​Self-driving cars could cut car ownership rates nearly in half: study

A single self-driving car could take care of multiple people's daily driving needs, so households might not need as many cars, University of Michigan researchers say.

The first fully functional prototype of Google's self-driving car debuted in December 2014.
The first fully functional prototype of Google's self-driving car debuted in December 2014. Google

With a single self-driving car able to shuttle people on multiple trips during the day, vehicle ownership could drop as much as 43 percent, according to a new study from the University of Michigan Transportation Research Institute.

Despite being expensive investments, today's cars idle away a large percentage of their existence in garages and parking lots doing nothing. Self-driving cars, though, could work more hours of the day and, for some, take the place of more than one car, said researchers Brandon Schoettle and Michael Sivak in the study.

The researchers concluded it's possible US households could cut car ownership from today's average of 2.1 to 1.2 with self-driving cars. They based their conclusion on an analysis of US National Household Travel Survey data (PDF) about how much people drive and how often the timing of those trips doesn't overlap.

The researchers cautioned that the number is an estimate of the most dramatic change possible and that several factors could in practice make for less of a shift. For example, it's possible that people would take more trips with self-driving cars or that people would be reluctant to share vehicles.

Still, the number points to how radically transportation could change with the advent of self-driving cars. Children and the elderly could get new freedom to move, safety could increase under the control of non-drowsy computers with a 360-degree awareness, and people could get new hours in the day for checking email, watching videos, telephoning colleagues, reading books or taking a nap.

The car industry is gradually adding more automation to cars, starting with smaller steps such as parallel parking and computer-initiated emergency braking. More sophisticated self-driving car prototypes could eventually do just about everything, getting a passenger all the way from start to finish. Google has drawn a lot of attention with its research into self-driving cars -- at one point, it was ready to do away with the steering wheel, but its latest prototype keeps traditional controls to satisfy California's testing requirements -- and traditional car manufacturers are also moving forward with their own programs.

As cars get more like computers, automakers will also have to make cybersecurity a higher priority. A study released this week by Sen. Edward Markey (D-Mass.) found that virtually all new cars -- the kind that people are still driving themselves -- may be vulnerable to hacking and privacy violations.

Car usage could increase with self-driving cars, according to the University of Michigan study. For example, a car could drop one person off at work, return home, perform errands, then return to the workplace later to collect the first person.

If that happens, cars will travel more per year and wear out faster, the study said -- 20,405 miles per year on average instead of 11,661 today. There's a silver lining, though: faster upgrades.

"In the hypothetical scenario we describe in this analysis," the researchers said, "with an average turnover rate that is 75% faster for these vehicles, the rate at which new self-driving technology would be replaced or updated in the on-road fleet would nearly double versus current average vehicle age and annual usage rates."

Featured Video