Actually, I like this idea. An FPV mower would be fun! and the parts are already out there. Still, a M3DP Mower would be ideal. I just can’t see me buying an $800 robot mower then stripping out the brains and spend another $200 + to add the RC/FPV stuff. A nice starting platform would be great. I can do the basic body in Fusion360 myself. I just am no good at making the rib re-inforcements for the model and getting the holes for the motor mounts and stuff to line up correctly. It would then have to be cut into pieces so that it could be 3D printed on the readily available printers, like an Ender 3 or a MP3DP. I also would not not know how to design it for the stresses from the motors, etc. Engineers are wizards at that sort of thing and spec-ing out the parts.
Considered design that’s 3D printed and CNC’d?
Could use something like ACM panels carved, cut and folded to make bulk of a weather proof rigid chasis (https://www.homedepot.com/p/Falken-Design-12-in-x-24-in-x-1-8-in-Thick-Aluminum-Composite-ACM-Black-Sheet-Falken-Design-ACM-BK-1-8-1224/308670291#overlay).
BLE, vision, ultrasound, lidar, so many navigation options…
Or, maybe EMT based structure with minimal 3D printed parts? Design Optimized for function, minimal print time, minimal assembly and cost?
Run on 20v DeWalt/Ryobi rechargeable batteries
An automated mower would scare me but an automated string trimmer would be pretty helpful. Easy enough to make it follow an underground wire and its a chore I hate worse than mowing. Just law down the wire like a dog fence and make a trimmer that hugs the wire and cuts to the left side of it. Let it run while I’m cutting the grass so I can watch it.
I always dreamed of mounting a string trimmer on the front of the push mower so it could trim then the mower would mulch the trimmings. Hard to pull off with a gas lawn mower and a corded electric trimmer.
According to Google, a person walks between 2.5 to 4 MPH (4-6.4 KPH). Lets just say 4 MPH for the mower. A brisk walking pace. 4 MPH = 4,224 Inches per minute. So, since circumference= 2 x Pi times the radius; a 6" wheel has a radius of 3 so the circumference is 6.2831 x 3= 18.8495". So, if we move at 4 MPH with a 6" diameter wheel, it would have to spin at 4,224 inches(4 MPH) divided by 18.8495 (circumference of our wheel) = 224 RPM. We need wheel motors that spin at 224 RPM. Probably BLDC motors since it’s a moist area and they might get wet. What would the needed torque be for the motors? The mower would probably weigh about 10-15 pounds. Lets say 15 pounds. Might have to deal with hills too. How do you figure out torque for a motor in this application?
Why would the automatic mower have to match a person’s walking pace?
The formula I just found is considering lifting a weight in newtons by a pulley on a motor. Torque needed is radius of the pulley times the weight being lifted in newtons. I’m no engineer but to me this scenario seems to fit. So, a 15 imperial pound, (66.723 newtons) mower with a 3" radius wheel would need 3 times 66.723 newtons= 200.19 newton meters of torque? Does that sound right all you engineers? I think we’d have to multiply that by some number. I think the 200 Nm value is continuous torque. It would need to be higher, I think, to get the mower moving. Is that stall-torque? I don’t know. I’m swimming in unfamiliar waters now. @Jeffeb3 , how does this sound? Where am I going wrong? Also, since we would have two motors in the rear, is this load divided by 2?
See, this right here is why mixing imperial and metric is a bad idea.
Why not use inch pounds if you are going to measure distance in inches and force/weight in lbs? 15 lbs on a 3 inch moment arm is 45 inch pounds torque. Converts to 5.08Nm
Or you can convert 15 lbs to 6.8kg and 3 inches to 76mm (or 0.076m) 1kg is 9.81N (1g gravity being 9.81m/s^2) so again I get 5.08Nm torque.
The issue come in with the 15 lbs. Using the mass as an approximation for force required to move is a fallacy. Torque is the ability of the motor to accelerate a mass, ignoring rolling resistance. My Supra weighs in at ~3800 lbs, and moves just fine on a 13" radius wheel with 520 ft-lbs torque thank you very much. It was still moving just fine on the bone stock 245 ft-lbs torque. It’s true that there are torque multipliers in the drivetrain, the gearbox and differential together mean that in first gear, I might get more torque than the mass of the car, which means getting more than 1g of acceleration, or whatever traction the tires can provide.
So 5.08Nm is the torque required to accelerate 15 lb at 1g, or lift it straight up at a constant rate.
More to the point, it is probably better to calculate the desired speed, and use torque multipliers for motor speed. Using a belt drive with a small pulley on the motor end and a larger pulley on the wheel end allows a lower torque / higher speed motor to deliver more power. So using commonly available GT2 pulleys a 20T motor pulley and 60T driven pulley would deliver a 3x torque multiplier, at a cost of 3x RPM needed. A 16T motor pulley would further increase that to 3.75x (A belt drive like that would also allow you to have your motors higher up and further from the potential wet.)
High speed motors are pretty readily available, and a system that incorporates multi wheel or axle drive can make sure that your movement limit is traction based, but since there are unknowns like the slope of the lawn, and younwill need to figure the rolling resistance of the wheels in the grass, some experimentation will be needed to know what is “enough power”.
Edit: by the way, just for reference, a Tesla model 3 has a total of 400Nm of torque, so while getting 200Nm of torque.from an electric motor is technically possible, it’s unlikely that you’d be able to use it with a 6" diameter wheel. 
a Toyota Prius hybrid’s electric motor is in the ballpark of 162Nm torque. Just saying that your calculation of 200Nm was suspicious…
Well, @SupraGuy, I appreciate your comments. That’s why I’m not an engineer. I did try to convert 15 pounds to Newtons, according to the Google conversion thingy. Then Newtons to newtons for figuring out the torque. But I wasn’t confident in my calculations which is why I spelled it all out. Yes, Belt drive would be nice. But if I can’t figure out a direct-drive I am certainly not going to get the gear ratios correct 
Maybe someone who knows a lot more than I do can come up with a solid Idea of what motors/gears would be ideal. Considering we will need two driving wheels for motion and steering. I guess just a large swivel ball-wheel in the front would be sufficient?
