Newsgroups: comp.robotics
Path: cantaloupe.srv.cs.cmu.edu!das-news2.harvard.edu!news2.near.net!news.mathworks.com!hookup!swrinde!pipex!news.sprintlink.net!news.indirect.com!seeker
From: seeker@indirect.com (Stan Eker)
Subject: Re: Ultra-sonic ranging
Message-ID: <D35out.GrH@indirect.com>
Sender: usenet@indirect.com (Internet Direct Admin)
Organization: Internet Direct, indirect.com
Date: Sun, 29 Jan 1995 07:30:29 GMT
References: <3fr88p$8hl@ixnews3.ix.netcom.com> <D2s9FL.AH@indirect.com> <3g88bg$164@synapse.bms.com>
X-Newsreader: TIN [version 1.2 PL2]
Lines: 27

David Allen (Allen@bms.com) wrote:
: seeker@indirect.com (Stan Eker) wrote:
: >
: > Sure has.  The strength of the return echo relates to what it bounced off
: > of, primarily, and (very) secondarily to the distance.  It's useless if your
: > robot thinks the couch is 20 feet farther away just because it's made of
: > fabric and foam rubber instead of steel.


: could you use the strength to indicate a class of object?  ie, differientiate
: a wall from a person or table from couch?  it would be additional info
: with very little additional hardware.

: da

You could, but an obstacle is still an obstacle, and ramming into it is
probably a Bad Idea.  Any cheap and simple source of information you can get
into a robot helps it to identify real objects, though, so I'd likely add it
in, even if the data wasn't used in an early version of the robot.  Walls
and boxes are good reflectors, books and magazines not very, and people &
furniture are terrible (too soft), so the Polaroid ranger (among others)
amplify the heck out of the signal and clip it to a useable range.

Any interface (difference in density) will give an echo, and the higher the
density difference, the stronger the echo.  You *should* be able to get a
mild return from the heat output from a vent, but it'd be exceedingly weak.

