Alt-BEAM Archive
Message #00687
To: beam@corp.sgi.com
From: Steven Bolt sbolt@xs4all.nl
Date: Sat, 20 Feb 1999 12:04:17 +0100 (CET)
Subject: [alt-beam] Re: hmmm...i wonder which subject
On Fri, 19 Feb 1999, Terry Newton wrote:
> >But it seems that it would be less complicated and just as
> >functional if you let the uC control the motors with nothing but
> >dumb drivers in between.
>
> In the case of walkers, you may be correct, but such a setup
> probably will -not- be any better than a good analog design.
> If I learned anything I learned one can't throw smarts at a
> CPG walker and make it smarter. Mechanical sensors on the body
> work just as well. I'm also old fashioned and believe there is
> some truth to the notion that a real microcore is a computer of
> a different kind and probably cannot be fully captured in a
> simulation. The sim might walk, but the real thing will probably
> walk better and is certainly cooler.
I recall an old discussion about feedback. The essentials are on
http://www.serve.com/heretics/discus/messages/225/219.html?WednesdayNovember2519980931pm
A sample of words in there:
Mark Tilden replied:
Ach! You have found out my secret, which was never a secret to begin
with. It amazes me how many people think the 4-element microcore is a
form of all-controlling processor that somehow gives advanced
"intelligent" function without designer mechanical skills. Even the
best brains cannot make up for bad mechanics -- just ask Stephen
Hawking, and believe me, the microcore is hardly the best brain even
when biased and fed-back appropriately. It was just the first,
easiest, and most stable.
It seems to me that a little uC can perform all relevant functions
of the microcore above - but why bother, and yes, it would be less
cool :)
[ Nervous Net a problem rather than a solution in your case ]
> These are problems with using a CPG to drive motors with no real
> positional feedback, not problems with the nervous network itself.
> A simulation would have the same problem.
Agreed, of course.
Btw, it has always surprised me that you added a uC to a two-motor
BEAM walker, which seems very handycapped to begin with. Imho the
five-motor `Walkman' is by far the most sensible configuration for
any experiment involving Nervous Net stuff. If you want no more
than two motors, something like my Spider is superior - walks and
turns easily and predictably.
> >What I was looking for in this discusion is an example of synergy,
> >a combination of uC and Nervous stuff which is more capable then
> >either would be on its own.
>
> I think I came close to achieving that, although most of the
> functions that the PIC ended up being good at could easily be
> hardwired. Whereas other designs increased in capability by
> magnitudes by addition of a memory brain, the walker gained
> only about 50% and then only after I simplified it.
I was hoping that a machine learning device would be able to
learn how to influence microcored mechanics effectively.
> > You conclude that
> >
> > I'm left with the feeling that much more will be needed to make
> > these devices useful in the real world.
>
> :) I further conclude that real world operation is not necessary
> to have fun and learn.
Hear, hear! But actual capabilities - as opposed to glorified
random complexity - do seem rather attractive to me.
> The statement implies that I thought that eventually real-world
> applications would be achieved, still do. The "much more"
> doesn't have to be cpu, I've been seeing some mighty fine analog
> solutions lately that push the state of the art.
Indeed. Actually, a significant increase in the number and quality
of sensors on a small, mobile `bot is imho impossible to achieve
using a single cpu. If analog/parallel processing is possible,
great! But my eye does tend to wander in the direction of uCs like
the AT90S2323/43: 8-pin package, in-circuit reprogrammable, 2K
flash, 128 bytes ram, 128 bytes eeprom, very low cost. In SMT, no
larger than an ordinary transistor...
> The fundamental problem, at least as far as 2-motor walkers are
> concerned, is gaining useful control over steering.
Unless you rearrange the mechanics. My 2-motor Spider steers just
fine, without servo's or positional feedback.
> Stamp walkers turn on a dime because they use servos with exact
> positioning. Of course by doing that you lose adaptibility to
> terrain.
That seems relevant only when you go to the 5-motor `Walkman'
configuration. I'd love to see a machine learning device achieve
control of that hardware, without losing its Nervous advantages.
---8<---
> There is very little information about how best to connect
> feelers and eyes to a microcore. Hook 'em somewhere? I think
> more motors are needed to solve the steering problem. More
> research too.
Right; five motors. As to sensor attachment, to me it looks like
too much trial and error would be involved, unless you stick with a
rather simplistic beast. That's where I tend to switch to uCs.
> >[ Is the `brain' overrated? ]
>
> Probably. Must match body capabilities or mass confusion results.
> For walkers, no brain at all is probably a better solution.
I suppose you mean 2-motor, BEAM-type walkers?
> Otherwise, intelligence is limited by the body and the senses
> no matter how smart you try to make it. For walkers anything
> more than 7 bytes proved overkill, the pic-bot popper likes
> around 64 bytes, and the owi spider hack with IR has hundreds
> of memories and had quite an attitude until his gears stripped.
It deserves to be repaired :)
---8<---
> > but imho much more than the present level of capability can be
> > reached without any learning at all.
>
> Like a hard-wired reversing obstacle avoiding photovore. Being able
> to back up makes all the difference in the world. Someone's gotta
> do a (beam) photopopper that backs up,
Hey! My three photovores all back up nice and dandy, and they
don't even have reversable motors.
> a pair of h-bridges and some minor logic, shouldn't be that
> difficult.
Easier than you think, actually. No h-bridges required:
Linkname: Photovore - search the light, avoid obstacles
URL: http://arlene.xs4all.nl/WWW/e-fotovoor_beweging.html
> The most useful thing about my PIC-bot popper was not that it
> could learn (always learns about the same thing anyway) but that
> it could back away from obstacles in its way and not get stuck.
> There is much work to be done with simple analog circuits and
> bodies, please keep it coming. Like those 1 chip reversible
> walkers! Now that's beam.
And fun :)
Steve
----------------------------------------------------------------------
# sbolt@xs4all.nl # Steven Bolt # popular science monthly KIJK #
----------------------------------------------------------------------
------------------------------------------------------------------------
eGroup home: http://www.eGroups.com/list/alt-beam
Free Web-based e-mail groups by eGroups.com
Home