Coats Jack <jocoats@AMOCO.COM> writes:
> ----- Begin Included Message -----
>
> [snip]
>
>   The reason that I joined this listserv
> is that I have a big interest in Robotics.  The only problem with this is
> that it is costly.  (any suggestions???)
>
>         2nd.  I'm got a slight ideal what a PIC is but can someone
> clarify it for me.... like it's applications, how hard they are to use in
> a robotic application.... and whatever you thinks would be usefull
> information.  It would be appreciated.
>
>         Thanks
>
>                 Andre Dault
>                 s635355@aix2.uottawa.ca
>
>
> ----- End Included Message -----
>
>
> Andre,
>
>  I might suggest looking into some papers on subsumption
> arechitecture, and implementing a "small" robot using it.  The
> reason I suggest subsumption is that you could implement self
> sufficient components, get the feeling of accomplishment of
> having each one work independantly, and then after having several
> working, you can have them work in unison to solve a 'bigger'
> problem.
>
> I see a PIC as being good for this, because of their small size,
> weight, external components required, and $ (especially for the
> small ones), you can afford to put a little computing power a lot
> of places, which should help in making your robot modular.
>
> I hope that helps ... JC

I agree. Although I'm not involved in robotics directly, my colleagues in the
AI group are, and this approach has been suggested. It would almost certainly
be the only sensible way to do it using PICs anyway, since they don't really
have enough program space for large applications.

Using the 16C84's would allow the devices to be programmed whilst still in situ
on the robot. I've been trying in-circuit programming using these devices and
it works well; we're considering the construction of a mobile robot using some
simple geared motors / sensors and a number of PICs programmable via external
jack lines. This is only one of a number of projects which may be constructed,
however! If it happens, I'll post some news.

Because of this interest, we'd be pleased if anyone who has already done any
work on a mobile robot in this way would get in touch and let us know how it
worked / is working...

Richard