|
[DPRG] Re: Right tool for the job
Subject: [DPRG] Re: Right tool for the job
From: John Swindle
swindle at compuserve.com
Date: Wed Aug 29 17:55:56 CDT 2007
The discussion about the quality of the data from the sensors,
and some discussion about discovering how some sensors work and
which ones are best: This discussion is precisely why I tried to
build my own sonar from scratch. It seemed that most people
treated sensors as simply something you attach to the robot, and
then tweak (or what I call "flutz") with the code until something
seemed to work, without really knowing what the sensor was doing.
I wanted to know exactly what was going on, before any code was
applied to interpreting the responses from the sonar. My goal was
to have no flutzing whatsoever. Some of the things I learned are
1) sonar frequencies that are too high will make big soft things
"disappear" and the robot will not see them, and 2) nothing is
truly directional nor truly omnidirectional, so without knowing
the polar pattern of the sonar, you're only guessing about what
it is "seeing".
Later,
John Swindle
More information about the DPRG mailing list
|
