Cheap DIY Robot Lawn Mower - IndyMower #1
HTML-код
- Опубликовано: 21 авг 2024
- The first 500 people to use my link skl.sh/nikodem... will get a 1 month free trial of Skillshare premium!
More about the project: indystry.cc/in...
Patreon: / nikodembartnik
Best 3D printers: shareasale.com...
BLDC motor: amzn.to/4bcSObA
DC motors: www.pololu.com...
Arduino UNO: amzn.to/4cyJiAD
Perimeter wire: amzn.to/4ca3Uzj
Mower blades: amzn.to/3zkOLfM
Subscriber count at the time of upload: 136 218
This video is sponsored by Skillshare
#robot #garden #arduino
The first 500 people to use my link skl.sh/nikodembartnik06241 will get a 1 month free trial of Skillshare premium!
😢q😢
Even remote controlling a mower without any automation seems so much more fun than pushing one, very cool project!
FPV lawn mower drone race
Others have said this, but I'll add to it: put the twin drive wheels at the front, so you're dragging the swivel wheel. This is called a "tadpole" for motor-trikes and pedal trikes, and it is both much more stable and means you're not jamming that single wheel into the ground and having to push past it. Putting it at the rear will drag it *over* the bumps instead of pushing it *into* the bumps.
Great Project. Robot Mowers are VERY EXPENSIVE so it makes sense to DIY. I'll be watching !
Ohey! I love making little robot designs like this that require rugged durability, and I have two big recommendations right off the bat for this based on my own experiences!
When 3D printing discs that need to spin without exploding, using the concentric fill for layers is super useful! If you use the default zigzag, your prints have a specific weakness in one horizontal direction, similar to layer separation problems. It can be particularly noticed when 3D printing gears. The proper solution to this, of course, is to waterjet some aluminum or other metal so you don't have a crucial part like that 3D printed!
Additionally, when dealing with high-speed rotation, a very quick and easy self-diagnostic method the robot can use to ensure it's not damaged or malfunctioning is an accelerometer! In the case of a sawblade (or the saw-adjacent rotary hinged blades on your grass cutter mechanism) it's quite standard to include an accelerometer near the center of rotation, on a stationary part of the robot. This way, the signal can be quickly checked to see if there's any unexpected wobble or vibration each time the device spins up to speed, ensuring there's not unexpected friction (things caught in the bearing or mechanism!) or wobble (blade missing, broken, etc!)
Hope these tips help! I love the idea of open source designs like this and I think youtube is a great community platform that allows creators to spread brilliant ideas :)
I've had a lawn mower robot for the past 7 years now (an husqvarna automower 315) and I can say it's a game changer. The garden is always tidy and we make no noise (vs my neighbors who uses their gas mowers every other weekend, and have their grass always high until next cut). What I can also say, though, is that the perimeter line is a pain. If at the time I bought my mower there was a GPS one at a similar price, I would have bought it without thinking a second. So, good work for now, if needed as a proof of concept, keep working on the perimeter-limited style mower, but think about rtk, it's on another level!
The GPS will need a beacon to match the accuracy constraints, won't it? Would the cheap Aliexpress GPS modules be able to compete? I fear it won't, but that would be really nice.
I was thinking the same thing (though I lack experience, just been researching the options). Great thing about this being open source is once it's out there, someone could add that as an option! Heck, Id bet someone could make it use both so it stays in the yard for certain but has the accuracy of where it's mowed of the rtk.
I could have said each word. Maybe have a look to OpenMower project.
Congratulations!
I once wanted to make such a robot, but I don't know enough about electronics to make one that works with the boundary wire. I saw on the internet that most people make robotic mowers with RTK GPS, which is quite expensive. I can't wait for the second part!
Powodzenia :)
After you tackle this, id love to see a diy roomba!
Windshield wipers motors, used, cost like 10€ and have ridiculus torque. Paired with 2 BTS7960B( 6-7€) makes cheap very strong drivetrain - I've used them for exactly same project. For blades I've used disposable blades for tape knife - some have hole premade and are very sharp. Be careful however, my main fear that stopped me from finishing project was my kids safety - If you could think about it in next video I would be very intrested to see your ideas
That's my main concern. All the safeguards that need to be implemented, with sensors and software, to make sure no human or animal can get hurt by a spinning blade without supervision
I never thought about using those blades... You could work up a tube, slider and spring system with weight on the back end for each blade. Centrifugal forces as it picks up speed would extend the blades and springs would retract them back into the sheath.
