Not too many people like pushing a lawn mower around the yard, but unfortunately, it is a necessary chore. Anti-push-mower advocate [imadethis2014] decided to (as his moniker suggests) make a solution to his pushing-problem. He was a fan of radio controlled devices so it made sense to convert his mower to RC!
The mower itself is powered by an off the shelf battery and has a 21-inch deck. The stock wheels and handle were removed and replaced with a new extruded aluminum frame. Out back are a pair of used electric wheelchair motors sourced from eBay. These drive motors are mounted to the new frame via a pair of aluminum brackets that [imadethis2014] designed and cut out on his Shakepoko CNC machine. Since both rear drive wheels work independent of each other, a pair of swiveling casters up front allow the mower to turn.
A large car-sized battery box houses the two smaller wheel chair batteries as well as the motor control and RC electronics. Check out the video after the break, the mower seems to do fairly well. [imadethis2014] admits he needs some new wheels as the current ones don’t get a lot of traction on the grass. He’s also thinking of adding GPS for automated mowing but isn’t quite there yet.
a remote control controlled lawnmower eh? Nice.
I would say radio control control personally.
Extremely cool. Looks really good. I gotta ask, though. Why extruded aluminum? For the weight? Ease of construction? Wouldn’t steel or even aluminum L bars/rectanglular “tubes” from the hw store been much cheaper?
Cheaper maybe, stronger, no. Fastening them together with L brackets or bolts is not something I thought would stand the test of time. Welding is not something I know how to do, nor have the equipment for. 80/20 extrusion seemed to be my best bet, plus it doesn’t rust.
with a position system like gps, but +- 5 inch precession, this thing could be easy programmed to do the job alone. I never found something that is precise enough in the range of 5 acres.
has anyone an idea?
I’ve done a lot of thinking on autonomous mowers. My best plan is to use GPS for rough coordinate positioning, i.e. boundaries and direction, but leave fine control (making sure paths are followed next to each other) and obstacle avoidance to camera detection.
2 cameras, one at each mowing edge mounted down low to see the difference in the height of grass, and a 3rd mounted high center watching for trees, kids, pets, basically anything not grass colored.
Tachometer to monitor blade speed, and sound sensing to monitor for problems like stick or rock under deck. Fuel monitor
Communications (via gsm modem) to send pictures back to a supervisor to ask if if a questionable situation is to be avoided or mowed over. For example a dark spot could be a dirt patch or a mud-hole, light spot could be dead grass or trash.
An alternative to the vision cameras for fine control, the use of light beacons could be used to triangulate position. The beacons could be of different colors of light or flashing a specific pattern that makes each unique. Placed at the corners of the yard or in other spots where at least 3 could be seen at all times. I had originally thought radio beacons, but I think the problem of reflection would be an issue. Plus the receiving antenna would have to be highly focused.
I have thought about using an invisible fence to mark the outside edges of where the mower should go and skipping the GPS. So many projects, so little time…..
My thought was to use a camera and markers on posts. The markers could be painted with a colored barcode on a cylinder so it would be visible from 360 degrees. I think you would only need one marker in your field of view if you’re only trying to follow a straight line but three markers would allow you triangulate your position. The camera direction could be controlled by a servo so all three markers needn’t be in the field of view at once.
The markers could also be IR emitters pulsed with different patterns or frequencies.
Alternatively put a marker on the mower and position two or more cameras around the field.
I’m too bone lazy to do it myself.
I’m considering http://store.swiftnav.com/ for the GPS with Centimeter-level accuracy. Someone in the reddit thread suggested a high sampling rate, rejecting the way off values, and averaging the rest – I’m not sure how reliable that will be.
The swiftnav uses a trusted immoble gps module on a station and a movable gps. These two modules talk to each other and compare the two signals together to find the error offset in the main signal; the station is the reference at a known position, and the difference is passed to the moving module.
You can achieve this setup with two gps module and two radios(like wifi).
Called “Differential GPS” or DGPS.
https://en.wikipedia.org/wiki/Differential_GPS
My first thought was a little quadcopter flying above it with an imu & vision tracking.
$600 electric lawnmower, plus the electric drives plus the rc kit – that’s a lot of money compared to paying the neighborhood kid $20 once a week to mow it for you (and the kid will probably do a better job – that thing cuts like a drunken barber).
