Re: [mowbot] Questions?

david steinke (dsteink nospam at dakotacom.net)
Mon, 03 May 1999 10:16:27 -0700

Lawrence Lile wrote:
> >c. Is there a simpler way?
> >
> >David
>
> Yes, there is a simpler way, David. LOSE the precision control, and don't
> worry about trying to precisely position your mower 4" from the last cut.
> Your mower will wander all over the place even if it is trying to go
> straight. The easy answer - let it wander. Come up with a scheme
> (barriers, buried wires, whatever) that keeps your mower in the yard, and
> just mow randomly.
>
> Sometimes it will re-cut an area, sometimes not. So what? Drink lemonade
> while watching it mow. The simplest scheme is just a barrier (even a row of
> stakes will do) all the way around the yard. Sense them with bump sensors.
> You can add wheel sensors if you have any bid holes, ditches, etc., but if
> you have a simple yard, just let the thing wander around like a goat.
>
> (GOAT stands for Grass Omnivorous All Terrain Robot. That's the name of my
> project.)

I thought about random wandering, but the problem came in power
consumption and run time. I can build two ways. One, low power
consumption "nipping" or higher power consumption "mowing". Right now I
have opted for higher power consumption "mowing" because of kids, dogs,
and watering. I have kids and dogs which trample the grass. Watering
makes it even worse. So I want to use standard mowing technology of
using air lift to make the grass blades stand up vertical to be cut.
After a little experimentation I found that spinning a 8x6 model
airplane prop at 4000 rpm 1" above the grass does a good job of lifting
the grass. It almost triples my power consumption, but I have less worry
about the condition of the lawn when the mower cuts.

So as it stands right now, my mower consumes 3.7 amps of current.
Cutting a 4 inch swath each trip 20' across my lawn takes 2 minutes at
2" per second. With my lawn 50' long, it will take 150 trips for a total
of 300 minutes. At 3.7 amps of current, that comes to 18.5 amp/hrs to
cut the grass. I can probably increase my travel rate to 5" per second
which will increase my power consumption about 50% to 5.3 amps, then it
will only take 120 minutes or 10.6 amp hours. But I am not sure what the
quality of the grass cut will be. I based my 2" per second cutting speed
on the power consumption of a standard 3 1/2 hp 20" lawn mower.
Travelling at 3 feet per second and cutting 20" wide, I gave it 720 sq"
per second of cutting for 3 1/2 hp, or roughly 2600 watts of power. I am
consuming 40 watts of power. To get the same ratio I could only cut 13
sq" per second, and with a 8" cutting width, travel at 1.6" per second.
Hence my 2" per second. I may be way off with my thinking, but it is
only a guestimation.

If I were to do a random pattern, it would take roughly 2 to 3 times the
amount of time to do the cut (does anyone have exact figures for random
vs non-random?). I would have to increase my battery capacity and solar
charging capacity to compensate. If I were to build a low power "only
when the sun shines" continuous nipper then the random movement would be
fine. As it is now, I figure I will solar charge three days and then
cut, syncronizing with the sprinkler system to cut before it waters.

David Steinke