@@enodenmad cool idea! But in general, sheath in enviromant of gras clippings probably would tend to jam a lot, so I suggest simplify your idea and add spring off center of blade pivot - so it would rotate inwards when stationary, and open up when spun up. Blades are on pivot also to avoid shattering on smal stones or weeds, if we limit their movement only to in/out we’d loose this feature
@@Niwen2058 I can think of a few fixes. Hybrid wheel design to adjust the rigidity of the blades and allow them to flex back into the design without breaking. Or notching the drive shaft and angling the blades so that it turns with the weight of the blade assembly enough to cut heavy grass but hitting a more solid object could kick it into a neutral or reverse position and this retracts all blades instantly. A 3D printed squeegee or felt wiper could keep particles from entering the sheath... I know it's a lot of engineering, maybe even over doing it, but my line of work you prioritize safety first then fix the other issues.
Awesome project - not a big fan of the perimeter wire myself, but that's because my ground is really rocky. Keeping the weight close to the driving wheels will help with traction, as would small spikes on them. Keep up the good work
when your motors don't have enough torque you can always reduce the wheel size, just like you did later on ^^
make a huge difference
I have the mower you took pictures of. It had a problem driving on surfaces with slopes > 45%. The solution was to buy sprockets, you can find them as "landroid wheel spikes". In the more expensive version, these are a combined 3 sprockets per wheel. Not only does the robot drive up steep inclines, it also performs better when the grass is longer and doesn't leave "trampled" wheel tracks on the grass, this is especially noticeable when driving along the cable towards the charging station.
Good luck with the project.
I used to run a website and forum for this kind of robot back in the late 90s. My robot had 5 IR sensors along the front to detect the blades of grass. So sort of like a line follower, the robot could follow the cut edge, gradually turning inward until the sensors no longer detected any grass and would beep to alert me. The robot was small enough that I could pick it up by a handle on the top, then take it to another grass area and start again. The whole top was mounted so it could move a little and would operate 4 micro-switches like a joystick, so I could get an 8-way reading on obstacles it might hit. The method resulted in neatly cut grass rather than the randomness of turning at the boundary wire. Because I used grass sensors, the robot made it's own boundary by detecting the edge of the grass, or concrete flowerbed edges.
I started (but never really finished) a very similar project a few years ago. In order to avoid expensive motor drivers, I DIYed some crude H-bridges with relays. The controllability of such a driver isn't great: the motors can only run at 100% power forwards or 100% backwards, nothing in between. But because all the current passes through simple relay leads, such a setup can handle a lot of power easily. For the relatively slow, highly geared down motors, the crappy controllability wasn't really an issue.
Awesome project! I’ve wanted to automate this task (because I’m lazy) for a long time. Cool to see someone do it!
cool project! some things i would suggest (using lawn robots since 8 years now): Make the not moved wheel in the back - so motors in the front with most of the weight. also bigger wheels if possible. also 4 wheels if possible. :D
If there's a dip in the ground, an unpowered wheel at the front will be pushed into it, whereas an unpowered wheel at the back will be pulled out of it.
I look forward to seeing your progress on making this autonomous. A good start might be running the code with it communicating to a base station while you tweak the inputs, then transitioning to something that follows the guide line to build out a map. Keep up the good work!
I wish someone would design an open source weed puller robot
I was surprised when you made smaller wheels. Won’t larger wheels move better through grass? What about using the swivel wheel in the rear? So cool! Congrats on the shed!!!
A bigger wheel takes more torque to start moving, and it looks like the real problem is getting going, not dealing with uneven terrain. (He also made the wheel wider which would help with the ground pressure problem of a smaller wheel.)
I agree that a tadpole configuration would be better for the wheels, though.
then y do people put bigger and bigger wheels for mtb, it rolls over stuff easier
So undervalued channel!! Keep up the good work!!
Thanks, will do!
Sir. Sir excuse me. You lost your mustache.
