Stories
Slash Boxes
Comments

SoylentNews is people

posted by janrinok on Thursday May 23 2019, @01:08PM   Printer-friendly
from the need-smaller-antennae dept.

The researchers, from the University of Cambridge, programmed a small fleet of miniature robotic cars to drive on a multi-lane track and observed how the traffic flow changed when one of the cars stopped.

When the cars were not driving cooperatively, any cars behind the stopped car had to stop or slow down and wait for a gap in the traffic, as would typically happen on a real road. A queue quickly formed behind the stopped car and overall traffic flow was slowed.

However, when the cars were communicating with each other and driving cooperatively, as soon as one car stopped in the inner lane, it sent a signal to all the other cars. Cars in the outer lane that were in immediate proximity of the stopped car slowed down slightly so that cars in the inner lane were able to quickly pass the stopped car without having to stop or slow down significantly.

Additionally, when a human-controlled driver was put on the 'road' with the autonomous cars and moved around the track in an aggressive manner, the other cars were able to give way to avoid the aggressive driver, improving safety.

The results, to be presented today at the International Conference on Robotics and Automation (ICRA) in Montréal, will be useful for studying how autonomous cars can communicate with each other, and with cars controlled by human drivers, on real roads in the future.

Sources:

[Editors Comment: The submitter is linked professionally to the last of the listed sources. Additional source material, including the original paper from Cambridge University as primary source, is also listed.]


Original Submission

 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 4, Insightful) by esperto123 on Thursday May 23 2019, @01:46PM (11 children)

    by esperto123 (4303) on Thursday May 23 2019, @01:46PM (#846638)

    The same was said about Uber and Lyft, but the reality is that it ends up increasing the traffic as it makes cheap to go from one point to another in a car, individually, instead of using public transport or carpooling, and empty cars going about looking to get a new passanger.
    It is very likely that autonomous cars will make traffic more efficient but will, in my view, very likely make traffic worse.

    Starting Score:    1  point
    Moderation   +2  
       Insightful=2, Total=2
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   4  
  • (Score: 3, Insightful) by ikanreed on Thursday May 23 2019, @01:59PM (9 children)

    by ikanreed (3164) Subscriber Badge on Thursday May 23 2019, @01:59PM (#846641) Journal

    Uber and lyft were never going to decrease traffic. There was no theoretical mechanism to allow that to happen.

    Driverless cars, while causing a similar increase in demand and total driving, at least have the theory of reducing certain kinds of driver-caused congestion by having better response times, when a magic percentage of cars on the road are driverless.

    • (Score: 5, Insightful) by AthanasiusKircher on Thursday May 23 2019, @02:49PM (7 children)

      by AthanasiusKircher (5291) on Thursday May 23 2019, @02:49PM (#846651) Journal

      Uber and lyft were never going to decrease traffic. There was no theoretical mechanism to allow that to happen.

      Well, there was a theoretical mechanism for it to happen. Anyone remember the several years when Uber was going around pretending it was about "ride-sharing" and carpooling rather than a taxi service? Carpooling could theoretically decrease traffic, but of course Uber pretty early transitioned to a substitute taxi service rather than "ride sharing."

      So yes, the way Uber originally presented itself, there was a theoretical mechanism for it to reduce traffic.

      Driverless cars, while causing a similar increase in demand and total driving, at least have the theory of reducing certain kinds of driver-caused congestion by having better response times

      Yep, and again, it's theoretical. Because, as noted by others already, what happens when people send driverless cars around for new tasks? What happens in major cities where parking is expensive (and hard to find), and a driverless car owner decides to send the car circling around rather than finding a parking spot (and paying for it)? What happens when thousands or tens of thousands of people do this in a major city on a popular evening?

      Obviously there will probably be new laws about such things, and maybe the advantages will outweigh the disadvantages. I think eventually they will -- but that could be in the far future. In the short term, I wouldn't try to predict whether the theoretical advantages will actually improve traffic overall.

      • (Score: 3, Touché) by ikanreed on Thursday May 23 2019, @02:52PM (2 children)

        by ikanreed (3164) Subscriber Badge on Thursday May 23 2019, @02:52PM (#846653) Journal

        What if we got a whole bunch of cars and linked them together with some kind of mechanical mechanism. And... instead of a road we put them on some kind of metal rails that force them to all follow exactly the same path reducing drag. And we only put an engine in the front one. The gas savings would be huge.

        Sure you'd have some challenges with planning, but you could have regular stops that everyone knows they can get picked up at fixed intervals.

        I think I'm on to something here.

        • (Score: 2) by Gaaark on Thursday May 23 2019, @08:55PM

          by Gaaark (41) on Thursday May 23 2019, @08:55PM (#846789) Journal

          But then you'd have to have some kind of standard for time, like dividing the country into different zones of time so everyone would know exactly what time the train was arriving.
          I say we call them 'standard time' and 'time zones'.
          I'd implement them in Canada first, though. (Takes off stove-pipe hat).

