• Turun@feddit.de
    link
    fedilink
    English
    arrow-up
    2
    ·
    9 months ago

    For what it’s worth 2 out of 3 subway lines are already fully automated. They started in 2008.

    Trams have the same issues as self driving cars though: you need to 100% reliably detect people in front of the carriage. And you can easily find tests with Teslas which just run over a child sized doll because they didn’t detect it properly. The tech is just not there yet.

    • Humanius@lemmy.world
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      9 months ago

      I agree with you entirely that automated trams are more difficult than automated metro systems. However I do think that trams are a most likely a more easily solvable problem than automated cars.

      • Trams are restricted to their track, so the number of unique situations in which they can end up is more limited.
        Because of this you can model the environment in more detail.
      • Trams are large, heavy and commercial vehicles. So you can justify shelling out for more detailed sensors such as lidar etc, whereas on a Tesla you have to make due with merely a camera sensor.
      • You could potentially hire a dedicated person in a central location whose job it is to remotely get trams out of tricky situations.
        This would not remove the need for drivers outright, but could reduce the number of drivers you need per tram.

      That is not to say automated trams are easy, or already viable. I’m just saying that they are likely more viable than automated cars will be in the nearby future.