Fully automated watering robot takes a big leap forward toward greenhouse automation

aquarius_robot

Greenhouse owners might find [David Dorhout]’s latest invention a groundbreaking green revolution! [David]’s Aquarius robot automates the laborious process of precision watering 90,000 square feet of potted plants. Imagine a recliner sized Roomba with a 30 gallon water tank autonomously roaming around your greenhouse performing 24×7 watering chores with absolute perfection. The Aquarius robot can do it all with three easy setups; add lines up and down the aisles on the floor for the robot to follow, set its dial to the size of your pots and maybe add a few soil moisture sensors if you want the perfect amount of water dispensed in each pot. The options include adding soil moisture sensors only between different sized plants letting Aquarius repeat the dispensing level required by the first plant’s moisture sensor for a given series.

After also digging through a pair of forum posts we learned that the bot is controlled by two Parallax propeller chips and has enough autonomous coding to open and close doors, find charging stations, fill its 30 gal water tank when low, and remember exactly where it left off between pit stops. We think dialing in the pot size could easily be eliminated using RFID pot identification tags similar in fashion to the Science Fair Sorting Project. Adjusting for plant and pot size as well as location might easily be automated using a vision system such as the featured Pixy a few weeks back. Finally, here are some featured hardware hacks for soil moisture sensing that could be incorporated into Aquarius to help remotely monitor and attend to just the plants that need attention: [Andy’s] Garden sensors, [Clover’s] Moisture control for a DIY greenhouse, [Ken_S’s] GardenMon(itoring project)

[David Dorhout] has 14 years experience in the agriculture and biotech industry. He has a unique talent applying his mad scientist technology to save the future of mankind as seen with his earlier Prospero robot farmer. You can learn more about Aquarius’s features on Dorhout R&D website or watch the video embedded below.


43 thoughts on “Fully automated watering robot takes a big leap forward toward greenhouse automation

      1. Irrigation that attempts to mimic rainfall tends to wast water. Often water that comes from underground reservoirs that recharge slowly if at all.

        1. You make a strong case against growing any kind of crops, or wells for that matter. So why single out just one watering method? If you want to single out any watering method then single out the one that is really causing the problems, pivots. Because it is pivot irrigation that is lowering water tables, not watering inside a controlled environment like a greenhouse, where if you wanted to you could reclaim all of the runoff.

    1. True, but this seems to offer better flexibility concerning layout and chances to layout, quicker setup and reconfiguration, and cheaper scalability (within reason).

  1. Looks awesome as a demonstration but I can’t see it as being commercially viable, it’s slow and wastes a lot of water. Water wasted is a big issue with growers.

    I also think it’s short of a couple of sensors, like to measure pot size and distance. I’ve seen machines become redundant when precise preparations are necessary to make them work properly.

    1. If you read through the forum posts linked toward the end of the article it looks like there has been improvement on the accuracy..

      “We later fixed the water spilling problem by moving the nozzles back a little bit and slowing down the robot a little bit more. It was quite the ordeal to get permission from Legal for us to shoot what you saw ”

      and they went low tech for a reason too

      “We intentionally went “low tech” with some of the systems so that the robot would be easier to maintain and operate by the greenhouse staff. We wanted them to be able to look at the robot and know exactly what was going on and how to set everything up correctly.”

    1. Tell Todd that he can keep the RFID tag idea and maybe next time just report on the actual story. 5/10 for his first post on hackaday.

      1. I disagree, I like the editorial otherwise it could have been just a link. I think it’s a good way of getting discussion going rather than waiting for the trolls.

          1. Well they aren’t literally “overshadowing” because we can still read all the original material, so you must be referring to some sort of imaginary or “virtual” overshadowing that probably exists only in your own mind. Intelligent people have developed the ability to “distinguish”.

            You see cross linking as some sort of nefarious plot, perhaps you could also see it as the entire point of the WWW and why we don’t just view pure text documents.

      2. I’m kinda with dALE on this one… there is nearly an entire paragraph dedicated to “how I would make your project better.” The post should be about the project, if you want to make suggestions or offer constructive advice, do it in the comments.

  2. I like the robot and maybe i is more cost effective then getting a bunch of pipes. I can also see this having the advantage of making the setup more configurable since you can move everything around without having to rip up the watering system.

