DPRG List  

[DPRG] Re: Robot Navigatiion

Subject: [DPRG] Re: Robot Navigatiion
From: RoboVac robovac at mninter.net
Date: Tue Nov 20 10:44:16 CST 2001

A good way to detect if the robot was picked up is to use tilt sensors.  Not
only do they detect being picked up, but driving over bumps/slopes.

Your robot will probably need two types of navigation, one for indoors and
one for outdoors.  The simplest indoor system is wall following.  This has
the robot drive parallel to a wall and use dead reckoning to measure how far
along the wall it goes.  Alternatively or in addition it could measure the
distance to the walls in front of or behind it to determine how far along
the wall being followed it is.

The outside wall of the house can also be used for navigation, but with only
a single wall in the field of view something else will be needed.  A cheap
and reliable way for navigation in a backyard is to use ultrasonic beacons.
The way these work is the robot sends a radio signal to the beacons.  The
beacons then emit an ultrasonic sound.  By measuring the time between
sending the radio signal and hearing the ultrasonic signal the distance is
known.  With the distance to 3 or more beacons the robot can calculate its

The advantage of this is that the speed of sound is slow and easy to
measure.  Also the direction to the beacons does not have to be measured,
which is hard to do.

The disadvantages are that: the robot needs to be able to hear 3 beacons at
all times, the beacons need a direct line of sight to the robot at all
times, there should not be any ultrasonic reflectors to cause multiple paths
for the ultrasonic signal, and there should not be any sources of ultrasonic
noise around.

William Crolley

----- Original Message -----
>From: Byron Williams <williamsbyron at peoplepc.com>
To: Eric Yundt <eric at facetcorp.com>
Cc: <Beaudch at auburn.edu>; Dallas Personal Robotics Group
<dprglist at dprg.org>; Adam McGhee <amcghee at suscc.cc.al.us>
Sent: Sunday, November 18, 2001 8:37 AM
Subject: [DPRG] Re: Robot Navigatiion

> I'm looking to basically deliver beer to a location in the yard and be
> to get back to the fridge. My problem is I'm looking to counteract the
> effects of someone picking the robot up and moving it 10 feet. If I just
> remember where I am I'll be 10 feet the same direction it was moved off
> fridge location. It would be like finding two homing places. They may be
> around walls and through different terrain (concrete driveway, Brick
> grass...) I'm afraid the concrete will throw out following a wire and the
> grass will throw out following a line. Triangulatable beacons is what I
> decided to be the most feasible, have you any idea where I could find
> or even an article explaining exactly how they work? I have a general
> but I was afraid if I built them I may be breaking some sort of FCC
> transmission law and have a bunch of Ham radioers hunting me due to the
> induced noise from my apparatus or something.
> Thanks again for your help.
> John
> ----- Original Message -----
> From: "Eric Yundt" <eric at facetcorp.com>
> To: "Byron Williams" <williamsbyron at peoplepc.com>
> Cc: <Beaudch at auburn.edu>; "Dallas Personal Robotics Group"
> <dprglist at dprg.org>
> Sent: Saturday, November 17, 2001 5:02 PM
> Subject: Re: Robot Navigatiion
> > John, et al:
> > cc: DPRG
> >
> > Jim Brown no longer gets email at this address, but I would
> > encourage you to join our DPRG Robot Discussion List and engage
> > the many varied roboteers there.
> >
> > Follow the "DPRG mailing list" link (13th on left) for info
> > about joining the dicussion.
> >
> > DPRG Website
> > http://dprg.org
> >
> > Re: your task, it sounds like an interesting challenge -- I'm
> > sure you will get plenty of feedback on the list.
> >
> > Are you trying to find a homing place -- like a charging nest?
> > Or just where to deliver beer from the fridge?
> >
> > Have you considered simply following a wire?
> >
> > Or using sonar to maintain a map of the world and some idea
> > where in the world Waldo is?
> >
> > Perhaps triangulatable beacons would fit the bill?
> >
> > BTW -- I believe you can (easily?) get centimeter acuracy
> > with Differential GPS -- although this year old text might
> > try to discourage you otherwise:
> >
> > An Introduction to Differential GPS
> > http://www.oreillynet.com/pub/a/wireless/2000/12/29/two_gps.html
> >
> > Don't take their word for it!  ;-)
> >
> > Please do tell us more about what you are trying to do.
> >
> > Byron Williams wrote:
> > >
> > > Jim,
> > > Sorry to bother you. I ran across the site of the DPRG in some
> > > I was hoping you could guide me to some information. I am very new to
> > > hobby electronics. I am a visual basic programmer by trade and have
> > > worked with a few pics and things. I have a plan for a robot but can't
> > > get through one of the foreseen problems. I want the robot to be able
> > > accurately navigate in an outdoor environment to say three or four
> > > preset locations. I need accuracy  within about foot. The operational
> > > area of the robot wont be any more than the average yard. I could
> > > add GPS but I don't think I could get the location accuracy from that
> > > type system. Can you give me any pointers or point me in the direction
> > > of information on a system that may meet my requirements. I am not
> > > afraid to assemble this thing from scratch. I appreciate your time.
> > > Thanks,
> > > John Williams
> > >
> >
> > Thanks,
> >
> > --
> > Eric Yundt
> > DPRG
> >
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://nimon.ncc.com/mailman/listinfo/dprglist

More information about the DPRG mailing list