🚀 go-pugleaf

In article <BsH5G2.I6t@math.uwaterloo.ca> Mark Tilden writes:
 >"Why" an insect robot should be regarded as insect-like does not have to 
 >do with how much it looks like an insect in operation, but rather if
 >it shows [...] survival ability. [...] "can it negotiate and traverse
 >complex terrains without getting trapped".

But this is a "robotic survivor", not a "robotic insect".  Besides,
insects seem to be pretty bad at avoiding traps; flypaper, for
example, can be very effective.  The insects' advantage mainly seems
to be in numbers.

 >We steal bug strategies for two reasons:  they're available and sufficient
 >for the task, and they seem to exhibit sufficiently simple behavior that 
 >we vainly believe we can figure them out.

I guess at this point I have to ask what those strategies are and
who's "stealing" them?

 >Point is: anthropormorphic designs may be valid for Ma Nature and computer
 >simulations, but don't expect them to work well for real-world robots.
 >Severe redesign is necessary to make metal behave like bone (or exoskeleton).

The material used for the linkages should not significantly affect
behavior.  Certainly, the effector dynamics will change, but as long
as the linkages and tendons are strong, why should the choice of
material hamper an anthropomorphic design?  Alternately, why wouldn't
your claim apply equally well to arthropodan design?
-- 
	-	-	-	-	-	-	-
Christopher Ian Connolly			connolly@cs.umass.edu
Laboratory for Perceptual Robotics		wa2ifi
University of Massachusetts at Amherst		Amherst, MA 01003