I'm a programmer with virtually no experience in robotics, that said,
I had the idea of building an autonomous robot snowblower for my dad. He's
got a flat roof, too much snow, and too many years behind him for this.
Is there a strong enough robot tank like platform that i can mount a small
snowblower on? It'd need to run on gas and be fairly sturdy. I can handle
all the microcontroller and sensor stuff, I just don't know where to begin
with a platform. Is this even possible?
After thinking about this some more, i think a snowblower with an 8 bit
mind might be a bad idea as far as fingers are concerned. What about a
robot snowplow, that makes it much easier to design. Is there such a strong
platform out there that I can buy and attach an avr and sensors to?
sam wrote in
A challenge, but not impossible. I would suggest starting from the
snowblower and working back, rather than the other way around. Does it
really need to run on gas?
I'm not really concerned
about pollution (after all, the electricity has to come from somewhere), but
an electric snowblower would be much lighter and easier to control with your
robot. The only thing you have to worry about is cable management. And since
you'll need to know EXACTLY where the robot is and in what direction it is
facing, avoiding tangles should be comparatively a cinch.
I don't think you'll need treads; a set of knobby tires should do fine,
since the snowblower will eliminate many of the traction problems associated
with plowing. Basically make a square and put a wheel on each corner, run a
chain on each side of the robot to two wheels and a treadmill or wheelchair
motor, fasten the snowblower on the front, and then you get to work on
sensors and position control.
That will be the hardest part (as always). Especially in this case, because
you don't need to have 50-pound robot falling onto someone. Not to mention
that this robot contains a spinning snowblower blade, rotating gears and
wheels and chains, and either a gasoline engine or live AC cord.
The time and money will probably be thirty times more expensive than hiring
a local guy to come and shovel off the snow. Even the rich lady down the
road only paid in the $100 range for someone to shovel off all the snow
after a really heavy storm.
I like that idea alot, the toro is only 12 pounds and the blade is hard
plastic. I don't think it could do as much damage as a metal blower.
However, I don't like the idea of attaching a live mains line to a robot
on the wet roof. So that makes power a major problem, as the thrower
engine runs on 6.6amp. I think that means 800w for the engine, plus the
20% loss from converting battery power to AC, and 200w(?) for the robot
itself? Snow is 15lbs per cubic foot, on a 3000sqf root, 1 foot of snow
on the roof is 45000 pounds. The power shovel says it can do
200lbs/minute, so it would need to run for at least 4 hours to clear it
all off. With lead batteries, I think i'd need about 450 pounds worth.
Did I do the math right?
I could mount something like a honda eu2000 generator instead, that
would weight about 50 pounds and do the whole roof on one tank. Is that
a realistic idea, or is there something better?
Does anyone sell a ready made platform like this? If not, I know some
pretty good mechanics that would love to build a mini tank :).
Yeah, this is the part I need to be most careful with. I might go so far
as to attach flame sensors and RF, incase it catches on fire. I don't
need the robot burning the house down either.
My father isn't the type to hire somebody to his work for him, so he'd
stay up there until he was 80. As long as it's less than the 15k to
replace the roof, it's a good solution.
thanks for your help!
I do this all the time with my gas powered leaf blower. I use it on my
deck. However, it only works well with REALLY dry snow.
It's funny, I had this same thought today. The Sears repair guy came to
tune up my 11hp snow blower (thank god for service contracts) and while he
had the panels off I was inspecting the drive mechanism for bot potential.
I noticed it had a friction plate type of speed control (very simple and
cheap). This would be a decent way to control the speed on the main drive
mechanism for a gas motor driven skid steering bot. The main issue with my
snow blower is that is has a single drive shaft and clutch for both main
tires. If you had two identical snow blowers and some heavy duty metal
working tools I think you could put the drive mechanism together pretty
quick. This would be fun albeit very scary project.
it's been done and documented on the web .. using a tracked snowblower
it was done by some univ types ... you should be able to find their
for a roof I'd think that bolting guide pipes to the short sides of
the roof ... then 1 moveable long pipe that rolls up and down the side
electric snowblower set so it does not remove the roof material
run forward .... then reverse the direction in the same track
move the snowblower over 3/4 of the chute width
rinse lather and repeat
motor controls would be to move the blower guide pipe
and to move the blower itself back and forth
got talc ?
Drysuit talc & bags
Yeah, I found 2, one by MIT and one by some guy in the midwest. The MIT
one is pretty fierce!
I didn't even think of something like that. Thats interesting. My plan
was to program in the blowing pattern, and give it a compass and edge
detectors. Since there is a layer of metal on the roof, I was thinking
of using simple metal detectors to check when it hit an edge. If that
didn't work, my backup plan was microwave sensors, since they seem to go
through snow well.