          --
          --- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
        • (Score: 2) by Gaaark on Thursday May 23 2019, @08:58PM

          by Gaaark (41) on Thursday May 23 2019, @08:58PM (#846791) Journal

          You'd also need to discover some way of communicating with the trains so they don't crash into each other!

          I guess you could start building parallel rail lines, though.....hmmmm...I like your idea. Just need to work out the kinks.!

          --
          --- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
      • (Score: 2) by bob_super on Thursday May 23 2019, @05:21PM (2 children)

        by bob_super (1357) on Thursday May 23 2019, @05:21PM (#846713)

        Car-sharing should reduce parking problems, because cars don't spend their whole day using a spot.
        Car-sharing could reduce congestion, with the right app scheduling people to actually optimize vehicle capacity, including changing cars along the way. (buses, but car-sized and with dynamic routes). Hasn't happened yet, but it is theoretically possible.

        This is about intelligent cars which are programmed to share the road capacity better than selfish humans.
        Proper dialoging and yielding between cars can remove almost all traffic stops, and likely most of the jams.

        My feedback: Thanks "researchers" for playing with robots, but the conclusion was known before you started, and you could have done it with a sim, like the professionals working on infrastructure do every day.
        No, the "results" will NOT "be useful for studying how autonomous cars can communicate with each other, and with cars controlled by human drivers", because that was fucking obvious to start with, and the method doesn't scale like the existing sims.

        Now, I'll go back to SupCom2, released almost a decade ago, in which hundreds of units of various sizes and speeds flow around each other and obstacles.
        https://www.youtube.com/watch?v=bovlsENv1g4 [youtube.com]

        • (Score: 2) by Gaaark on Thursday May 23 2019, @09:01PM (1 child)

          by Gaaark (41) on Thursday May 23 2019, @09:01PM (#846793) Journal

          Ever played Total Annihilation, the forefather of SupCom?

          Man, the units used to get in each others way ALL the fecking time.

          Good game, though. I still play it.

          --
          --- Please remind me if I haven't been civil to you: I'm channeling MDC. ---Gaaark 2.0 ---
          • (Score: 2) by bob_super on Thursday May 23 2019, @09:20PM

            by bob_super (1357) on Thursday May 23 2019, @09:20PM (#846803)

            Yes, but the problem with these old games is to find humans to play them with, because the AIs were pretty stupid at the time.
            Supcom FA + LUA mods started having decent AIs that didn't result in one side getting crushed easily.

            Put a good AI in TA, and you don't need any newer strategy game.

            Put a good I (Organic or Artificial) in all cars, and 40 miles to L.A. could stop feeling worse than leaving to cross the Atlantic during peak hurricane season.

      • (Score: 0) by Anonymous Coward on Friday May 24 2019, @09:17AM

        by Anonymous Coward on Friday May 24 2019, @09:17AM (#846986)

        What happens in major cities where parking is expensive (and hard to find), and a driverless car owner decides to send the car circling around rather than finding a parking spot (and paying for it)?

        Here in Europe (Belgium) there's talk about new car taxes based on when you're driving and where.
        Driving in rush hour in inner city? That's €x / km or €x / h. Driving in normal hours would be cheaper and sticking to main roads as well. The idea is to give incentive for driving outside rush hour times and to use public transport where possible. (e.g. higher cost for driving in cities)

        The idea is nice, but the consequences on implementation are to scary to be for it. (Trackers in all cars at all times)
        Next to that, it's probably a money grab to prepare for when a large part of the cars becomes electric. (Bye bye hefty gas taxes)

    • (Score: 2) by Freeman on Thursday May 23 2019, @08:39PM

      by Freeman (732) on Thursday May 23 2019, @08:39PM (#846784) Journal

      All it takes is one car to hold up a huge line of traffic. Unless all cars are mandated to be driverless, which I don't see happening for a goodly long time. It's all well and good to say driverless is the wave of the future! Yeah, how much does that future cost?

      --
      Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
  • (Score: 2) by Immerman on Thursday May 23 2019, @02:42PM

    by Immerman (3985) on Thursday May 23 2019, @02:42PM (#846650)

    If we're essentially talking a giant autonomous taxi service anyway, what if we add a major additional change: almost all the cars are single-seaters that can lane-split to fit 2-4 of them in the same road-space as a normal car, depending on speed. If you need more space you can call a 2-seater, 4-seater, or cargo van, but the cheap ubiquitous option is little more than a chair in an weather shell. Park them packed in like sardines in any available nooks and crannies so there's always a few parked nearby. Sort of a hybrid of car-hailing and electric scooter rentals.

    There's also another benefit to automation and the large-scale coordination it enables - far more seamless mass transit. You can take a local single-seater from your doorstep to the nearest high-speed bus going your way, knowing that another single-seater has already been booked and will be waiting for you when you reach your stop.