DIY AI Weed Burner Robot | Robot weeds lawns, gardens, farms | Full build plans, parts list
ฝัง
- เผยแพร่เมื่อ 18 มิ.ย. 2023
- Hello Weed Burner, Goodbye weeds!
This is a DIY Autonomous Intelligent Weed Burning Robot. It AI to identify weeds and the power of the sun, focused to a point to burn weeds to death. In this video I detail how I built it including parts list, schematics and code to run it. Links in for those resources are in the description below.
Resource links:
Take Photos script - drive.google.com/file/d/1RGre...
Divide images and sort script - drive.google.com/file/d/1kI0E...
Build model script - drive.google.com/file/d/1qyks...
Control Motors script - drive.google.com/file/d/1qHJB...
Weed Killer, main program - drive.google.com/file/d/1RwLf...
Wiring Diagram - drive.google.com/file/d/1TAQd...
Raspberry Pi setup commands - drive.google.com/file/d/1ARah...
Parts List:
1. raspberry pi & peripherals
1 pack SD cards - 32GB each x3 - $12 - www.amazon.com/dp/B07YXJM282?...
1 raspberry pi zero 2 w - $15 - www.adafruit.com/product/5291
1 raspberry pi camera 3 - $25 - www.adafruit.com/product/5657
1 mini HDMI to HDMI cable - $8 - www.amazon.com/dp/B08DK4LVYX?...
Assume you have a monitor and keyboard/mouse used during debug, so those aren't to be purchased
2. fresnel lens
1 2 pack 300mm diameter, 600mm focal length - $68 - www.amazon.com/Fresnel-Acryli...
(actually you only need one, but this has the 2 foot focal length)
3. control arms/motors & sun tracker photo detectors
2 MDD10A, Cytron Dual channel 10A ESC - $46 - www.amazon.com/dp/B07CW34YRB?...
1 Pack Digital photo detectors (for sun tracker), Songhe LM393 - $8 - www.amazon.com/dp/B07XFZ99XL?...
1 10" heavy duty linear actuator - $50 - www.amazon.com/Happybuy-Actua...
1 30" light duty linear actuator - $45 - www.amazon.com/gp/product/B07...
1 six pack servo motors - $28 - www.amazon.com/gp/product/B08...
1 20kg servo motor - $16 - www.amazon.com/ANNIMOS-Digita...
2 motor wheels - $30 - www.amazon.com/Bringsmart-JGY...
Note - I recommend better wheels/motors. This is just what I used because I had it. A chain drive wheel would be best. See my mower build video and parts list for details
4. Frame, wheels, power, wiring
8 2x2x8 - $25
2 bike wheels (pick up from recycling or from old bikes) ~$40
2 mower wheels (pick up from recycling or old mower) ~$10
1 3s lipo battery - $35 - www.amazon.com/Zeee-5200mAh-5...
1 pack DC buck converters - $11 - www.amazon.com/gp/product/B01...
(optional) 1 shore power supply, 12V 5V - $21- www.amazon.com/gp/product/B07...
1 pack wires for wiring connections - $7 - www.amazon.com/gp/product/B01... - แนวปฏิบัติและการใช้ชีวิต
![[TH] 2024 PMSL SEA W3D4 | Summer | ฮึดเข้าไว้ ถ้าอยากไปรอบไฟนอล](http://i.ytimg.com/vi/W0eoG-wrVVU/mqdefault.jpg)
The depth of thought that went into this build is just impressive
There is an industrial agricultural robotics company that builds a giant version of this that uses lasers instead
I just want to say that this is *really* inspirational. The construction -- cardboard! random sticks! -- looks like something I might make. I can't tell you how amazing that is: to see something *work* that looks like something I might actually be able to accomplish. I keep telling people that my idea of a great build is when I use the *clear* packing tape; this project is within my capabilities, and that is just so inspiring. Thank you.
Wow, thank you! My construction methods always veer towards the crude when I'm getting started. I'll polish it up someday perhaps... but until then just let it run!
I find it refreshing to see a creator that for once does not use a 3D-printer for every mounting bracket or connection.
(Balsa)wood, plastics, clay or metals can actually be shaped with simple hand tools faster than your typical [3D.software-Slicer-plastic selection-#D-printing...] and the accuracy is usually "good-enough" to get the concept proven.
