Alt-BEAM Archive
Message #01985
To: beam@corp.sgi.com
From: Justin jaf60@student.canterbury.ac.nz
Date: Fri, 02 Apr 1999 23:57:35 +1200
Subject: [alt-beam] Re: Socer bots/ Hive behaviour
> > Actually, maybe just coloured lights, and light sensors with coloured
> > transplarent plastic over them is the easiest way to think about it.
> > Yeah, that makes it heaps simpler to talk about.
>
> To *talk* about, yes :)
>
> You seem to be quite convinced that BEAM soccer is feasible. But
> your simplifications sound a little suspect; I can see many ifs and
> buts under the carpet. How about some detail design?
In my first message I noted that this example was not about being
real-world ready, and that lots of issues would need addressing. To be
frank, I don't really care whether BEAM soccer is feasible, I don't find
the idea of robot soccer very interesting, my position is that
non-intelligent systems can work as a team. Soccer is just the context
in which I'm discussing this.
Anyway, regarding design detail, from my experiements with bicores, I
think the two bicore design (left/right, forward/back), with
colour-coded sensors wired up how I described (four banks: two in series
+ diode, these in parallel with the third + diode) probably would in the
real world, result in the kind of behaviour I described. The big thing
that is lacking from the example is a design for ball mechanisms, mainly
because I don't see this as relevant - it's needed for soccer, but the
issue I'm interested in is teamwork.
I can think of several ways to implement extra functions into the top of
the basic two-bicore design and really don't see it as especially
problematic. A 3-seond solution would be to put a magnetic pole, or IR
source, or whatever between the two posts at each end, disable the light
sensors (so that it becomes pole-tropic, or IR phototropic) when the
ball is in inside whatever the grasping/moving mechanism is.
And I'm thinking that this might be the real problem - the control
circuitry (be it BEAM or CPU) is fairly straightforward to get a
minimally functional design. A minimally functional and reliable ball
manipulator is probably a greater challenge. In terms of solving
real-world problems and designing for application, discussing the
control circuitry may be barking up the wrong tree.
What are the big ifs and buts that you see?
To tell the truth, yes, I think BEAM soccer is technically not very
difficult (with batteries I should add), but I don't really see much
reason, to the point where I'd say it's socially infeasible that someone
would bother to do it :) I wouldn't find it very interesting unless
there was competition between builders (ie lots of people make teams),
and then it becomes a really stupid idea because who wants to spend
their life making an entire team of robots for one competition? Everyone
would go for sumo instead.
------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/list/alt-beam
Free Web-based e-mail groups by eGroups.com
Home