|
[DPRG] Are representations what makes Can Can hard?
Subject: [DPRG] Are representations what makes Can Can hard?
From: Ed Okerson
ed at okerson.com
Date: Thu Dec 7 17:33:43 CST 2006
> Should we really be encouraging non-odometry, non-sensing
> versions of Can Can? It's not an idle question.
>
> Someone coming up with an easier way of doing things can really
> get one thinking.
>
> However for the most part we'd rather see broad skills promoted
> rather than arena specific solutions which have no broader
> purpose.
As I said at RBNO, I am amazed that the discussion has not changed in a
decade. It was always my understanding that the original intent of the
contests was to start someone out gently in the art of robot building.
They were steps of difficulty:
Line Following
Quick Trip
T-Time
Can-Can
>From Line Following you learned how to detect the lines used as end zones
in all the other contests. Quick trip encouraged some basic form of
odometry and adding wall sensors. T-Time added some turns to the
odometry. Can-Can was a kind of wrap up that added detection of something
other than the walls.
Over the years I have seen many course specific designs that tryed to jump
to the end, but were totally useless for anything outside the confines of
the specific rules of that one contest. Compare that to the more general
design of SR04 as stated on :
http://www.geology.smu.edu/~dpa-www/robots/sr04/sr04.html
The SR04 was conceived around the following very loose design criteria:
1. Survive in a wide range of (cluttered) human environments
autonomously and continuously, without getting stuck.
2. Provide a robust and reliable platform for developing navigation
and behavior software.
3. Be entertaining and aesthetic for the local human population.
No where does it mention collecting cans in the DPRG contest, but it
happens to do it very well.
Ed
More information about the DPRG mailing list
|