The reason I came here is that I had thought about building something along the same principle of a Fresnel lens on a moving platform controlled by a RasPi. It is great to this approach which I can (and will) use to improve my own design. Thanks!
This is so awesome, just the sort of thing market gardens and small scale organic farmers could desperately use, if there was a commercial version of this I’d 100% buy it
Oh man, I've got a long way to go to commercialize :) I'd like the next version to be smaller, faster and use more 3d printing so others could print and use, not so much setup/debug required.
@@nathanbuildsdiy I’m very interested seriously!
Could use polarizers to turn the lens on and off instead of the flap. A couple of salvaged polarizing sheets at 90 degree angles to each other would block all of the light coming through, then rotate one to let light through. Or you could use an LCD for the same thing. That might be a little more foolproof than lifting a flap that can suddenly catch wind.
Good call. Any idea where to find polarizing sheets for cheap? Or should I turn to online suppliers?
The LCD is really cool, but I'm thinking the next revision of this is going to be much smaller and less expensive, so I hesitate on LCD
Polarizer sheets are like 1USD.
So awesome. Fresnel lens, polarized sheets and delta guide wire setup might be great for my little garden. For extra credit it could focus just enough extra light to help good plants when cloudy. For extra extra credit, fricken lasers to kill specific pests
AFAIK, there is a polarizer sheet in any LCD screen.
Getting one from a trashed TV/monitor/laptop should be doable.
What about a camera shutter mechanism? Here's a guy taking one apart. It's actually simpler than you might think. th-cam.com/video/CqKqu8lsCa0/w-d-xo.html
This is a great solution. The University Hohenheim testing/building weed control robots that detroy weeds mechanical or by applying hot water or foam. You might find some ideas for improvement in their published work.
You don't need to burn the weeds completely. The humidity in the plants will be transferred, even down to the roots. I really like your solution. The burned material is even beneficial for the other plants. I could imagine that heating water with sun energy and spray the weed with this hot water is simpler to build.
To avoid problems when wind hits the opened lens cover, it might be better to move the cover sideways?
Thanks for the pointers to University Hohenheim. I'll check it out. I was thinking about what you mentioned as well - the burned plant matter will be good for the soil as well.
I've started planning v2 of the weeder, I agree moving the lens cover to the side would be better. I have to figure out how.
@@nathanbuildsdiy ⚠️ God has said in the Quran:
🔵 { O mankind, worship your Lord, who created you and those before you, that you may become righteous - ( 2:21 )
🔴 [He] who made for you the earth a bed [spread out] and the sky a ceiling and sent down from the sky, rain and brought forth thereby fruits as provision for you. So do not attribute to Allah equals while you know [that there is nothing similar to Him]. ( 2:22 )
🔵 And if you are in doubt about what We have sent down upon Our Servant [Muhammad], then produce a surah the like thereof and call upon your witnesses other than Allah, if you should be truthful. ( 2:23 )
🔴 But if you do not - and you will never be able to - then fear the Fire, whose fuel is men and stones, prepared for the disbelievers.( 2:24 )
🔵 And give good tidings to those who believe and do righteous deeds that they will have gardens [in Paradise] beneath which rivers flow. Whenever they are provided with a provision of fruit therefrom, they will say, "This is what we were provided with before." And it is given to them in likeness. And they will have therein purified spouses, and they will abide therein eternally. ( 2:25 )
⚠️ Quran
In a "production series" you can make something like the shutter of a camera (or the stargate :)
@@1islam1 Islam is a false religion and soon will be destroyed
Wow! Thank you for this work, the energy you put into showing us, and the commitment you made to building the complete thing! Inspiring in so many ways! Thank you!!!
Wow, you're a genius! LOTS of really smart skills coming together for this build. I'm subscribing!
It is fascinating how such a complex and wonderful solution was built using so crude materials. Well done Nathan, I am genuinely impressed with your machine. I am thinking on building something similar but for de-weeding a big lawn of mine.
Bringing all those random bits of tech together for something useful is amazing.
It's great, I've been dreaming about it since childhood, and you made it happen, thank you
I like it. Brillant project and reminds me of my childhood when I used a lens to play around with the sun and burn all kind of things!
Quite brillant in the concept, I am very tempted to try and build something like that.
Super innovative. Thanks for sharing!
You are killing it as an educational TH-camr. 👍
Thanks man! It's what I use youtube for myself, so it's time I gave something back.
