[Robotgroup] Divide and Conquer!

Acy Stapp acy.stapp at gmail.com
Thu Sep 13 13:13:44 PDT 2007


You want to read Rodney Brooks' "Cambrian Intelligence". One of the first
robots he built at the MIT media lab was Genghis, an insect-inspired walker
with distributed movement control using what Brooks calls the "Subsumption
Architecture". In his robots there is no "main processor" because all the
"heavy lifting" happens in a distributed fashion. Fascinating stuff.

On 9/13/07, Thelostcircuit <thelostcircuit at gmail.com> wrote:
>
> Hello everyone,
>
>                 I had an idea for a project. A sort of divide and conquer
> method of approaching a task. I want to design a 6 leg walking robot with
> the ability to use AI algorithms for object avoidance and things like
> that.
> However, that is not really, what I wanted to ask an opinion on.
>
>                 The ideat was since PIC chips are not that expensive and
> do
> not take up that much room. Why not divide the work into small tasks. Like
> one chip controlling each leg. The chip responsible for that leg knows how
> to operate the servos and things that move the leg. The chip hooked up to
> the sensors knows how to read the data and convert the data into more
> meaning full numbers and so on. This would leave more cycles for the main
> processor to to the heavy lifting. In addition, it would give you more
> room
> for more complex programs and algorithms to be performed by the main CPU.
>
>                 The only downfall I would think would the platform would
> have to be large enough to accommodate the extra electronics. It would
> also
> use more energy.
>
>                 Just wanted to pass the idea through the group to see what
> all your thought where.
>
> Jon
>
> _______________________________________________
> Robotgroup mailing list
> Robotgroup at puremagic.com
> http://lists.puremagic.com/cgi-bin/mailman/listinfo/robotgroup
>



-- 
Acy Stapp

"When I'm working on a problem, I never think about beauty. I think only how
to solve the problem. But when I have finished, if the solution is not
beautiful, I know it is wrong." -- R. Buckminster Fuller (1895 - 1983)


More information about the Robotgroup mailing list