😆
Good video!
nice project niko, and remember, we should fail fast and fail often when inventing, so we get to the success faster :)
Great job! Can’t wait to see the next video in the series. Thanks for sharing!
I'm here for the technical part, but I just wanted to mention: Nice video effects :D
Very cool project. Did something very similar some weeks ago (didn't do a video on it yet). I started with a defective mower from ebay (mainboard was dead) and converted it to remote control. While watching I'm printing a new blade disk. Looking forward to the next part :)
Awesome video! I've been working on a robot lawnmower for a while now. Your video gives me new insights...!!
Here's a comment for you, I subscribed 5 seconds into the video. I know a quality content creator deserving of a follow when I see one.
I appreciate it a lot! Thank you!
A serious gem in a pile of coal.
Witam szanownego. Algorytm mi zaproponował to jako 1 Twój film. Imponujący międzynarodowy kanal mordo. Szacun.
Super project! Love your approach, although mayge, just maybe give yourself enough time to think/brainstorm a little more. My first reaction is the mowing bed: it should be able to cut taller & thicker grass. Many other machines do it successfully with the same principle as yours. This might come back to haunt the project later. Either way, I love this and look forward to your next video. Good job!
Looking forward to your next video!
Do you have any plans on giving it adjustable cutting height?
Yes, I decided to skip it for now but it will be added in the future, I am not sure how to do it in a proper and minimal way but most likely it will be a simple knob on the top with 3-4 height settings to choose from.
just build a battlebot, if it fails you still have a lawnmower.
Also... Paint it bright orange, or add some reflective tape or something! 😉
Or 3D print an orange protective cover?
Something that screams "caution" at anyone who sees it.
Great Project,
.
Also the fact that you show the YT video followers the option to see your progress in this project is great.
But to be interested for others , to jump on your bandwagon as well , it needs more update , needs to be more mature .
.
We have a small lawn but still keen to to see what the final end stage project will be and against which DIY price.
I'm really interested in the next part.
does the big roller wheel in the front not push the grass down so its harder to cut? thats why most have 4 wheels so the blades can reach better. You could also experiment with just a piece of plastic wire at high rpm, that is what used for handheld mowers and is ment for higher grass :)
i'm doing a similar project currently, but I'm using RTK-GPS and ROS, so the mower will cost a bit more :D (the gps modules and raspberry pi are not cheap). I'm also considering using UWB for localization, but I have not managed to make it reliable yet. Maybe I will also make videos about it :D
i also have a jetson nano ready for computer vision, but I don't know if it will be possible to use as navigation.
Following you in hopes of seeing your project someday soon :)
What GPS-RTK are you using? Do you have your own base station or do you use NTRIP?
@@nikodembartnik I have my own base station. I use a set of two u-blox C94-M8P modules. They have amazing accuracy (within 2cm if the mower has good view of the sky), so the mower can follow the path nicely. But it is not exactly cheap...
great project, but maybe you chose too high rpm motors, for a nice mower you need around 30rpm
I think the problem is not the motors but rather that it has a single small wheel in front and that generates a lot of pressure on the ground.
In regard to having to give your first motor prototype a push to get it going... Some motors require a start up capacitor. I wonder if you had used such a capacitor if it might have operated better... Just a thought. Cheers.
Great job Nikodem! Can't wait for next part about boundry wire, I am curious how you solved it. I also did try to make mover but more heavy duty one, with wiper motors as a main powertrain, but these were overmade with too much power consumption. I have ended with motors from window window lift motors. As a blade motor I used 12V DC motor from power drill. I stopped now because I have no idea which sensors to use to sense the boundry wire.
Hi I see that you are using an aero ESC. this may not be the best choice to drive the blade as they normally get more air cooling. Check current draw of the cutter under load. may need to add a heatsink to use a brushless ESC intended for an RC car.
Świetny projekt, sam planuje się za coś podobnego zabrać. Niektóre części już zamówione ale z mikrokontrolerami i w ogóle elektroniką mam raczej mało wspólnego haha, może jako programista będę w stanie coś zdziałać. Pozdrawiam
Nice. How about bigger driving tyres?
@4:01 that is why I only print with white PLA, it's to write sharpies notes on my design!
Great video. You should put four motors on.