Strong work! Very impressive!
One of the best, if not the best project I have ever seen.
I can very well imagine that this idea can be further refined.
The community and comments clearly show that.
Keep up the good work, great project - great implementation, super explained and absolutely inspires to rebuild.
Best regards
it can be refined, my idea is to burn all the weed personally
Thank you! I agree it could take off, the next revision will be lighter and less expensive. I'm collecting ideas here for the next round :)
@@atomictraveller we got a different definition of weed i thinkt^^ - certain weed shouldn't be destroyed by a machine, I agree :-P
@@Lashandan what can i say, i see "weed burner robot," i see pristine internet posting fields, i see 200+ totally dry comments. a tucson boy has gots certain things they gots to do.
plus there's that whole "i need ideas" thing just glaring.
Very cool. I love the ingenuity! It would be great if u added chapters to the video so ppl can go back and forth with ease
I have two hands and couple of options. A small kitchen knife is one. By the time your project finds and eliminates grass I am already chilling on the terrace admiring my capable two hands - and the feeling of achievement.
Two things:
1. This is really awesome, great work!
2. "... save time weeding..." 🤣🤣🤣
Man you nailed it.... Impressive👍👍👍
I am very impressed by your method of finding and killing a weed with a machine that has the repeatability of a felt hat. I spent nine years 30 years ago working on using computer vision on a robot sprayer to identify and trigger a sprayer to kill weeds or apply fertilizer to wheat. The academics needed to allow user input on weed identity to kill weeds well. The user is the only one that can define the weeds in their garden. The user defined the fertilizer setting for wheat that part worked fine.
Making the solar weeder smaller and lighter and mounting two or three of them cn one chassis scales your robot.
There isn't one size frame that works all season long. A low frame that covers two rows may work best early in the season. Later a tall boy that straddles a single row like the one you use now. Finally, a small robot working between rows using other methods to kill weeds.
Thanks for the encouragement! That's good advice about different size robots for different times of the year. I see a path to a much less expensive, small weeder in the next revision for early season or between rows. Then I'll probably come back to this one. Adding multiple arms/lenses would be pretty neat after that.
I'm so frekin excited for his next video!
'How to clean your pesky gutters using simple thermal nuclear fusion'
This is an excellent low-power approach proof of concept! Very efficient. Thinking ahead I can see lasers being used also when the light conditions are not ideal or to zap insects. Yes, this could be used against pests with lasers so as not to damage the plants. Just a thought
Totally possible, but I think getting lasers with enough power would really increase the cost, right?
@@nathanbuildsdiy Yeah, and energy usage too- more battery and perhaps a solar panel etc. Components really add up quick. Device gets heavy. Still it would be nice to zap the cucumber beetles. Also I think it would be cool to have a doorway entry sentinel to laser flies and mosquitoes coming in behind or from an open window even. Imagine training that to work! I find this DIY approach to the problem very refreshing
You crack me up, I love this.
Why not hang the lens off of 3 wires? This way you can move, tilt, raise/lower the lens however you want just by three spools. :)
The wires can double as signal/power/ground for the motor in the center, which removes the lid.
Then you have something above it covering the sun
I love the whole thing!
A great prototype! 🤗
What a cool idea!!
Very impressive!
TH-cam algorithm is amazing, I was looking up Fresnel lenses recently with the idea of using them for killing weeds. But your AI system is on another level!
Edit: I love the Predator style targeting system 😂
Cool! Inspiring!
I'm happy to see you build this. I have a donkey car and a stack of wallet - sized fresnel lenses waiting for me to put them together for 5 years!
I love the 4 points of light trick! Turns off the cleansing light and provides a perfect focusing tool. You have a big enough frame to support a water dripper to reduce fire risk.
th-cam.com/video/wL9PcGu_xrA/w-d-xo.html to isolate the sun tracker
Yeah, the lid and targets turned out to be crucial so as not to kill everything
Amazing!
Impressive work! New subscriber here
Thanks and welcome!
I wonder if two lenses would work, such that one focuses on the other.... you could rotate the top lens to 'disengage' the second frying lens?
This is brilliant. You can make this even solar driven, it works in the sun only anyways. Maybe it could be an improvement do direct the focal point of the lens into a light conductor fiber? This way you can let the lens be pointed always towards the sun, and only move the fiber where the light comes out of it over the weed. So no need anymore to get all three moved in one straight line: sun - lens (at 90°) - weed.
