RSS

Robot cars may triple highway capacity

12 Sep

Gridlock: now imagine it moving at 100 kilometers an hour.

We’ve been hearing an awful lot about robot cars over the past year or so – about how they’re quickly advancing and will soon be here – but there have been few hard numbers on why we’d actually want them. Sure, self-driving cars offer the promise of making long road trips more productive and enjoyable and they’ll go far in eliminating road rage. But surely the developers behind such vehicles must think there’s more behind it than that if they’re going to go to all that trouble, right?

Some new research out of Columbia University, presented to the Institute of Electrical and Electronics Engineers conference on vehicular technology last week, hits that sweet spot. According to the study by Patcharinee Tientrakool, a fleet of self-driving cars could nearly triple the capacity of our highways.

She believes that a lone robot car that is able to manage its own speed would see an efficiency gain, in terms of fuel and time spent, of about 43 per cent. But the gains would really become apparent when the autonomous vehicle is pooled with similar-minded robot cars that it can communicate with. After all, a single robot driving on a highway full of actual people still has to contend with an awful lot of dumb humans.

If all the cars on the road are robots, however, things can really speed up. The report envisions the same sort of bumper-to-bumper traffic we currently experience on highways at rush hour (or here in Toronto, basically all the time), but the swarm would be moving at a much faster speed. Humans generally stay a second and a half behind the car in front of them; a robot could close that gap considerably, thereby allowing for a higher density of vehicles on the road.

This is the sort of thing that’s music to the ears of not only commuters, but city planners, road builders and politicians. Here in Toronto, our city council and mayor should perhaps spend a little more time thinking about how to implement robot cars and a little less on arguing about whether to build subways or light rail transit.

Advertisements
 
2 Comments

Posted by on September 12, 2012 in robots

 

2 responses to “Robot cars may triple highway capacity

  1. Marc Venot

    September 12, 2012 at 2:28 am

    The best would be to form a “train”, which means having a physical link between the components.
    Toronto should follow the lead of Vancouver and build a Skytrain network.

     
  2. russellmcormond

    September 12, 2012 at 2:11 pm

    Triple efficiency is tiny compared to the savings we could get if people didn’t own their individual cars, but instead used a car sharing system (public based as part of public transit network, or private like http://www.vrtucar.com/ )

    Being able to call up an APP on our phone and have the right sized vehicle (number of people, do we need to carry any large items, etc) just pick us up is far more efficient than having parking lots and other such costly overhead of the current car ownership culture.

    Good luck with the politicians part, though: Our transportation network in Ottawa is nasty partly because they can’t even move to a hub-and-spoke public system and have massive BUS congestion downtown. We also barely have a single light rail line, no subway or street-cars, and very little in the way of rational organization. Most of those on the transit commission aren’t even users of the system.

    Politicians will manage to mess up robot cars — ask yourself who should be in control of them? The allegedly “conservative” property-rights believing feds don’t even like the idea of computer owners controlling their own computers, so just who will be in control of these cars and for what purpose? (Note: As we know from the DRM/TPM debate, owning the car does not give you any more control or even access to know who is actually in control).

     
 
%d bloggers like this: