Yeah, the brain is the real challenge, isn't it? Mind you, my handy
skills are low enough to make a base a challenging proposition, but
the software is where the cool stuff happens. It's also where
having a goal comes in handy.

I like the idea of the Brooks subsumption architecture - lots of
little behaviours in a hierarchy, ranging from self preservation
through to awareness (or whatever). I believe that the irobot
roomba uses some modified form of it to do it's cleaning tasks. The
whole concept is nice in the way that you can build a little, test a
little, and you always have some complete subset of your target
behaviour, and the lower levels (ie self preservation, roaming, etc)
can be reused pretty easily in different scenarios.

But I haven't tried to apply it to something like robot soccer, so
I'm no expert on how useful it really is.

M

--- In robot_sig@yahoogroups.com, "Linus Dillon" <ldillon@...> wrote:
>
> Well, that shouldn't have any problems with bumps in the track.
>
> My own one is still a work in progres, but I have recently started
work on
> it again. It is also a modified RC car, but a very different one.
I have
> already hacked in a proper server for the steering, and have
modified the
> drive motor with a high-res shaft encoder (about 5kHz when running
at normal
> speed). I had already developed and built a motor driver board too
(cleans
> up the inputs for the encoder, and isolates the motor/servo
signals from the
> rest - uses a simple FET to switch the motor on/off and a dual-
pole relay to
> allow the direction to be changed).
>
> I had stalled previously designing/building the brains. It was to
be a
> custom board using a large 40-pin PIC chip, complete with LCD,
512kb memory
> (for logging), sound, IR, etc etc. All in all overkill and way too
much work
> (although I had finalised the design and had a shot at etching the
board -
> but the board traces were fine enough to give me trouble). I am
now looking
> at using a PIC-AXE chip or two instead, and just doing what is
needed.
> HOWEVER, having said that I still plan on adding one interesting
addition to
> it to hopefully give it an edge over the others. We'll see how it
turns out.
>
> I'm currently sorting out the logic/etc for an 8-pin PIC-AXE which
will
> handle the main motor drive PWM seperate to the main controller.
>
> I'll post photos when I get a little further along.
>
> Not many responses yet to the recent posts here. Anyone interested
in a
> meeting sometime? Maybe early next month? With luck the new
Mindstorms
> should be in the stores, so we should have stuff to play with. I'm
happy
> enough to put together some stuff to talk about/look at, or even
organise at
> a pinch (provided more people express interest in comming).
>
> -Linus
>
>
> -----Original Message-----
> From: robot_sig@yahoogroups.com [mailto:robot_sig@yahoogroups.com]
On Behalf
> Of Marcin
> Sent: Tuesday, 29 August 2006 9:56 PM
> To: robot_sig@yahoogroups.com
> Subject: [robot_sig] Re: Quinn's Project
>
>
>
> I've put a photo of my current robot in the 'photos' section. It's
> about 30x30cm, uses a cheap RC crazy car (called ugly predator) as
> the base with all the electronics disabled, and has a couple of
> sensors (rotating sonic rangefinder, ir rangefinders, odometry)
> motor driver and mcu on it. It doesn't do too much other than
> wander around at the moment (it would be good to give it a goal)
but
> it's good for trying ideas.
>
> Marcin
>