Robot Lives in Your Garden and Eats the Weeds

You can’t deny the appeal of gardening. Whether it’s a productive patch of vegetables or a flower bed to delight the senses, the effort put into gardening is amply rewarded. Nobody seems to like the weeding, though — well, almost nobody; I find it quite relaxing. But if you’re not willing to get down and dirty with the weeds, you might consider deploying a weed-eating garden robot to do the job for you.

Dubbed the Tertill, and still very much a prototype, the garden robot is the brainchild of some former iRobot employees. That’s a pretty solid pedigree, and you can see the Roomba-esque navigation scheme in action — when it bumps into something it turns away, eventually covering the whole garden. Weed discrimination is dead simple: short plants bad, tall plants good. Seedlings are protected by a collar until they’re big enough not to get zapped by the solar-powered robot’s line trimmer.

It’s a pretty good idea, but the devil will be in the details. Will it be able to tend the understory of gardens where weeds tend to gather as the plants get taller? Can it handle steep-sided raised beds or deeply mulched gardens? Perhaps there are lessons to be learned from this Australian weed-bot.

44 thoughts on “Robot Lives in Your Garden and Eats the Weeds

  1. I would worry about delicate plants (parsely, carrot greens, and dill for instance) suffering damage to their stems from a line trimmer sturdy enough to knock down the tougher weed seedlings.

    1. They sell wire guards that go around baby plants. The weed eater is completely underneath the body of the robot; doesn’t stick out any. It’s a pretty good design actually.

      1. If you’re only managing the weeds that are in the bare areas around the plants you’re just plucking the low-hanging fruit. Removing weeds in the open areas is easy to do with a hoe; the time consuming part is getting all the weeds right next to the plants.

        1. If you can’t solve all the problems of all possible operating conditions/environments perfectly with a prototype, don’t waste time sharing your invention on the Internet, amirite?

    1. The version pictured is like the second generation of prototypes. The sent a press kit with renderings of later generations, which are getting to look more like the Roomba form-factor and have a much greater area for solar cells.

    2. I was thinking the same until I thought that maybe it moves into a good spot for sunlight and just sits there for twenty three and a half hours. Spends 30 minutes working, over 365.25 days that would work out the be 10957.5 minutes (little and often is really what is required for good weeding).

  2. For some odd reason I’m now thinking about an old Michael Crichton film called Runaway with a malfunctioning robot rolling through the middle of a corn field (in the opening scene).

    1. In addition to the cast of robots it also stars Tom Selleck, Gene Simmons, Kirstie Alley, and Cynthia Rhodes.

      I like the triangle shaped camera drone with a prop in its center. The technology exists now (and has been featured on HaD) to make one of those for real.

      “Runaway” could’ve been a blockbuster hit, but for crappy marketing and Beverly Hills Cop, Indiana Jones and the Temple of Doom, Gremlins, and Star Trek III sucking up all the money.

  3. Many weeds grow faster and taller than the food plants, so tall vs short won’t work there.

    You’d have to manually weed until the food plants are above this robot’s threshold, then turn it loose to keep all new growth cut down.

    How about an updated version of the Sizz-Weeder? It used fast moving blades of flame to kill small weeds while not harming mature row crops.

  4. If only it could use openCV and big data to identify the plants, give you a list of plants it found in your garden and you just select the ones you want to keep.

    1. So why bother putting it on HaD…

      Because the concept is interesting?

      Because some of us are willing to sign up to actually read about?

      Because even the amount of info they give you on that first page is enough to inspire some of the smart people around here to start work on their own, without needing any further details?

      I think you were looking for instructable-a-day.

        1. Don’t mistake plants for bacteria. :-) The sun has plenty of UV. So most plants are quite used to it. At least if you do not concentrate it e.g. with a big magnifying lens or use a (UV) laser of sufficient power.

        2. You’d need very short wavelength UV (which is hard/expensive to produce with LEDs) and even then it would take a long time and a lot of power…plants are very resiliant to UV…

          A 1-3W NIR or blue laser seems like a much better option, they’ve become really affordable and at close range you won’t mind the crappy beam characteristic ;-)

        3. I researched online an experiment with using microwaves to sterilize soil by destroying seeds. Didn’t work well, but they didn’t test it on live plants either.

  5. the problem i foresee with this is many weeds will regrow quiet happily from a root or clipped stem, meaning the ultimate outcome is a garden full of weed nubs.
    Realisitically they should release a version which can blast the weeds with glyphosate.

    1. No need for glyphosphate on the small scale. Even on the medium scale. If you don’t till & mulch you should only have to weed once or twice in the spring as weed seeds germinate. For hardier weeds like established dandelions you can pluck the growth every week or so to starve our reserves. Or just cover with a section of your newspaper.
      Glyphosphate may be safe enough (TM) for humans but it’s not great for soil health.

    2. That’s actually not a bad thing. In Permaculture they use chop-n-drop and leave the unwanted plants right where they are cut, for added mulch and nutrients. If more spring up, more chop-n-drop. This would do that automatically.

    3. If I really want to use glyphosate in the garden, I would prefer to spray it out manually, just to be sure it gets only where I want it. Spraying is much less work than manual weeding.

  6. You should not trim bad weed, but remove them completely otherwise you end up with a lawn of weed.
    There is also a lot of weeds that do not stand tall but cover all the surface quickly (sorry don’t know the english name of these), and they are the worst when they end up sticking on the plants themselves.

  7. +mac012345 Was just going to say that too… with some kinds of weeds, you’re doing them a favor by trimming them off. They need to be pulled. It may be possible to make a robot with a set of rollers that actually grip and pull the weeds up as it drives over them.

      1. Well, I guess that’s up to the consumer. I’ve only grown a garden once or twice, and weeds were not a big issue for me. I think I just went out once a day to check on the garden and pulled them by hand. I’ve seen people buy far more dubious things in my life.

        I told these guys what I really want is a robotic lawn mower. It doesn’t even need a fast and dangerous spinning blade. If you could figure out a way to safely clip the grass, you could let it run pretty much constantly, without having to worry so much about it hurting anyone. [jellmeister] had an interesting start some months ago. Hopefully he just put it on the back burner for winter and will return to it soon: https://hackaday.io/project/12642-solar-lawn-sheep

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s