DPRG List  

DPRG: Cognitive Navigation Question

Subject: DPRG: Cognitive Navigation Question
From: Robert Singleton slugmusk at alias.flash.net
Date: Wed Feb 24 16:12:23 CST 1999

Dan Miner wrote:

> What if the robot successfully navigates the course in a way that
> is pre-declaired by the builder and clearly demonstrates that it
> is not simply following a wall because of the shape of the path?
> T-time example:  Start on LEFT side of A, go down to B (passing
> area C), turn around in B, go to C, turn around in C, return to A.
> Other paths could also demonstrate non-wall following.  Another
> example: the robot travels DIAGONALLY from A to B, then ...

Very good solution, particularly if the bot happened to have to dodge
out of area C to complete the trip from A to B, indicating that it had a
concept of where it was and where it was supposed to be.

Something else that would give a strong indication that it was cognizant
of it's location would be to start the bot in an arbitrary orientation
in area A.  Let it run around and figure out which way it's facing and
reorient itself.

> Perhaps we add this rule for the next contest?

It wouldn't really need to be a rule... DPA tossed a box into the arena
of his own accord.

> Or, maybe a Micromouse contest!??!?

I think there is good interest in this type of contest anyway... There
seems to be a segment of the club that pines for smaller and smaller
bots.... :)

Robert Singleton


More information about the DPRG mailing list