That's an interesting idea! Is there any heating in the fiber or can it transmit a lot of light/power easily?
@@nathanbuildsdiy I don't know. That was just an Idea that came to my mind. But the fiber should just pass the energy through with almost no resistance. So I think there should be no, or almost no heat inside the fiber.
This is awesome as long as it doesn't start a wildfire. You should make an autonomous mower /s 😂
LOL, on it boss.
But yes - the fire risk on this one is there. I actually stopped to give a warning about starting fires at the start of the build section.
@@nathanbuildsdiymaybe combining it with plant irrigation so after the weed is burned the area is wet and fires or smolders are stopped before they can spread
WOW, Impressive! Anybody who says this is too complicated for a simple job does not understand gardening, ag, energy... Likewise the fire danger is over rated, farmers will deal with it, flame weeding is an established method. Hence the others who have said the weed does not need to smoke to be killed, check out the "wilting point", the ultimate model will be much faster and know exactly how little heat it takes to kill weeds of various types at different sizes. A huge advantage of this that's not obvious is the lack of soil disturbance bringing up weed seeds, that is huge to organic producers. Funny that people say you should just use a laser, like saying why should you ride a bike? you can just take a jet. Maybe a large solar array could power a laser or focused microwave and do the same thing, but the cost, size and speed would suffer huge innefficiencies. Fiber optic is fascinating, excellent idea, but I like the lenses also. An idea I've had since 2006 was another variation of this with an array of stationary parabolic reflectors aiming their points in turn at nearby weed species, that was in a rangeland situation, your small mobile device is much better suited to weeding garden crops. The point being that reflectors, parabolic or otherwise, could be aimed and coordinated to accomplish the same thing as a concentrating lens.
what an awesome idea!!! you could use 3d print to accelerate the iteration of development :)
what about fire danger though?
Yeah, that's a problem for sure and I haven't figured it out yet other than "don't operate this if there's any chance of starting a fire that could spread" I only run it on the garden where there's lots of dirt and a few plants spread out a bit.
Underrated bro ❤
Hopefully it'll get some views. I think someone with motivation could turn this into a real product.
@@nathanbuildsdiy yeah just polish it a bit more and make a kick starter
For areas with a lot of shade, where the creeping charlie is, you would need something like a periscope to allow capturing the rays in sunny area, and using them in shade. But finding one that wide could be hard.
Some have suggested fiber optics, put the light in the cable, then point the cable at the weed. It's worth a try
Oh this looks like a fool-proof contraption that absolutely can't go disasterously wrong in many creative ways including starting accidental (or possibly premeditaded) fires in or around people's homes.
InconTFSeivable! 😆
This is incredible. With a little solar power you could keep charging up the batteries to run a wheeled Rover with this and other attachments. Have you seen the Amiga made by Farm-NG? I'm working on putting together a wheel base that could connect to different attachments for vegetation management and site monitoring of large scale solar arrays.
I've thought about the solar charger as well, it'd provide some protection from sun/rain. I think in revision 2 I'll do that.
I hadn't seen farm-ng before, neat ide and products! Let me know how your project goes.
@@nathanbuildsdiy you could use a smaller solar cell and use the lens to focus the sun onto it. i dont know of the power economics work here, but the idea is entertaining.
ha! That's a good idea. I wonder if I could put the solar cell under the weed burner fresnel lens (since I already need it for the robot), then move the cell out of the way when I want to burn weeds. Cool idea!
Great great work.
Will the lense work in 20 Degree steep land.
I want to get one of this in Australia.
How could I contact you directly?
I bet it would work, but you'll want better drive system/tires to handle it without tipping!
I'm at doublegdevelopments@gmail.com, and i'll delete this in a day or 2 to avoid spam
I would like this to be much larger, but we get a lot of sun in the summer and spot spraying could be replaced with this during those times and it wouldn't have to stop when it got too hot. Paraquat can not be applied legally past 85 degrees and it hits 105 here some days in summer.
0:51 The s in Fresnel is silent. That you built this amazing robot is testament to your intelligence, talent, hard work and dedication, but for half a second there, you managed to sound like a rube.
Yeah, I had only read about it, not heard it pronounced. Ah well, live and learn.