Hi Nikodem! Great project once again. Just want to ask if you have tried to put more weight between the wheels, for example the acid battery. My guess is if you can put more wait between two wheels, the wheels will get more friction to turn. Putting battery on top of the idle third wheel makes it hard for the robot to move.
be careful charging a warm or hot battery because it can cause permanent damage, cool video.
I don't know if somebody else mentioned before, but why do you drive the robot from the rear instead of driving from the front? I mean just the opposite direction, will be much easier to drive
I also working in same kind of project remote control rc car using Mobile phone
Use 18650 battery
+
You can get cheap type c bms + charge controller
And use it
You will get good capacity for cheap price and less weight your old motors are powerful enough for running I see the problem is lead acid battery it's heavy that's why it struggling
Great project. I’d love to see your own version of an ai powered robot like the ones from Larry’s workbench, makingmadeeasy, and RobGPT.
Very cool project! Good luck. FWIW - some years ago I bought a low-end robovacuum. As far as I can tell, it does a random walk to clean an area and the software is not smart enough to go recharge when the battery is low. Not a particularly efficient vacuum. I'm very curious to see how you handle the mowing algorithm(s). Thank you for sharing the video and information.
Sounds like you might not have a start capacitor connected to the motors and so the current draw is way to high to get the robot to start moving.
You need bigger front wheel
Have you thought about addings some bracers to the chassis? It seems to be flexing a bit in the video, especially with the gel battery
It's not flexing, it's simply not flat, cheap plywood rarely is
How about making solar with film panels connected to supercapacitors?
You probably want a big gear reduction for the wheels. It doesn’t need to go so fast and it will be much easier to modulate the speed in the ranges you can actually mow at.
I'm excited to see the approach you are going to use for autonomous navigation or robot
have the pivot wheel be the trailing wheel so it iant driving into the ground
polak jesteś dobrze po angielsku gadasz też spróbuje zbudować takom kosiarke lecz użyje jeden silnik 24v 350w do napendzenia a przednie kółko użyje do skrencania używając mocnego servo lub silnika 775 też może spróbuj
How about a robot weed picker?
Nice work man 👍
Hmm. It can be cool to add AR game to RC controlled version. I think I saw Dyson vacuum cleaners had similar idea. Maybe you can explore or do color with some one
You could get away with drill motors at this weight as a heads up. You can get them from pawn shops for like 5 bucks.
why do it manually 2 hours a week when you can automate it in 50 hours a week :P
So how do you get the files for the 3d printing? And the arduino code? Or do you plan on working on it and finalizing it more?
Great project! I've signed up on skill share but I cannot find your project once I'm there. Please give a link that I can take to your project. The url you provide just takes me to the sign up page and then the main landing page.
Thanks
Good start Nikodem! I know you're in an easly stage of the project but do you have an idea of the final cost?
Will it be able to deal with non flat gardens?
Not sure yet and I hope so, if not I will try to upgrade the motors to something better :)
I hate to be that guy, you even called me out in the video... And I love the idea of this... And I know prototyping is expensive... But as soon as you started spending hundreds on this, it got *more* expensive than the commercial ones from China... Like... if those parts break... It's the same cost to replace them as to get a whole new commercial unit... I appreciate the DIYness of it though, and the gumption to reuse parts... But this is a rabbithole I went down with 3D printing 16 years ago... It used to be cheaper to DIY, and so I did for a few years but eventually, it became cheaper to just buy a better product than to purchase the replacement parts for the things that broke, but I was resistant to change, and that cost me a lot.
Honestly, I'd be more interested in modifying an off the shelf robot mower, like the cheapest one from China, and do an OSS Firmware replacement project. Find a way to hack in an i2c GPS and Camera modules and some AI (classical AI, like A*, PID tuning, obstacle accordance, etc.) to have be able to monitor the robot in real time from my phone / browser, and to avoid setting any magnetic lines in the ground.
