that however beneficial such a system would be, it' would require a complete restructuring of government and society to shift that kind of spending into centralized transportation infrastructure.
2) These fairly affordable unit economics only work if everyone participates. But the system also only works if you continue to maintain the existing road network, which makes the switch from private cars that much more difficult to justify.
3) A p2p network is an attempt to get people who live in non-transit areas to use transit. But many (many many many) of those people live there specifically because they want privacy and seclusion from other people, so any kind of transit system is wrong on principle.
Which obviously leads to an otherwise not obvious major design factor: round. Once the norm is for the occupying humans to not have much, if any, control in driving, the paramount drivetrain issue will be how to move the vehicle in any suitable direction on very short notice. Replace the "front wheels turn a few degrees" paradigm with zero-turn-radius or any-direction movement, and the vehicle's ideal shape is a circle.
Seems obvious to me.
What's not obvious to me is how crowdsourcing will come up with such a model. To wit: how does crowdsourcing overcome non-ingrained memetic patterns?
ETA: BMW is partnering with crowdsourcing pros Local Motors of Phoenix to run the contest. Local Motors already has a crowdsourced, on-sale car to its credit, the Rally Fighter. Then the outcome of this contest will be amusing, as the Rally Fighter seems the utter opposite of the design pursued here.