Alt-BEAM Archive

Message #03896



To: beam@sgiblab.sgi.com
From: "John A. deVries II" zozzles@lanl.gov
Date: Fri, 28 May 1999 16:46:28 -0600
Subject: [alt-beam] Re: Efficiency and Good Old Stryder


At 03:45 PM 5/28/99 , Bruce Robinson wrote:
>Actually, Dave's comment is technically correct, at least in a
>qualitative sense. Efficiency can be expressed in another way: (Total
>Energy - Waste Energy)/(Total Energy). Lifting a leg (or the body of a
>robot) consumes energy. If that energy is not doing anything useful and
>can't be recovered, then it is waste. So the higher a robot lifts it's
>legs, or it's body, to no useful purpose, then the less efficient it is.

To a certain extent, yes. However, that robot is likely to be inefficient
because it drags several legs (at least two, more probably three) each time
it steps. Friction caused by those dragged feet eats up bunches of energy.
A Scoutwalker style robot is likely to be more efficient -- fewer legs
drag for each step. Although energy is wasted by lifting, I'm still
willing to bet that more is wasted to friction. I'm also willing to lose
the bet.

I will admit to having had another thought and can hardly wait to see the
video (which is supposed to arrive today.) This depends on exactly how the
gait works. I ain't gonna speculate until I see it.

>So how much weight is the robot supposed to lift?

I don't think it actually matters -- "as much as it can" obviously is the
limit although you'd probably use less (half the max, just for jollies?)

>And how far is it supposed to lift it?

"As far as it can in a certain allotted amount of time" although you would
have to be limited (in the method I describe below) to the height from the
resting point of the weight to where it interferes with the pulley.
Actually, you can trade this off for weight if you wish. All you are
measuring is force times distance: you'd also have to measure the amount
of electrical power it consumes in that time which might be more of a
challenge.

> And what if the lifting mechanism wastes energy -- does the robot
>get penalized for that, even though the mechanism is simply there to
>make the measurement?

This is a bit of a problem, but you'd use the same lifting mechanism for
each and thus the amount of waste there can be (a) calculated and thus (b)
accounted for. Again, appealing to Watt's method, you use a string (or
rope for a really STRONG robot ) connected to the "horse", thread it
over a pulley (which, hopefully, has very little friction), and connect the
other end to a weight. If you want to get really picky you'd have to
include the average weight of the string... The mechanism makes the
lifting consistent and is essentially there to make the measurement, but
isn't making the measurement the point in the first place?

If you can account for the lifting mechanism, you really don't care about
the force and distance and the electrical energy used as individual things
-- you'd only be interested in something like

(force * distance) / [(1/t) * integral from 0 to t of (instantaneous
voltage * instantaneous current)]

although you might come up with some other approximation of the electrical
power used. Once it is a ratio, that is the "useful work efficiency" (at
least in terms of moving) of a given robot. It might very well turn out
that the results would be quite surprising.

>That's the challenge, isn't it? To come up with a simple method that the
>typical amateur can use and still get a useful measure.

I dunno -- seems simple enough to me and I think it would be a gas to try
out (he wrote, using ancient slang.)


Hope y'all have a good weekend!

Zoz

---------------------------------------------------------------
John A. deVries II
zozzles@lanl.gov

------------------------------------------------------------------------

eGroups.com home: http://www.egroups.com/group/alt-beam
http://www.egroups.com - Simplifying group communications



Home