Fun project though. Gives me ideas. 👍
Buy a used, cheap, beat up hover board. U can hack the firmware on the controller and even use that as a motor controller. They have A LOT of torque, motors are sensored, have position feedback, Ah consumption etc. A used hoverboard is like 20-40 euros
LOL, this is 2024, we need to use an accurate GPS+ system like the new Kress machines - NO boundary wires please 😁😁😁
He built a DIY automated lawn mower and shared how he did it for free. You come along with your double chin and chubby fingers and say it should be better. Get a grip of yourself
Kress machines require a reference station at the dealer to work, if it gets turned off (if the dealer closes down), you have no mower and youve lost your investment. Husqvarna, Segway and Stiga have a better solution. Husqvarna are overpriced and unreliable though! Interestingly Stihl announced they won't go wireless for safety reasons. Also, GPS doesn't work with heavy tree cover because the manufacturers won't invest in better GPS technology so in many cases wired machines are the only option.
@@mattphillips6515 Good info Matt, thanks - didn't know that. Kress doesn't mention that in any of the online vids I've seen - saved me some money 😊
Amazing project, let me complain about microphone once more.
Lavalier mic on the beginning - great
Voiceover parts - great
Unknown microphone at around 3:00 mark - terrible, echo, completely different than rest of the video
7:40 - lots of noise, don't see your mic around the shirt
Now as for project itself, brushless motors + cheapo VESC is the way to go. I am currently in the middle of designing my own VESC hardware and you would be surprised how much more torquey BLDC motors are. Price is also not much different, cheapo vesc is around 40euro + 20 euro per motor and you will have zero issues with power.
Alternatively if you want to go even cheaper you can repurpose those cheap hoverboards, they have two great motors + battery in them and even electronics (altho repurposing electronics may not be so easy)
How about getting a couple of sheep?
Do you know about openmower? Might be worth a look.
Nice video
@nikodembartnik I was going to suggest the same. Buy cheapest one, and replace electronic. I bought cheapest yard force mower and I'm going to modernize that. I haven't disassemble it yet, but it reports as esp chip to router, so it may be even easier.
Update: yardforce 280r and 400i are stm32f103 based, it's almost the same as bluepill that can be used as an arduino :)
now i wish i had a lawn just to build this robot lol
😅 you can try building floor cleaning robot
Może warto by zamienić kierunki ruchu? Tzn duże koła to przód a nie tył?
Myślałem o tym ale patrząc na inne komercyjne roboty tego typu to każdy jest zbudowany z dużymi kołami z tyłu. Dlaczego? Nie jestem pewny, być może robot o takiej konstrukcji lepiej kosi kąty i wąskie miejsca, to będzie coś do sprawdzenia podczas testów na całym ogrodzie.
08:56 - I think the weight of the robot is too low, which is hindering the movement of the robot.
Go with gps and lidar instead of perimeter wire
How many people saying it is lazy are hand washing their clothes instead of using a washing machine?
He just understands how mentally unhappy people are om youtube.
You need 2 front wheels
Is there any chance you're looking for design/prototype help? Little bits, pieces, connectors, etc.. I've been looking to start the same project, but I'd rather colab if it's possible.
Brother when upload my program and start and going suddenly stop and there is showing grbl 1.1h ($ for help) give me solution. I have no idea❤
Battle Bot next video?
maybe :)
cool, I was building a Rep-al mower, but feature creap at the time made me not finish it :D. in the end I backed a mamotion luba mower on kickstarter. Mine was using sabertooth controller with ardurover, and GPS / RTK, as perimeter wire drives me nuts, we have a husqvarna, and the amount of times it has just nicked the wire. than later it corrodes and trying to find it is nuts. if you want to go gps and ardupilot way, check out "Yuri's Homebrew & DIY" here on youtube.
9:30 skyblock
What about GPS?
great! but i think it go too fast and your motors are too small
Dont bother with the perimeter wire and look up ROS2 and NAV2 :)
what about sticking some blades on a roomba
No one older than thirty years should have a robot mower. They need the exercise of a hand-pushed one.
❤👍
Child ride on car motors are dirt cheap, 200 dolla is a waste
TOO MUCH WEIGHT ON THE FRONT WHEEL.
lmfao you mow your lawn every week?!?!?! you have no life or what?
Agree, is a waste of time and money. Many people waste a lot of time and money on nothing but appearances. But I like this video and channel very much. Lets make sure to give him respect and encouragement so he keeps sharing awesome videos of awesome ideas with us.
In past times maker youtubers was more skilled and could amaze me, this generation sucks