Newsgroups: comp.robotics
Path: brunix!news.Brown.EDU!noc.near.net!howland.reston.ans.net!agate!ames!decwrl!decwrl!netcomsv!netcom.com!nagle
From: nagle@netcom.com (John Nagle)
Subject: Re: Very-short-range finder
Message-ID: <nagleC7F1Gp.EL1@netcom.com>
Organization: NETCOM On-line Communication Services (408 241-9760 guest)
References: <May19.175549.78132@yuma.ACNS.ColoState.EDU> <nagleC7D3GD.HB5@netcom.com> <1tj2sl$ctm@elroy.jpl.nasa.gov>
Date: Sat, 22 May 1993 06:46:01 GMT
Lines: 19

gat@robotics.jpl.nasa.gov (Erann Gat) writes:
>In article <nagleC7D3GD.HB5@netcom.com> nagle@netcom.com (John Nagle) writes:
>>       Relative IR sensors are good for this.  These consist of one
>>emitter and two detectors, with one detector closer to the target than
>>the other. 
>Does this really work?  I came up with this idea independently a few
>years back and thought it was really cool until I did a sensitivity
>analysis which revealed that minute errors in the measurement of the
>returned signal would translate into large errors in measured distance.
>(I don't recall the details, but it was something along the lines of
>1% signal error --> ~50% distance error.  It's not a difficult calculation
>to reproduce.)

        If the ranges of interest are 1-2", and you can get a distance
difference between the detectors of, say, 1/2", the error sensitivity
isn't that bad.  That's about the range used for grasp control.

					John Nagle