It's possible for permanent use, a version of this device could be operated from something like a linear clothesline, especially over uneven or muddy terrain. But maybe not in wind.
I think instead of trying to focus the lens above the weed every time, maybe you could track the sun with the lens, and use front surface mirrors and a focusing lens to focus the beam
You may concentrate the rays into fiber optic cable so You will have much more freedom in terms of aiming at the target while guiding the optical cable instead of the whole lens itself.
Good idea. I'll give it a shot for future revisions
been loving the lawn and garden robots man! wish i had the expertise you do, it would be cool to make a steam weeder or something haha
Keep at it (and keep watching), you'll get there.
Can you help me the design part of this invention. I want build one and use in the field where we grow ground nut.
I'm actually working on a version 2 of this that will be much easier to assemble and not have all the customization steps for each build. I'll post it in a month or two, check back and it'll be much simpler to build. (I hope!)
i think this has been done before, but the general idea was that if you go as far as having a computer to identify weeds it would be easier and faster to just have an arm remove them with surgical precision instead of creating another potential fire hazard burning
I just saw an article about a startup doing that. There's probably going to be a lot of entrants into this area soon
HI 🤠 " Divide images and sort script" not available to be seen or download, it says that file is in the trash. 🙁
Sorry about that! I think I've got it fixed now.
Fresnel is pronounced like frenell.... but this is genius!!
I would like to say about my correction of your pronunciation is it indicates that you are very smart and knowledgeable. It means that you read the word and are trying to pronounce it the way your read it. READING IS POWERFUL!
Good to know! I had only read about the lenses, never heard it pronounced. Thanks man!
After rewatching this video a month later, I think the safer route would have been to make it an autonomous weed sprayer instead with a cheap more eco-friendly approach using vinegar, dish soap, and salt. Have it drop a predetermined amount of the mixture directly on the plant. It may operate a bit faster and hit more weeds but it's going to remove the risk of accidentally starting a wildfire. 😅
Just curious, Why not use a laser? Seems like a laser might give better control and maybe easier to build? Plus you wouldn't have to rely on perfect conditions. Id love to see a diy on a laser based weeder.
A laser would be simpler, but one that can burn weeds would have to be pretty high power. I haven't found one that's reasonably priced and high enough power, do you have some leads?
There's a company named carbon robotics which builds a professional weeding machine that also uses AI but lasers. Creatively they call it the Laserweeder.
Yup, same concept. But that one costs quite a bit more :)
2:40 clever idea
YOU should be the right person to develop a autonomous lasergun to fight locust plague in Australia.
First, I think this idea is really cool, but is this as efficient as drilling out the weed? You are moving a lot of servos.
There is a startup that's pulling weeds directly I think. I think the act of pulling would use a fair bit of power as well though, so it's a toss up. Worth benchmarking though!
OMG! this mans a genius! what an amazing machine! machines that destroy weeds! huh!
whats next neighbor, a mouse gallows??? kidd'n...probably wouldn't hurt though?????
I had one of those as a kid, but I only had my meat computer to move the lens.
💖💖💖💖
MacGyver for vegans. Nice. But, could a laser be used to selectively turn on when required?
Certainly. But it'd use more power and cost more. I think Carbon Robotics is doing exactly that, but it's a pricey product they're making.
look at those cute little hands at 1:53
HA! I just saw it, the shadow of my hands inside the focused projection of light in the lens shadow. That's pretty funny.
Which lighthouse is now missing a lense…?
Isn't that just a nightmare of a fire hazard?
A firestarter machine that goes around unsupervised in a field of vegetables dried up by the sun... What could possibly go wrong? 😂
Nice realization, tho.
This is the most over-engineered week killer I've ever seen, how can I order 2 of them?
LOL, home build is the only way :) I'm not much for building a business, just building the robots ;)
So it only works in the daytime? Seems kinda inefficient.
Man... i am doing the exact same thing as this... i started 2 years agò
JFYI: The 's' is silent in "Fresnel"
didn't watch the vid, but I don't want Skynet in charge of this one ;p
You don't have to roast the weed to a crisp. Heating it up over 150F will do the job fine.
