[chbot] Fwd: RE: [HBRobotics] Micromouse Fans check this out

Andrew Errington chchrobotics@lists.linuxnut.co.nz
Sat, 03 Feb 2007 11:32:44 +1300

Here's a good account from a Micromouse neophyte.


----------  Forwarded Message  ----------

Subject: RE: [HBRobotics] Micromouse Fans check this out
Date: Fri, 02 Feb 2007 13:51:11 -0800
=46rom: Jim McGuffin <hbrobotics@spamex.com>
To: <a.errington@lancaster.ac.uk>

I have been building a Micromouse and maze over the last four weeks. =
I have
some links at home that I will send later. Here is a status report on=

I purchased a starter platform from Trossen robotics. It consists of =
stepper motors and a simple aluminum frame and some wheels. You have =
provide everything else. The people who win these events do not use
steppers; they use regular DC motors, but this was a way to get up to=
quickly and start testing sensors and algorithms. I have a Robostix (=
Gumstix) for the main processor. It has a AVR that I programmed in
 assembly. So far it just moves forward a few feet, turns around and =
 back. I could enter it in this year's TableTopChallenge (tm). One th=
 that is harder than I expected is getting it to turn precisely 90 or=
 degrees. The wheels slip a little. I put a gyro on it but haven't
 integrated it into the code yet.

The last few weekends I worked on the maze itself. They are somewhat
difficult to build. I bought some posts and walls that are IEEE regul=
size and color. A standard maze is 16x16 squares, which is about 10x1=
 feet, which is too large for most houses. They sell the walls in set=
s to
 build 5x5 mazes. I bought two, so my maze will be 6x8 with some piec=
 left over. I bought a sheet of MDF and drilled the holes, which is t=
 hardest part. They have to be perfectly placed, and they have to be =
 It is very hard to get 7mm drill bits. The nearest non-metric sizes =
 17/64ths and 9/32nds. One would not fit and the other fell out. Then=
 painted the board with "blackboard paint" as is specified by the rul=
 Blackboard paint has some abrasives in it which help the traction. T=
 maze looks pretty good.

I planned to bring the mouse and maze to the last meeting but was una=
ble to
because of work issues.

As far as sensors are concerned, I plan to use Sharp GP2D120's (4-30c=
range) for the 4 side sensors and GP2D12's (10-80cm) for the front an=
d back
sensors. I have noticed that others do not use these. It looks like s=
people just illuminate the wall with IR and measure the strength of t=
returned light. This is simple but subject to problems from ambient l=
 I think they calibrate before each race. I would assume that the lig=
ht is
 modulated at 40khz or something to help but I do not know if this is=

A key design element seems to be to keep the parts close to the cente=
r to
reduce rotational inertia. And of course they are very light.

There is a Micromouse competition on Feb 26th in southern California.=
 I am
going. It is in the Disneyland Hotel. Here is a link:
<http://www.apec-conf.org/content/view/7/12/> It is affiliated with a
conference but the Micromouse event is free for competitors and spect=
(sadly, I will be the latter this time around).

Jim McGuffin

PS - as you know, I do not like to name my robots, hence A2 and A3 (t=
he A
just means autonomous). Since all of my bots will be autonomous from =
one, I am calling this one M1 (for mazebot 1).

>I love how fast they go.  I have done some google searches in the pa=
st on
>Micromouse.  Does anyone have any links to a site that explains what=
>sensors, motors, etc, people are using, to get the speed up there?

Valentine=92s Day -- Shop for gifts that spell L-O-V-E at MSN Shoppin=


| Silicon Valley Home Brew Robotics Club Mailing List
|    For Club info Visit:  <http://www.hbrobotics.org/>
| hbrobotics@wildrice.com               Send a post to the list.
| hbrobotics-join@wildrice.com=09       Join the list.
| hbrobotics-digest@wildrice.com=09Join the list in digest mode.
| hbrobotics-leave@wildrice.com=09Leave the list.