Autonomous Electric Lawn Mower
 

Hi guys,

We are running into issues trying to source an appropriate motor for our cutting blade for our lawn mower. I was wondering if you guys had any options to suggest. System voltage is 24V. We are looking for a motor that satisfies the following requirements:

Voltage
DC: Any
AC: ~120V to ~240V (60Hz)
Wattage: ~1500W+
RPM: ~2600RPM+
Torque: Variable
Power requirement Equation:
P (Watts) = (Torque(N*m)*RPM)/9.548

We would prefer DC motors that meet the requirement but could work with the AC motor options.

Seems like our current options have been mostly 360V and 400V in very odd frequencies which require vfd's. Makes the whole setup way to complicate and expensive for a lawn mower.

I would look at motenergy motors. I use one of theirs for my battery riding mower. It's a permanent magnet dc brushed motor and works awesome. The permanent magnet motor self regulates it's speed based off the voltage so you really don't need to get fancy with controls. Mine just uses a forklift contactor. Their ME0909 is nice and small and would easily do the job of spinning a blade.

I like it. What voltage are you using on yours? What blade width? How many kW (or amps whatever works for ya)?

Currently we have a crazy expensive setup with a 205vac 1.35kW (4.15n*m 3450 rpm) motor, an allen bradley vfd, and a big inverter on our grocery list. All together I assume our current robot cost will be around $4500 with all the gps, sensors, 5 electric motors, etc...
Puts into perspective if I ever make another one for private use I will for sure just use a gasoline cutting motor. Much cheaper.

My mower runs on 48V. It takes very little power to push it around, probably 10-20 amps max. Engage the blades and you're sucking an additional 80. Start cutting and that jumps up to about 130 in thick grass. It has a 3 blade 36" deck.

Tim Fulton's Simplicity Broadmoor 728

My interest is in the autonomous strategy. Picking a motor to spin a blade should be cake compared to operating in a complex environment.

First word in the title, then it goes into the weeds. :)

Start with a drone mower?
https://www.youtube.com/results?sear...trol+mower+diy

Why not start with a commercially available electric push mower? Many of the battery ones (at least the newer ones) spin the blade slowly to save power until a load is sensed. Then you can focus on the autonomous parts like redpoint said.

I guess to add back to this post we ended up finding a solution pretty easily. We will end up using a single phase input to three phase output VFD for our 1.5kW Ac motor.

Why not use an existing unit? Donations for the most part. All of our parts have been without cost so far. And actually some of the reverse engineering was a little more complicated than just making your own simplified system. Seems like some of the oem options were made purposely more complex for people to work on for what i assume to be intellectual property protection.

We have a fantastic strategy so far. Can't share too many specifics but we will have multiple GPS inputs, collision avoidance sensors, and collision detection, and probably bumpers. lol

Full tilt we are looking at 2000W/h for the entire robot.

The existing robots all use blades that are relatively safe if they hit a person or animal. Maybe it would cause a laceration, but not bad. I've seen videos of the robots going over plastic toys, and they leave the toys intact.

Sounds like a fun project. It's something I've thought about back before they made consumer products. My current musing is how to make an asphalt shingle laying robot. I'd have the material on a spool though instead of "squares".

Will this robot mow a pattern, or random?

This sounds much more interesting than the original post. I'm glad I'm subscribed.

So the mowing a random pattern concept has been driven into the ground. The plans are even already available on maker space for one of those three-wheeled electric fence bouncing robots already.
We are going to be mowing a pattern. Accuracy of the sensors is up to 10 mm. But we will see. Setup and implementation heavily effected the accuracy from 10mm up to 1 meter accuracy.
Also, my goal is to be able to mow any pattern we input, not just the users personal yard (which I think is lame on top of the random pattern and very limited mowing/cut).
Lastly, I'm sure our robot will be as dangerous as a standard gas mower. Except when a gas mower hits a big branch and stalls, ours gets more torque strength as the mower approaches zero rpm! So keep the small chillens away. :P
Nah I'm sure we can monitor input amperage vs rpm and see what we are mowing and if it is outside of acceptable range we can stop the blade. (think smart power tools)
For the most part we are only going to the final concept stage of development and will not be selling this so we don't have to worry about sue happy snowflakes! (see: tesla door handles)

I will for sure be building me one for personal use afterward though as I will finally be an engineer. Also, it will probably be converted to gasoline and possibly even add an alternator. A pure gasoline input automower would be heavenly and cheaper. *woot cash money* *woot no more poverty*

If it's still in the conceptual stage, could you swap out the [rotary?] blades for reels? They are much friendlier to the grass, shearing instead of battering. Less moisture loss.

