I've got a whole stack of papers here that are very relevent to mobile
robots, but obviously copyright exists on those.
> Teaching our Mowbots the exact dimensions of our yards and how to navigate
> it is probably much less useful than teaching it what to cut and not cut,
> how to seek food (recharge) and self preservation (avoid damage/theft).
> "Teaching" is used here a substitute for design/program/build.
I agree totally, I don't see any relevance in worrying about mapping. If
we construct the senros properly and connect them efficiently to the
behaviours, the robot will perform very well without any explicit
programming. Just by using the grass sensors discussed before, sensors
to determine concrete/grass, some bump sensors and a low battery sense,
should be sufficient to achieve some very useful emergent behaviours.
Like Connell, I am not keen on persistent states, preferring a mostly
reactive system.
-- Dave Everett Email: deverett nospam at idx.com.au (c) 1996 - Copyright remains with the author unless explicitly stated