Subject:
|
Re: Autonomous Robot
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Thu, 10 Aug 2000 01:01:39 GMT
|
Original-From:
|
Steve Baker <sjbaker1@airmail#NoSpam#.net>
|
Reply-To:
|
SJBAKER1@AIRMAIL.NETantispam
|
Viewed:
|
827 times
|
| |
|
|
John Barnes wrote:
>
> I think that this autonomous robot thread has been the most interesting
> one on lugnet for a while.
Fun isn't it!
> I have a suggestion, which I think may date back 50 years to the early days of
> radio direction finding. I believe the British Isles are (were) surrounded
> by a system
> of transmitters with counter-rotating antennas, one rpm. Each had a different
> signal, so you listened and counted the time for each signal and deduced your
> resulting bearing from the transmitter.
>
> Can't you do this with two counter-rotating lasers! The RCX sits out there and
> counts out the time between three hits, 1 - 2 - 1 for example. The relative
> time
> gives the angle. There is an ambiguity but if the transmitter is in the
> corner of the
> room, then the ambigous position is not real.
Wait, wait, wait. You must need more than two numbers. Mathematically,
you can't specify a 2D position without TWO numbers (eg X and Y, or
Heading and Range, or two angles). Your timing between three hits from
two co-located lasers is only ONE number if the lasers rotate once per
minute - then if you know the time between hit-1 and hit-2 then you can
deduce the time for hit-3 - so that carries no new information. If you
only know one number then you know that your position lies on some
particular line (straight or curved) - but you can't know WHERE along
that line you are.
So, I think you'd need at least THREE rotating lasers...but then, how
do you know which one is which. The British radio detection system
(which my Father worked on BTW) used radio beams of DIFFERENT FREQUENCIES
so you can tell them apart. We can't (easily) do that with lasers because
cheap pocket lasers come in "any colour you like - so long as it's RED" :-)
I guess you could flash them on and off at different rates or something - but
then the exact moment of detection becomes problematic - and the arrival of
two beams close together would confuse the heck out of the detection logic.
> A crude approach at range, is then to move at right angles to the beam some
> known
> small distance using odometry which should be good for a short straight
> line and then
> resample the transmitter's new angle. You now have all the info to solve
> for the range.
But moving at right angles to the beam isn't possible - you don't yet know
where the beam came from because you need an omni-directional detector.
Hence, all you know is that your position lies on one imaginary line - then
you move some distance (in an unknown direction) and then know your new
position lies on a new imaginary line. Those lines aren't straight - and determining
where you are isn't exactly trivial!
> I have no idea if it can be made to work, I haven't tried it.
It sounds *possible* - but very tricky.
> Also, I wonder if you need lasers or if a pair of counter-rotating
> mini-maglites whould do it!
> They'd be a lot cheaper for testing out the idea.
I doubt it. My local Fry's electronics has laser pointers for $5 each.
Maglites are pretty costly compared to *that*. I'm rather interested
in getting away from lasers though - just from a safety perspective.
A passive digital camera might be a better solution...when they get a bit
cheaper!
--
Steve Baker HomeEmail: <sjbaker1@airmail.net>
WorkEmail: <sjbaker@link.com>
HomePage : http://web2.airmail.net/sjbaker1
Projects : http://plib.sourceforge.net
http://tuxaqfh.sourceforge.net
http://tuxkart.sourceforge.net
http://prettypoly.sourceforge.net
|
|
Message has 1 Reply:
 | | Re: Autonomous Robot
|
| What if instead of two counter rotating lasers one tower, we use two towers each with one laser and one non-lego IR emitter that counter rotate at a set speed under the control of an RCX. I believe we can make non-lego emitter/detector pairs that (...) (24 years ago, 11-Aug-00, to lugnet.robotics)
|
Message is in Reply To:
| | Re: Autonomous Robot
|
| I think that this autonomous robot thread has been the most interesting one on lugnet for a while. I am fascinated by all the creativity which pours out in response to this most interesting challenge. Personally, I have cheated this issue by using (...) (24 years ago, 9-Aug-00, to lugnet.robotics)
|
4 Messages in This Thread:
  
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
Active threads in Robotics
|
|
|
|
|
|