I'd prefer a dangerous mower. They aren't meant to be safe, and if they are safe, they aren't so effective. The videos of consumer robots shows them mowing an already mowed lawn, which isn't impressive. I want to see it take a couple inches off and hack down weeds. I'd still put a big red EMO button on it though.

Might some of the bicycle motors work with 24V? Bicycle motors are sometimes, 36V & 48V, but could be some are 24V?

I started to go that way at first, but I noticed the continuous heat capacity on them was only like 200W-250W (btu/hr). Ended up being too small. The bigger versions were closer. But these guys came through faster than that whole idea.

Blade: Standard 20" lawnmower blade from wal-mart. I think those razer blade wheels and all that nonsense are so cringe.

Yes, most of the bicycle motors are too small. Cities try to limit bicycle motor sizes to avoid HP wars & bicycles moving at 30-50+MPH. I had a Crystallite(?) 48Volt, 2000Watt motor on a big-tired used Schwinn that powered up, even steep hills.... except for one. The Crystallite came in three torque & speed levels. An old style, I think it could have been more efficient. But it would have powered through most grass easily.
There are other (more efficient?) electric bicycles with strong motors. Got to be a 24Volt bicycle motor, you can use.

Alright final-ish lawn mower specs:

Synopsis of the report: The design consists of an autonomous lawnmower with a 1500W three-phase ac cutting motor (20" blade), a single VFD controller, 4500w inverter, four-wheel-drive system with (4) 100W DC drive motors (7 height adjustable... not easily adjustable yet... :( ), a modular 3d designed chassis, and a 24V battery solution. For the sensor array, the robot deploys multiple Ultrasonic sensors for obstacle detection and avoidance, a multi-axis accelerometer/magnetometer for directional control (0.1-degree accuracy) and event detection (object strikes, anti-rollover), and a weather sensor for inclement weather detection. The robot also houses an internal GPS unit with a connection to the on-premises GPS base station for correction data to achieve the 10 mm placement accuracy since the closest public GPS station is out of range at 75 miles away. Finally, for the code, the if-then statement lines will control safety, object interaction, traction control, and GPS/Compass data-logging and comparison for route-following.

With that said, that's the plan. We had to resubmit our sensor order so I hope it gets approved again. Also, design contingent on how ambitious I still feel after next semester between the race-car and every other project going on personal and academic.

I think retail for the parts is about $5000 dollars so far. No labor cost included. GPS/magnetometer components around $800.

Now I'm looking into getting into testing and adding the project-specific sensor output values to my previous navigation code from the maze solving robot I worked on this semester. That consisted of around seven pages of code for the obstacle combinations, x tracking, y tracking, pointing direction tracking, three sensor inputs, and left and right motor forward and reverse. :O

Hopefully, there is a picture attached to this.

Lastly, to add. The next big problem to face will be a lightweight adequate and affordable battery solution. We have to pay out of pocket for it and this thing needs to be similar in weight to an existing push mower! (75 pounds?)

I would highly recommend looking here for lithium battery solutions:

Battery Hookup

You'll likely have to make your own ~24V pack out of the cells, but the prices are great.

I know I’m digging up a pretty old thread. But very curious about this. For 10 years I thought about a robo-mower every weekend when I was mowing the lawn. How did this turn out??

Segway released a new model that looks cool. I'd link it, but I just joined so I'm pretty sure I can't post links yet. Look up "navimow".

A lot of autonomous lawn mowers just wander around, while this one is able to have boundaries set through your phone and then it follows a path to be efficient. I'm not a homeowner currently so I'm not going to get one soon, but I'd get one once I get a home. Lawn mowing is no fun.

Covid happened and they kicked us off campus and told us to stop working on it and just write the report on what we had. So we weren't allowed to finish. That was back when everybody thought they were going to die from covid and also a tornado took out a lot of the college kid neighborhoods so the dean told everyone to go home before spring break. :(

I will have to ask my old advisor what they did with it. I am assuming it will be recycled into another senior engineering project.

There's modules that plug into your R/C equipment on your, say, drone that have commands to do a preprogrammed return flight back to some point. Get some big servos and a steering mechanism or you could use skid steer.