    1. I think it has better promise in another venue. Like a mobile water cooler that drives around an office and waters plants. Better yet drive around and refill coffee cups.

    1. Something like this do you ?

      If a HortiBot^H^H^H^H^H^H^H^H Aquarius does go out of control, then could always call in Tom Selleck and the rest of the runaway robot squad.

      (In case other readers have never heard of it, there was a 1984 film called “Runaway” written and directed by Michael Crichton, staring Tom Selleck)

  3. Something reminiscent of Huey, Dewey And Louie from Silent Running (:
    It can open doors, Now if they could get it to use some vision system to identify a plant, test the soil. (And maybe play the odd game of checkers… :D)

    1. On further reading it can test the soil (:
      Quiet a neat little project. I’m wondering, with a few additions, how well it would cope with planting seeds. Monitoring which ones germinate. Analysing the end result and learning maybe even crating it’s own hybrids.

  4. This isn’t really any better than a bunch of pipes yet. With some clever CV algorithms it could see pots and their sizes based on their diameter. What’s more even sensing moisture levels could be done using CV because soil changes colour as it gets drier. Plus it would need some sort of robotic arm with a nozzle at the end to dispense water

      1. Having used automatic watering in a greenhouse, I can tell you it’s a pain! At least the stuff you use for smaller greenhouses.

    1. Running a bunch of pipes is not at all simple. this robot is mobile and flexible. All they have to do is put new tape on the floor for it to follow.

      1. It’s simple. Have you two ever installed irrigation?? I have and I actually worked in a greenhouse for a couple of years. No offence but you lack first hand experience to judge what’s practical.

        Once you set it up there is no need to move anything in most situations. In cases where you HAVE to move stuff you need something that doesn’t require putting tape. If you have to put tape to respond to changing conditions you might as well water them yourself. As I said in my original comment it is not any better than piping YET. For it to be useful it has to be able to respond to changing conditions, hence need for CV.

        1. I’ve installed drip line irrigation. I’ve done a variety of other plumbing jobs too. I’ve also been on a job where we had to burn the fibermesh out of a concrete pad for a line following robot with ox-acetylene torches. Those little hairs sticking up messed the robot up! No offense but you probably lack first hand experience compared to me. Although I have to admit I’ve never worked in a greenhouse. I demolished a greenhouse once though. For whatever that is worth. You pansy picker you!

          1. The point here is practicality and financial feasibility. Drip line irrigation systems are very cheap and do their job well and require almost zero maintenance. This robot doesn’t offer anything better than that. Sure once they can respond to changing conditions and navigate well, estimate soil moisture and condition of individual plants and deliver precisely the mount of water they need, have the ability to move plants, apply pesticides and fungicides or even harvest then they might make sense.

            Watering/irrigation is just something that’s already automated so it offers nothing in terms of cost reduction. There are other task that are very labour intensive. Robots are the future of farming for one particular reason – labour is expensive. If you can replace workers with machines you can increase profit (it does sound harsh, but that’s reality)

    2. When testing to see if a plant needs water you won’t be checking the moisture content of the soil on the surface… So a visual check is a no go. Also, as “F” and “dALE” mentioned, the robot is far easier to maintain and is more flexible than a bunch of pipes and valves. Also, small valves and “drippers” get easily clogged from minerals found in the water.

    3. Kerimil
      Sorry but using CV to jusdge soil moisture at the surface is not a good solution because the top layer of the soil can be dry but a few mm down it can be wet.

      Judging soil moisture content is hard for most humans much less a robot.

  5. I have always wanted to build a robo farmer to tend the crops while you do more productive things, (like building a greenhouse or something) but WHY would anyone waste a greenhouse growing CORN? or maybe it isn’t corn. perhaps the greenhouse is to protect the rest of the world from the pollen from this special corn plant. i just don’t see a greenhouse worth of corn being worth the space, unless, of course, the corn is bioengineered to produce some über valuable, hard to produce medicinal substance, what you should do is cultivate Senzu beans. MUCH more useful than corn. But who am I to judge? I should post pictures of my 3-4 month old plants that are 2-3 inches tall. I seem to have mastered the art of bonsai, or gotten a bunk batch of GA3.

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