My electric mower cost $20 on Craigslist.
What if you have a couple acres to mow?
Probably does not need new wheels on the rear for traction. He needs to move the traction battery to the rear of the mower to get the weight off of those under sized caster wheels on the front. Replacing the front casters with more turf oriented wheels might also help.
I put the battery in the front only cause that’s where there was room without making it wider/longer. It’s actually pretty well balanced, as those motors and wheels are heavy I didn’t want to put more stuff in the back because I wanted to be able to open the door and attach the grass bag. New wheels are on order and I’ll post a video probably next weekend.
I’ve built one and I’ve tried 3 different sets of tires and wheels. He definitely needs traction. I’d say change them.
Am I really so old that I’m the first of five commenters to think “hope nobody shrinks his kids!”
I was thinking he was twenty-six years behind, but then I realized Maximum Overdrive was three years earlier then I remembered it took someones eye out….
Hope it doesn’t take someones eye out.
At least he needs to move the battery to the back over the drive wheels.
I really want to build one of these so bad but the shipping cost to Australia is almost as (if not more) expensive as the motor/gearbox combo.
next step, for when you eventually also get bored wit RC-ing the lawn, is to add a gps-assisted autopilot, i’m betting the 3DR APM is already almost plug-and-play ready for this.
https://www.youtube.com/watch?v=OhjR7ZJyvb8
i might have been too fast in posting this, i found it in a forum thread about an APM mower, but it seems he’s not yet using one in that video ..
there’s lot’s of stuff on youtube about automated mowers, but most still seem to use some kind of underground cabling.
i’m getting an itch..
Check out http://www.ardumower.de open source, still a need for a wire but lots of sensors to maybe do away with it…
Lawnmowers in general are extremely outdated and haven’t really advanced. They still can’t cut up to the edges of obstacles, they still use a large centered blade and are huge.
All you want is an updraft to pull the grass up straight and a wide-linear cutting line to chop it off. We need something like a flowbee for the lawn. It could be much, much smaller and would allow cutting right up to the edge. And using a system like an electric shaver with reciprocating scissors smaller than fingers and toes (I mean seriously – grass is tiny), would be MUCH MUCH more safe to humans and animals.
My lawnmower died last summer and I haven’t had the motivation to fix it. Hmm….
Also a sort of comb on the front and a low profile could push rocks and branches out of the way instead of picking them up and tossing them at people, cars, and house windows.
How about a set of Hedge trimmer blades?
A petiole of a leaf would take out a fine clipper, not to mention twigs. Acorns are a gob-stopper.
Any autonomous mower must figure out how to get out of dips and holes, or shut down.
I like the idea of manual control, but tank treads are necessary. We have hills to do, and that rope trick is dangerous.
Make money by renting it to neighbors with the same hills.
Then there is the option of other green ground cover. Or get the latest in “mulch”, industrial rubber waste. Cover the earth with dead waste. Easy!
Look at the ardumower project , it ‘s quite advanced atm and open source/hardware:
http://www.ardumower.de/index.php/en/
There’s already perimeter management, and I hope RTLS/UWB location soon enough :)
Did you really have to name it “Ardumower”? lol
It s not my project :-P By the way the name would probably change for the next version with stm32f4 ..
I built one of these a couple of years ago. An old electric wheelchair, old Victa mower, a Sabretooth controller run by an Orange RC kit and some welding did just fine. Most expensive was the Sabretooth. Worked a treat though.
Maybe add Roomba brains and a flat top so a cat can sit on top … the you’d get lots of Youtube views.
Always really interested to see developments in the semi-autonomous mower world, this looks good! I wonder what’s stopped there being a viable market for a commercially produced model, or whether it’s just that the jump to robotic mowers has it covered (albeit more expensively so)
Probably the reason is that giving a 21″ steel blade spinning at thousands rpm fully autonomy out there is a little scary. Vacuuming the dust seems good, but chopping stuff needs a lot of control and safety
Buy your mower nothing. Make it too !
It looks like a childs toy
https://www.youtube.com/watch?v=DFnYdSP-pBw
Good to see some genuine radical design and innovation in the lawn mower market. A trendy update of the traditional push mower.