For example, you could have a lump of aluminum heated up somewhere safe and press just this onto the weed. Fire hazard eliminated🤗
That's a neat idea. Could still use the sun to heat the metal, then use the metal to fry the weed
@@nathanbuildsdiy The idea with the heated plate sound nice, but adds a lot of complexity. Try to stick with the first part, and just don't focus the sunbeam as much, so the weed only get's cooked to death.
Freaking genious
Spread the word, I'm hoping this idea goes mainstream. I've not seen it before.
Needs a 2W laser to be operational at night and an update to identify slugs...
Every time I look at the code of DIY people, who can actually build things, I start to get convulsions. It works, but is so wrong. But if I would try building such a thing (anything in the physical world), I would probably die from a splinter within 20 seconds of building it. I am jealous.
LOL. Yeah, I studied EE and I know my code is pretty crude.
How about the next version I'll 3d print the parts then there's no splinters so you don't have to worry when building it?
👋
The s in Fresnel is silent. "Frenel lens."
I would just add a small solar panel on top. And if safety wasn't a problem I would use burning lasers...
I'm planning to add a solar panel in the next revision, but high power lasers are expensive, aren't they? I'm trying to keep costs down
Really cool idea but I wouldn't try to build one.
damn, dude - did you get any job offers from this video?
Dear Nathan, could you please provide your contact info?
I'd rather have a yard of dandylions datura, mothers wart, and Mullien, than useless grass.. Can the machine be converted to Destroy grass and leave the very valuable weeds is the real question. You can eat dandylions every part of them..
Yup, just change the image classification model - whatever you consider a "weed" gets zapped and the rest gets left alone.
I hate seeing all these cool gadgets i can't make cause i live in America with absolutely zero chance of buying a pi.
Pretty pricey... I suspect there's a way to reduce the number of moving parts.
Very much so - I'm planning that for the next version
Could potentially cause forest fires.
The "s" is silent in the word fresnel - thank French for the funny word - another one.
This is a great diy weed burning machine. You state though that you have never seen one before. Laser versions have been coming out already. Here’s one: th-cam.com/video/_2s-0wgQWXM/w-d-xo.html
Right, I'm saying a fresnel lens version is new. Carbon robotics burns weeds high power lasers, and I mention that a laser version would be very expensive and high power. I think they're talking $1M for their system. Sorry if I wasn't clear - I've not seen an AI weed burner with a lens before
3D printed parts are not a good idea in the blazing sun. Heat and UV will destroy it. I would keep the size, but use aluminum.
I didn't realize that. Man... I like the idea of rapidly reprinting prototypes. Maybe I get something working then make it out of metal.
@@nathanbuildsdiy I'm curious. You got some very exciting projects going. :) Greetings from Romania
could bugs be fried? lol.... like burning ants, except think grubs, and other crap that would eat on your plants... brainstorming here.
I just tested on a poor pillbug. He was toasted in a few seconds. I think it'd work! Though if it was on the plant it'd probably fry the plant too
@@nathanbuildsdiy so then, if you detect a bug, then check for it to detect a plant, if it's a weed, burn both, if it's a food plant, then spray for bugs?
@@brightsourcemedia yeah, that'd work! Next version of it I think. I've got plans to make it way lighter and somewhat cheaper in the next revision.
Or… just use a magnifying glass!?
Liability will the the problem with this. No matter how safe you make it, if there is ever a fire in this area you're going to get blamed.
We use gangs of propane weed burners on some farm implements to kill weeds. A magnafying glass is not likely to be a problem.
@@gordoncouger9648 'We use gangs of propane weed burners on some farm implements to kill weeds.'
That is turned earth, agricultural use, and generally has a good set back from other properties.
That is nothing like this video including 'lawns, gardens'. If you do this in a lawn around neighbors' homes it is an extremely dumb idea.
'We use gangs of propane weed burners on some farm implements to kill weeds.'
You'll be amazed how much you will still get blamed if you cause an uncontrolled fire.
And you're not spot killing weeds in the middle of other grass like this is talking about doing. It's not all about what is OK just on your farm.
Not "FREZ-nell" but "fruh-NELL".
He was French.
Yeah, I learned that after posting, oops. >
Cool til you burn down da place
Forest fire start in 3.2.1
This won’t do shit to Canadian Thistle. It will only kill the top layer. Not the roots.
Yeah, it needs to operate on baby weeds. Once they're big this won't work.
Skynet v1.0
Why burn cannabis plants ?
There are only few countries left that cannabis is prohibited