DIY robot lawn mower

[Tommy]

It looks like you've taken on a huge cutting area behind many houses -- are all those people getting some free mowing?

That reminds me of last winter, I asked everyone on my block if it'd be ok for me to use George
to plow their snow, not one said no. but one time we got hit hard(5ft drifts) and I slept throw
it, while I was working on a recharge of George one of them brought over a snow blower and
got us(me and george) out of trouble. what comes around goes around.

As it turns out mowing that large field is much easier then either mine or my mom's.
because we are both covered by trees, storage sheds, bird feeders, etc.. It seems
possible to do that field by GPS alone with the 38" cut path.

one thing I find interesting, it's possible to view the speed of george on the software map,
in the first picture above I plotted the GPS data as lines, then the last three as circles(the size of cut path)
as george went faster the circles got farther apart per second, so, going slow gives a solid color,
going fast gives line of circles.

This is where limited Mechanical aptitude can slow a project down, so far I'v been working
on developing a Yaw Senors for over six months with very, very limited success.


Tommy

 

[Two-Stroke]

I can relate to the idea that George would be OK with GPS only in an open field but not OK where there's a lot of tree cover, etc.

Keep up the good work, Tommy.

 

[Tommy]

Video of george mowing in a confined area.

Tommy

 

[Two-Stroke]

Video of george mowing in a confined area.

Tommy

I was impressed by how well George articulated around the obstacle (it looked like a phone box).

Does he go faster when the area is more open?

Thanks for posting, Tommy.

 

[Tommy]

Does he go faster when the area is more open?

Two-Stroke, this test was using what I call CNC Mode, which mimics a CNC machine used
in the manufacturing industries that uses what is called G Code programming, in it's simplest
terms a G Code program is a text file that contains step by step instructions to accomplish a
task.

In george's case a line in the file may look like "GC3 180.040.0"
where the first character defines the type of move(G=george moves forward, X=moves Backward).
the second character defines the sensor to use(C=compass, Q=Yaw, S=GPS, G=Gyro ect..)
the third character defines the Max speed george is to move(1 to 6, 1 is very slow, 6 is his top speed)
the actual speed is a function of how on course george is (slows down when off course).
the next 6 spaces sets the direction that george is to move, the last 8 spaces of the line tells
george how far to go.

so the line "GC3 180.040.0" tells george to go forward, using the compass, at half speed
traveling at 180.0deg.(south) for 40.0ft. once the move is completed it goes to the next line.

All moves done in the video was at speed 3(half speed), but with that said george will never
win any races.

Tommy

 

[Two-Stroke]

Two-Stroke, this test was using what I call CNC Mode, which mimics a CNC machine used
in the manufacturing industries that uses what is called G Code programming, in it's simplest
terms a G Code program is a text file that contains step by step instructions to accomplish a
task.

In george's case a line in the file may look like "GC3 180.040.0"
where the first character defines the type of move(G=george moves forward, X=moves Backward).
the second character defines the sensor to use(C=compass, Q=Yaw, S=GPS, G=Gyro ect..)
the third character defines the Max speed george is to move(1 to 6, 1 is very slow, 6 is his top speed)
the actual speed is a function of how on course george is (slows down when off course).
the next 6 spaces sets the direction that george is to move, the last 8 spaces of the line tells
george how far to go.

so the line "GC3 180.040.0" tells george to go forward, using the compass, at half speed
traveling at 180.0deg.(south) for 40.0ft. once the move is completed it goes to the next line.

All moves done in the video was at speed 3(half speed), but with that said george will never
win any races.

Tommy

Thanks, Tommy. Knowing how things work is a great interest of mine and different types of computer coding is a particular interest.

 

[Tommy]

Last mowing season I first tried using rotary style mowers(3) in there friction drive mode,
but found them to be unreliable because small stick, corncobs just about anything could
stop them causing the mowers drive wheels to stop turning. I then used two standard type
electric mowers connected together to give me a 38" cut path, but found them to be very
under powered and very noisy. I'm now trying rotary style mowers but with electric motors
to drive the blades and casters so George can still do ZRT.

Tommy

 

[Tommy]

I understand this is an old(dead horse) topic, but just want to share a short video
of snow plowing with George.

Tommy

 

[lawn mower fanatic]

Video of george mowing in a confined area

I understand this is an old(dead horse) topic, but just want to share a short video
of snow plowing with George.

That is really cool! :thumbsup::thumbsup:

 

[Grok]

... I'm now trying rotary style mowers but with electric motors
to drive the blades and casters so George can still do ZRT.
Tommy

Just stumbled across this thread and I think it is amazing. I would love to hear an update on how the electric drive reel mowers worked out.

I've been knocking around the idea of building an autonomous mowing rig myself for over a year now and it's great to see what worked for you and what didn't. Professionally I'm a computer programmer so while the software side is immensely complex, I know I can figure that part out and doesn't worry me that much. My dad is a mechanic and I've got friends that are good with circuits if I get stuck. The big problem I wrestle with is the pure engineering options and scale of the device. Here is my current thinking for anyone interested:

* Small as possible. While George can tackle a 4-acre field, I think most needs are smaller and I can always use multiple to scale up if needed.

* Small gas powered engine generating electricity. All the off the shelf generators are too large but there are really good small light 4-stroke engines available these days to build a custom small generator I think. I have a 4-stroke blower that is as light as any 2-stroke and is much quieter. Going gas saves a lot of weight since gasoline is a very dense energy source. It also means I could have very long run times. I can skip this during prototyping by using small light batteries with limited run time.

* Powered rotary reel cutters in a 3-way gang configuration. This is why I'm so interested in how this worked for you. Because I want mine to be small, I really want each reel to be 6-8" wide for a total deck size of less than 20". Nothing seems to be commercially available though. I've even considered a crossbar style cutter which can easily be scaled down but it doesn't lift and cut like a reel would.

* Mower unit would be dumb. I would use blue-tooth/wi-fi to send sensor information back to a control unit and have decisions made there. This has serious limitations but it's the fastest way to make everything work for a prototype. Once validated a plan to move more smarts down to the mower could be done.