Path: utzoo!attcan!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!swrinde!elroy.jpl.nasa.gov!zardoz.cpd.com!dhw68k!felix!dennisg@felix.UUCP
From: dennisg@felix.UUCP (Dennis Griesser)
Newsgroups: comp.robotics
Subject: Re: Robot Fencer (was RE: robot pool player)
Message-ID: <155039@felix.UUCP>
Date: 22 Dec 90 02:25:49 GMT
References: <12247@life.ai.mit.edu> <21680@ttidca.TTI.COM> <2323@greek.csd.mot.com> <154798@felix.UUCP> <11560@goofy.Apple.COM>
Sender: daemon@felix.UUCP
Reply-To: dennisg@felix.UUCP (Dennis Griesser)
Organization: FileNet Corp., Costa Mesa, CA
Lines: 43

lance@motcsd.csd.mot.com (lance.norskog) writes:
>This brings up a subject dear to my heart:  I'd like a robot fencing
>partner.  Are there low-cost robots around which have the speed and
>flexibility to handle this?  I've got the Power Glove already...

I replied:
>How about taking your Power Glove, clamping on some Eyephones (or a cheaper
>equivalent), and firing up a Virtual Reality with a simulated partner to
>practice with.  If the simulation breaks down, at least it doesn't thrash
>around with a foil!

In article <11560@goofy.Apple.COM> dgiles@apple.com (Darren Giles) writes:
>Great, with one major problem... how do you get the foils to interact?
>Ignoring the math involved in the simulation, this is a case where there HAS
>to be strong tactile feedback.  Otherwise, you can't even parry!
>
>Which raises an interesting question (although it might fit better on a
>VR group): how could you give that kind of feedback?  Ideas, anyone?

Agreed.  Another poster mentioned synthesized sounds for the clash of steel.
Certainly a requirement, but tactile feedback remains the problem.

OK, so pick up an instrumented sword/foil/saber mockup.  Make sure that it
contains some weights attached to some type of fast effectors (solenoids?).

Then when your sword is hit by your simulated opponant, the solenoid on the
appropriate axis is energized, throwing the weight and pushing the sword in
the opposite direction.

Pretty silly idea, but it's a first cut.  :-)

OK, how about this:
	o For every degree of freedom measured by the PowerGlove, add an
	  effector to give you feedback.
	o Limit the force that can be applied by the effectors so that if the
	  simulation runs amuck, it can't hurt you.
	o You will have to add instrumentation at other parts along the 
	  human body so that you can gather additional information and assert
	  feedback in different positions.  Currently, other techniques are
	  used to locate the position and orientation of the Glove.

Hmmm.  If you increased the power of the effectors and made the software a
simple feedback loop, you would have a motorized exoskeleton.