Subject:
|
Re: Positioning
|
Newsgroups:
|
lugnet.robotics
|
Date:
|
Wed, 26 Jan 2000 18:46:57 GMT
|
Viewed:
|
826 times
|
| |
|
|
Hi Carsten,
The other day I was trying to get to Target. I didn't know where the store
was, so I asked directions. I was told:
<i>Go down the street to the second light and make a left, go two more lights
and its on the right.</i>
Strange. There was no timing involved or counting the turns of wheels, but I
got there.
It seems to me that humans use landmarks to find there way. It also seems to
me we build maps from a relatively small set of landmarks. A gas station. A
McDonalds. Go left. Go right.
Can you solve your positioning problems using a landmark system? Are there
colors or light patterns in the room that can be recognized?
If a landmark system is useful for your problem, I've got some code that will
store landmarks as the robot discovers them. Then you can play back the
landmarks and your robot can find its way home.
If this is useful for you, the code can be downloaded here:
http://www.geocities.com/Athens/Column/2507/Dowload.htm
The design of the code is here:
http://www.geocities.com/Athens/Column/2507/SoftBricks.htm
Specifically, look at the Mapping design.
The on-line reference manual is here:
http://www.geocities.com/Athens/Column/2507/Catalog.htm
And I'd be happy to show you how landmarks can be stored in these maps.
David Leeper (is not a GPS)
In lugnet.robotics, Carsten Müller writes:
> Hello,
>
> I think about a little robot, which is able to move around a 'normal'
> room/flat and always knows where he stands. The current version of my
> robot is able to detect the speed and direction he is moving. But as
> everybody can imagine, the error grows and the accuracy of the position
> fades away. Now i think about using 2 or 3 IR beacons in conjunction
> with a custom IR Sensor to determine the actual position of the robot
> inside the room
>
> Now my question:
> Has anybody played around with triangulation or are there other
> suggestions in how to calculate the position of the robot. My girlfriend
> insists, not to paint rectangular shapes on the floor. So I thought of
> another solution.
>
> regards Carsten
>
> --
> "All parts should go together without forcing. You must remember that
> all the parts you are reassembling were disassembled by you. Therefore,
> if you can't get them together again, there must be a reason.
> By all means, do not use a hammer."
>
> -IBM maintenance manual, 1925
|
|
Message has 2 Replies:
 | | Re: Positioning
|
| (...) OTOH, how many mid-course corrections did you make as you walked down the (linear, visible) street to maintain the precise orientation when you reached the light? Landmarks work if they are universally visible, but not if they are beyond the (...) (24 years ago, 26-Jan-00, to lugnet.robotics)
| | | Re: Positioning
|
| (...) that's one way. the other is to interalize a map and visualize your reference to it. there was just a (short) article in Backpacker mag (the one with "Europe" on the cover) on this. -- -Steven "Nothin' But Net!" (URL) (24 years ago, 26-Jan-00, to lugnet.robotics)
|
Message is in Reply To:
| | Positioning
|
| Hello, I think about a little robot, which is able to move around a 'normal' room/flat and always knows where he stands. The current version of my robot is able to detect the speed and direction he is moving. But as everybody can imagine, the error (...) (24 years ago, 26-Jan-00, to lugnet.robotics)
|
28 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
|
|
|
|
