• piranhaconda@mander.xyz
    link
    fedilink
    arrow-up
    28
    ·
    5 days ago

    Beautiful use of machine learning and machine vision, get out of here with facial recognition everywhere, give me shit like this

    • cows_are_underrated@feddit.org
      link
      fedilink
      arrow-up
      3
      ·
      4 days ago

      The Problem is, that these machines are hella expensive. I heard about costs of about 1.00.000€ + support. The one that I saw also has a newer version, that is probably even more expensive. Additionally you can not even closely match the efficency of herbicides, since equipment designed to apply herbicides has a working range of up to 40m and you can drive with much higher speeds.

      So even tho laser looks like a promising solution they are not even anywhere close to being able to replace pesticides. There are other options for mechanically reducing weeds that may not match the efficiency, but are economically far more viable.

    • ShittyBeatlesFCPres@lemmy.world
      link
      fedilink
      English
      arrow-up
      10
      ·
      edit-2
      5 days ago

      Expert on shooting plants with laser beams here. Not quite enough power yet but a swappable battery system and a renewable source with a molten salt reactor could easily solve that problem.

      • anotherandrew@lemmy.mixdown.ca
        link
        fedilink
        English
        arrow-up
        2
        ·
        4 days ago

        EE (specifically embedded systems) here: just how much power do you need to zap a weed effectively? I would’ve thought a 40W laser would have been more than enough, and then scale that up for a hundred acres.

        only solar would be tough, but a small EV battery with a large panel in the sun for 12h seems like it’d be a lot of juice to run WeedZapper2000 with a topping off charge overnight, no?

        • cows_are_underrated@feddit.org
          link
          fedilink
          arrow-up
          3
          ·
          4 days ago

          The problem is not that you need to run a single laser, but multiple lasers in order to have some form of usable working width. Then you also need to power the cooling and the computer for the plant detection. If you build some form of small scale robot it should work with solar, but any machine that is designed to get the job done relatively quick probably won’t work.

          • anotherandrew@lemmy.mixdown.ca
            link
            fedilink
            English
            arrow-up
            1
            ·
            4 days ago

            Interesting; I thought this worked be a single high power laser (or a few) with galvanometers for targeting.

            Would love to learn more about how it’s really done (as opposed to how I imagine it’d be done).

            • cows_are_underrated@feddit.org
              link
              fedilink
              arrow-up
              3
              ·
              4 days ago

              The problem with a single laser is, that, in order to achieve bigger working widths, it would have to be able to rotate very accurately, very fast and not shoot your crops at the same time. Especially the latter makes it impossible to use a single one.

              You also dont have to “burn” the weeds. It is enough to destroy certain cells to stop them from growing.

      • BigTrout75@lemmy.world
        link
        fedilink
        English
        arrow-up
        1
        ·
        4 days ago

        Cool freaking tech. Build a micro version of this for yards that kills dandy lions and you’ll be rich.