I wasn’t trying to be disparaging. I see what happened, looking now. You converted directly from inches to meters.
A drive setup like the Rolling Plotter build might not be bad.I think that he originally used belts and then went to a gear drive system. Even simple R/C cars use some sort of drive reduction as a torque multiplier. I don’t know if stepper motors are quite required here, but maybe some R/C motors with speed controllers could do the trick if you have something to keep track of the amount of steer that you have.
For a mower application, I think that trying to keep to 2 tracks would be better, so 2 swivel wheels on the front, or more likely, have 2 driven wheels on one axle, so no differential drive for steering, but maybe one free swiveling wheel and one steering wheel that is servo controlled? I think that most lawn mower type wheels will slide on grass a little.
I figure a walk-behind mower. Google search says a standard Tractor style Riding mower also does 4 MPH and a Zero-Turn type does about 8 MPH. If this is going to be an RC/FPV type then I just figured 4 MPH would be a good speed. Especially for an un-proven design. The blades have to keep up as well. I suppose a tractor style mower is limited for safety reasons.? I was just thinking what a walk-behind mower does but at a brisk walk. By all means, if it can be done quicker, I’m sure Tim the Tool guy would like it!!!
Well, wheel drive can be a PTO from the main motor, so that reduces the complexity of the thing. Blades for cutting grass benefit more from speed than torque.
Oscilating blades like a hair clipper could be safer and need less power, maybe, definitely less overall space.
Riding mowers are usually governed for speed, but people soup them up for racing, (removing the whirling blades of death) and getting them up to much higher speeds, like 60mph for the extreme cases, so there’s lots of room, if you really want Moar Powaa.
So, what do you think, Ryan? Does this sound like something you’d want to tackle and offer? Even an RC/FPV ready version without the GPS/RTK stuff would be appealing to a lot of people, I think. And a kit with the 3D Printed parts, motors and ESC’s for the motors is something I would definitely purchase. The receiver, I think would depend on the Transmitter so maybe not put that in the kit. Or, offer the receiver/transmitter pair as an option? I don’t know. Sounds like then you’d have to warehouse more parts. Not sure that’s your thing.
What do you think, Ryan? Sitting on your back deck with a cool beverage remotely controlling your lawn mower with FPV goggles? Hmm? Hmmm? Whatdy’ think? Hmmm? You gonna’ do it? Huh?
If you DID make one, I will buy the kit, build it and make the video for free. I mean, I’ll pay for the kit and make the video for free. And I have a drone so I can take cool drone video too. Just another excuse for me to use my drone. AND, I will edit the video. I can send you a couple of links to my YouTube channel of my current drone flying and editing skills. I won’t put them here unless you say it’s OK. Even if you just designed the base and spec’d the motors and ESC’s and sold those as a kit, you could then CNC carve a puck and Vacuum form the covers you sell that mate up with your base?
It would be fun, but That is pretty far out of my current inventory. It is in the back of my mind. I was kinda thinking along different lines, but lawn/garden robots have intrigued me for a while. Nothing along these lines from me anytime soon.
An honest reply. Thank you, Ryan.
Well, Maybe the rest of us here in the forum can knock our heads together and come up with something?
Does that sound interesting?
Guessing PTO stand for Power take-off - Wikipedia? Would you just attach stepper directly to each of the two omni-wheels, or, use stepper with integrated gearbox?
Friend has a lawn robot that’s slow and non lethal enough that kids and pet roam free while the bot relentlessly does it’s thing. Personally prefer v1 be a reliable bot, absolute speed’s less critical, for me, requirement is to do maintenance trimming of ~5000sqft contiguous area, must self charge, should weed. I don’t expect v1 to tackle a neglected lawn that’s become a jungle/field.
I burn too much time every spring, and later summer, thinking about a lawn bot that uses OpenCV to laser/microwave/flame-throw/machete weeds…
+1, automated string trimmer for a v1 seems faster to design, mod/improve, maybe easier to source parts, less dangerous/liability.
Ugh, this topic is making me want to tear down my dewalt battery powered strimmer now…
James Bruton probably designed, built, coded, and shared the plans, for something like this already, that guy churns out so many robots…
Doesn’t sound like @vicious1 is interested in doing this.
What do you all think? Shall we try to tackle this here and figure out how to do this?
If there’s enough interrest, I would think we first need to set a weight limit. From that we can spec the motors/drive system. Then the power source. I would strongly suggest batteries, but deep-discharge type Lead/Acid or LiPo? Also, the motors needed would guide the ESC’s to drive them. Also, Sensors needed. Types and sources for them. And a microcontroller. I can’t code in any language. So whoever decides to write the code gets to choose the microcontroller. I would hope it would be a newer 32 Bit ESP32 device with Bluetooth and WiFi so it can be controlled with a smart phone maybe. Using the MIT App inventor found here: https://appinventor.mit.edu that way it’s free and opensource and anyone can modify the code. Once the hardware needed is established, A 3D printed base is designed in CAD that fits all the hardware in a logical way.
Then, once we have a working platform with the parts mounted and working, we figure out how to make the cover. 3D printed or create a CNC file to cut a puck and vacuum form a cover that fits the base unit and has places for the sensors.
Anyone interested?
I think most of us either don’t have the skills or are preoccupied with other projects. Otherwise someone would have already said he/she was in.