Newsgroups: comp.robotics.misc
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!fas-news.harvard.edu!newspump.wustl.edu!news.ecn.bgu.edu!vixen.cso.uiuc.edu!howland.reston.ans.net!news.sprintlink.net!newsfeed.internetmci.com!server2.ols.net!redstone.interpath.net!sas!mozart.unx.sas.com!sasrer
From: sasrer@unx.sas.com (Rodney Radford)
Subject: Re: How to make simple wall-detection?
Sender: news@unx.sas.com (Noter of Newsworthy Events)
Message-ID: <sasrer.803314583@sas.com>
Date: Fri, 16 Jun 1995 14:56:23 GMT
X-Nntp-Posting-Host: cinnamon.unx.sas.com
References: <3rnimf$530@driene.student.utwente.nl> <3rp9lm$n0l@sundog.tiac.net> <3rrbub$nup@server.elysian.net>
Organization: SAS Institute Inc.
Lines: 43

silveroak@server.elysian.net (Ray Cochener) writes:
>Tom Cicatelli (cica@tiac.net) wrote:
>: Try looking at studfinders.  These devices are sold in hardware stores 
>: in the tool section and are used to measure wall density.  They could 
>: easily be hacked to detect walls.  Simply tie into the leds.  Three 
>: would probably work for you.  Left, front and right.

>	Reminds me of another device... don't know the name, but when my 
>parent's house was being appriased for taxes, the appraisor used a sonic 
>measuring device.. layed it along a wall and measured the length by 
>bouncing sound off the far wall. Check with your local house appraisor, 
>see what they cost, and if you can get one...

The problem with most of these is that they require a minimum of 1' or so 
from the target. This is because they typically use the same transducer to
both transmit and receive the pulse, and the transducer is still ringing
for awhile after the pulse.

I worked on a robot that used IR emitter detector pairs as wall avoidance
systems. If you are working on a maze running robot where you know the
material used in the maze and how it is painted, it is fairly easy to
even work out approximate distances from the wall based on the amount of
reflected light. Of course this is very dependept on wall type, color and
texture, so it is not good as a general avoidance system

We pulsed the transmitter at 40khz, and the receiver was basically a 40khz
5th order butterworth bandpass filter, with the output being fed to a peak
detector and then on to an A/D. We got pretty linear response from as little
as less than an inch to over 2 feet away with a flat black plywood wall.

If interested in more information about this, let me know. I should be able
to find the schematic and PCB layout we generated for it (both in postscript
form). The total unit was about 1.5"x2" and operated off a 9v battery. 

Several were built for a robot that competed in an IEEE robotics contest, and
then the same design was later used in another maze running robot for a local
robotics contest.

--
-----------------------------------------------------------------------------
Rodney Radford,     Senior Systems Developer,  Image Technology Group
sasrer@unx.sas.com  SAS Institute, Inc, Cary, NC 27513  (919) 677-8000 x7703
-----------------------------------------------------------------------------
