I’ve spent many hours thinking about driverless cars, and have even drafted a few blog posts. With the announcement the other day from Google, and the subsequent flurry of news coverage, it is time for me to join the party and get my thoughts out there.
First, my prediction: Self-driving cars will become standard.
The fact that self-driving cars will become the norm going forward has very interesting implications for cities and traffic. First of all (and to quote The Register), “everything we know about traffic is wrong”.
Transportation analysis is largely predicated on the idea that people travel for a reason. Most trips are made to move a person from one point to another. While there are a tiny minority of trips that are made purely for the sake of moving (the classic “”Sunday drive””), most trips serve a purpose for the driver and passengers (if any). The upshot is that traveling is considered a cost, both in terms of the cost of operating the vehicle, as well as in the time wasted traveling and not doing the desired activity at the destination.
The disutility of travel to the driver is actually a major component of the disutility of travel; take that away and travel isn’t quite so expensive.
Put another way, consider parking. One parks at a destination in order to get out of the car and do whatever activity is planned. The parked car takes up space, and in popular areas paying for a parking spot can cost a lot of money (even here in car-centric Southern California). With a driverless car, however, there is never any need to park. The passengers can just get out, and the car can circle the block endlessly until it is needed again. Given that parking costs $10 a day on a suburban college campus like UC Irvine, it would be significantly cheaper to pay for the cost of sending the car away, perhaps even sending it home again, until it is needed.
The impact on traffic might be drastic. Instead of the canonical “home-work-home” daily activity pattern, with two trips, you would see “home-work-drive-work-home” patterns, in which the car is potentially doubling its impact on congestion by driving home and back to work empty.
However, before we run around chanting that the sky is falling, consider that the only reason traffic is a problem is that travel is considered to be time that cannot be spent doing something the driver would rather be doing. So traffic jams won’t be such a big deal, because the entire time spent stuck in traffic can be devoted to work or leisure. So even if we have mega traffic jams resulting from double the numbers of trips being taken, we won’t care so much because the perceived cost of that time traveling will be significantly lower.
So really, we in the traffic profession will need to do a lot of rethinking of our methods and rerunning of our models. We are busily predicting traffic patterns 20 and 50 years out, and Google just demonstrated that we’re probably making mistakes.
To be continued
This is a parbaked post. I’ll be expanding and revisiting the issues in the coming weeks. But if I don’t get something up, I’m going to keep putting this off and refining and revising and nothing will come of it.