Alt-BEAM Archive
Message #08266
To: beam@sgiblab.sgi.com
From: Bob Shannon bshannon@tiac.net
Date: Tue, 14 Dec 1999 19:35:51 -0500
Subject: [alt-beam] Re: Regarding Solar heads, such as Solarbotics model, and The tildens
--------------AAE258EEE62028F3C35155DC
content-transfer-encoding: 7bit
What you need to do, is recognise when you are in a 'good enough' light
pool.
If your solar cells are delivering maximum charge current to your
batteries then
looking for a 'better' location is a waste of energy and time.
In this condition, you know that the best behavior is to sit still, and
store energy.
Dennison Bertram wrote:
>
>
> Here's an Idea. Tilden say's that a robot should constantly
> look for better food. And while this is a good idea, what if
> it already happens to be in the brightest light pool? It
> doesn't make sense to leave. Sure it's not as interesting,
> but personally I've been on the side of the robots staying.
> Anyway, I used to be interseted in finding a way that a
> robot could tell if it was in the best source of light or
> not. I wasn't sure how to do this, thinking 'cds cells by
> the solarcell, some fancy circuitry?' well heres a
> simplified idea. Just use a solar head, that uses that new
> fancy stuff, I forget, 139 bridge, poppernets, or something?
> You know the one that blinks when it's "locked on". Well,
> instead of routing the outputs to one motor, route it to two
> so it's like a photovore, buzzing around looking for the
> brightest source of light. Then, once you find this source
> of light, the robot starts to blink to indiciate it's in the
> brightest light pool, and it aint moving. This would have
> several emergent behaviors, one of them would be, if another
> bot came by, and cast a shadow then the bot would very
> quickly move out of the way, avoiding the shadow. Flight
> behavior. Stuff like that. Anyway, someone should try it. I
> haven't studyed the poppernet (that's what it's called
> right?) circuitry for a few months so I'm a little rusty,
> and unsure about the conditions for a 'error' bit message on
> the 139. but it sounds feasible. dennison
>
--------------AAE258EEE62028F3C35155DC
content-transfer-encoding: 7bit
What you need to do, is recognise when you are in a 'good enough' light
pool.
If your solar cells are delivering maximum charge current to your batteries
then
looking for a 'better' location is a waste of energy and time.
In this condition, you know that the best behavior is to sit still,
and store energy.
Dennison Bertram wrote:
Here's
an Idea. Tilden say's that a robot should constantly look for better food.
And while this is a good idea, what if it already happens to be in the
brightest light pool? It doesn't make sense to leave. Sure it's not as
interesting, but personally I've been on the side of the robots staying.
Anyway, I used to be interseted in finding a way that a robot could tell
if it was in the best source of light or not. I wasn't sure how to do this,
thinking 'cds cells by the solarcell, some fancy circuitry?' well
heres a simplified idea. Just use a solar head, that uses that new fancy
stuff, I forget, 139 bridge, poppernets, or something? You know the one
that blinks when it's "locked on". Well, instead of routing the outputs
to one motor, route it to two so it's like a photovore, buzzing around
looking for the brightest source of light. Then, once you find this source
of light, the robot starts to blink to indiciate it's in the brightest
light pool, and it aint moving. This would have several emergent behaviors,
one of them would be, if another bot came by, and cast a shadow then the
bot would very quickly move out of the way, avoiding the shadow. Flight
behavior. Stuff like that. Anyway, someone should try it. I haven't studyed
the poppernet (that's what it's called right?) circuitry for a few months
so I'm a little rusty, and unsure about the conditions for a 'error' bit
message on the 139. but it sounds feasible.
class=860300616-14121999>
class=860300616-14121999>dennison