|
Posted by D Herring on February 18, 2008, 1:28 pm
MisterQuan@gmail.com wrote:
> I've been building a small navigational robot that (so far crudely)
> maps it's environment on the fly. To help eliminate drift where the
> robots path bends slightly instead of driving straight, I thought I
> could simply use an electronic compass module and self-correct using
> a constant heading from the compass.
>
> It has been suggested to me that this would not work as there is too
> much interference either from the robot itself of for other 'stuff'
> inside buildings that would render the compass more trouble than it's
> worth.
>
> Has anyone had experience with this?
Yes. Some rooms are worse than others. Get a cheap navigational
compass and sample readings across the floor (and across your robot).
That should give you an idea of the situation.
> Is it possible to overcome this?
If its a problem, no. You'd have to map out the magnetic fields
throughout the room -- which rather defeats the purpose...
> What could be used in place of a compass to get an accurate
> heading indoors (GPS won't work indoors / is too expensive).
There are some ultrasonic solutions and some optical solutions. Both
rely on placing beacons at known locations in the room. Neither is
particularly cheap. Many people solve this by using an overhead
camera (e.g. robot soccer leagues).
- Daniel
| 
Yahoo! 
Google 
Windows Live 
del.icio.us 
digg 
Netscape 
XML site map
> maps it's environment on the fly. To help eliminate drift where the
> robots path bends slightly instead of driving straight, I thought I
> could simply use an electronic compass module and self-correct using
> a constant heading from the compass.
>
> It has been suggested to me that this would not work as there is too
> much interference either from the robot itself of for other 'stuff'
> inside buildings that would render the compass more trouble than it's
> worth.
>
> Has anyone had experience with this?