DPRG List  

[DPRG] Walking robot

Subject: [DPRG] Walking robot
From: futurebots fuboco at bellsouth.net
Date: Thu Sep 25 12:59:01 CDT 2003

Dan Gates wrote:
> I have to chime in here, I have been working hard on the algorithms for
> making my Biped (now dubbed S.A.M.M. for Servo Actuated Mechanical
> Man). There is definately a LOT of computing to do. Since this thread
> has started, I've decided to try a more ridged approach. I never could
> understand why robots like the Honda and Sony use the bent knee method,
> it always seemed like an awkward and power consuming way to go. I
> decided to take a walk, paying attention to my knees especially. My
> conclusion is that my knees are locked in the ridged position durring
> most of the stride, then bends and snaps the shin/foot forward quickly.
> There is an initial side shift and forward lean that starts the stride
> in motion, after that there is very little side shifting, it becomes
> more of a pivot and snap motion. Think of a man on stilts, he doesn't
> have the ability to shift onto one leg and balance there while the
> other leg is suspended in the air, but rather he does a pivot and snap
> to maintain his upright position. There's really no reason I can see
> that the Honda and Sony can't adapt to such a motion, I am going to
> give it a try with mine.
>  As for biped usefulness, I can think of a few.
> *Can a wheeled robot go up stairs that are common to a humans world?
> *I think people often forget that one of the largest industries, if not
> THE largest, is entertainment. People pay thousands a year on being
> entertained. There are many aspects of entertainment that a biped robot
> can do better than a wheeled robot. As was mentioned earlier, we tend
> to relate to robots better when they take on some human form. Soccer
> bipeds and boxing bipeds are already entertaining the Japanese in
> RoboCup and Robo-one.
> *If someone gets on the ball fast enough, I know of one HUGE area that
> bipeds would riegn supreme. We are now in the retirement stage of the
> Babyboom erra, there are assisted living homes going up everywhere. A
> biped robot would be accepted more readily than a wheeled robot, many
> of the assisted living people have recently lost a spouse and desire
> companionship (someone to nag at). I can see bipeds being a useful tool
> in assisted living centers. Imagion the busy young professional in his
> office thinking about how he just stuck dear old dad in a home. He gets
> on his computer, connects with cyber droid and sits and has a
> conversation with dad durring his lunch hour.
> *My point is, there are lots of uses for bipeds in this biped world of
> humans.
> *a biped can be programmed to climb a ladder.
> *a biped can climb down a revein
> *We are bipedal, it makes sense that a bipedal robot would do better in
> a humans world than a wheels robot, I think the point that Chuck was
> making is that a biped can go from one medium to another and still fair
> pretty well where-as a wheeled robot has to be modified or custom made
> for each type of terain. One little step in a house that has a sunken
> living room prevents most wheeled robots from freely traversing their
> own home. To a biped, that little step is nothing. Ask a person in a
> wheelchair what a small step adds up to, like a curb on a sidewalk, to
> most of us it's never even an issue, but to a wheeled device it becomes
> an unsurmountable barrier.
> -Dan
> --- Sanjay Dastoor <sanjayd at uclink.berkeley.edu> wrote:
> > <<Sorry for the long email, but I have actually condensed what I
> > wanted
> > to say.  I'll try to make it to the January meeting and have a more
> > involved discussion.>>
> >
> > The reason stair climbing cannot be sufficiently handled by passive
> > walking is that it is the most expensive movement for humans to make,
> >
> > in terms of energy.  You can't recycle energy when you climb stairs.
> >
> > It is one of those movements where balancing is necessary.  That's
> > also
> > why it's very tiring.  Now imagine how Groucho Marx walks.  He walks
> > everywhere with his knees bent.  Try it.  You get tired really quick.
> >
> > That's how Asimo walks.  When you're not using your legs as rigid
> > objects to pivot on, you expend way more energy.  This is why
> > passive-dynamic walkers are far more energy efficient than
> > Asimo-style
> > walkers.
> >
> > The reason I support passive walkers more than Honda's or Sony's is
> > because the latters' method involves the use of brute force in
> > providing the power and the massive amounts of computation needed for
> >
> > active balancing.  They have taken many years and many dollars to
> > produce these remarkable robots, but these robots are programmed for
> > the environments in which they are demonstrated.  It takes lots of
> > pre-programmed instructions to move one leg to the next stair, shift
> > weight so it won't fall over, pick up the back leg, ....etc.  Their
> > computers and their batteries are huge (a backpack).  Passive walkers
> >
> > are so simple that they can work with no brains, no actuation,
> > nothing
> > except mechanical principles.  THAT, to me, is impressive, that we
> > understand walking to such an extent that we can create a walker so
> > simply.  Granted their capabilities are limited right now, but the
> > potential they have is enormous, and I think they'll surpass their
> > rivals soon.
> >
> >
> > > Truthfully, the only real advantage I can see for a walking robot
> > over
> > > a rolling robot is on the scale of Absolute Coolness, where the
> > walker
> > > wins hands down.  For all more practical considerations, wheels
> > have
> > > the decided advantage.
> >
> > Incidentally, part of my support for walking robots is in terms of
> > planetary exploration as opposed to movement on paved or flat
> > surfaces.
> >   In this context, wheels are awful.  They are the reason that the
> > only
> > exploration of Mars has been accomplished on the most boring parts -
> > the flat parts.  There is substantial research at NASA and elsewhere
> > on
> > alternative methods of locomotion - legs, snakes, etc., where such
> > limitations wouldn't exist.  I think that wheels achieve more, but
> > only
> > because legged motion is expensive, difficult, and slow.  I think
> > that
> > will change, at which point legged robots will become more versatile
> > than their wheeled counterparts.
> >
> > Sanjay
> >
> > _______________________________________________
> > DPRGlist mailing list
> > DPRGlist at dprg.org
> > http://nimon.ncc.com/mailman/listinfo/dprglist
> __________________________________
> Do you Yahoo!?
> The New Yahoo! Shopping - with improved product search
> http://shopping.yahoo.com
> _______________________________________________
> DPRGlist mailing list
> DPRGlist at dprg.org
> http://nimon.ncc.com/mailman/listinfo/dprglist

Hello Dan,

Is there any sites out there that have some code for biped robots?
interfacing gyros, force sensors, etc., I'm trying to build a small
walking robot system..

Dan Mathias  
Future-Bot Components        Web:  http://www.futurebots.com
203 N. Pennock Lane             Email: fuboco at bellsouth.net
Jupiter, Fl. 33458  USA            Phone/Fax  (561) 575-1487
Robotic and Electronic Components for the Hobbyist and Professional..
We have Motors, IC Chips, PC-Parts, Sensors, Robotic items......
I take credit cards through http://www.paypal.com

More information about the DPRG mailing list