Newsgroups: comp.robotics
Path: brunix!sgiblab!swrinde!cs.utexas.edu!utnut!cannon.ecf!hertz
From: hertz@ecf.toronto.edu (Roger B Hertz)
Subject: Re: Fast/Flexible Manipulators 
Message-ID: <CH05yI.CF8@ecf.toronto.edu>
Keywords: Fast, Flexible, Manipulators
Sender: news@ecf.toronto.edu (News Administrator)
Organization: University of Toronto, Engineering Computing Facility
References: <1993Nov23.124719@iastate.edu>
Date: Wed, 24 Nov 1993 15:35:05 GMT
Lines: 68

In article <1993Nov23.124719@iastate.edu>,
Chad E Bouton <bouton@IASTATE.EDU> wrote:
>-------------------------------
>Fast and Flexible Manipulators:
>-------------------------------
>
[some deletions]
>
>I have read about 30 papers so far related to this topic, and it seems
>there is always at least one more new paper in the IEEE robotics journal
>each month.  The problem for me is that I want to have an application that
>is worthwhile.  

Flexibility is a problem common to all manipulators, because it is a main
constraint limiting the performance bandwidth.  Whether you are dealing
with industrial or space manipulators, compliance limits the speed and
accuracy performance of the robot.  For most robots, this compliance 
is contained in the joints, rather then the links (space manipulators
are no exception).

The problem of compliance is being addressed by researchers in one of two
ways:
	1.  Working with it (understanding it and designing better
		control algorithms to reduce its effect)
	2.  Designing around it (redesigning the actuators, or entire
		manipulator to reduce compliance)  

>I can imagine many pick and place type applications where
>a lightweight, fast, and flexible robot with minimal vibrations would be
>needed.  But my question is: Does the lightweight arm (which leads to a
>flexible arm) necessarily lead to faster operation.  

A question in return: Why does a lighter arm imply a more flexible arm?
This is a *huge* misconception held by most roboticists.  Are you familiar
with work on using parallel mechanisms in robotics?  Although such
a mechanism will reduce the workspace or the arm, it gives the designer
the possiblity of using mechanisms with higher specific stiffnesses,
and placing the actuators closer to the base of the robot.  The result:
a much lighter mechansim with favourable stiffness properties.
Cases in point: most robots made by ABB, Fanuc, and Motoman now employ
a five bar mechanism, and several new robots use `in-parallel' actuation
exclusively (Smartee by Hughes, Delta from Switzerland).

But I agree, for the most part a lighter arm translates into a faster arm.
Without trying to trivialize the subject, remember that `F=ma'.

>I know motor
>dynamics play a big part in all of this and I have done some studying
>on that too.  I find this area fascinating, but I am concerned with
>finding a good real world application.

The bottom line (in industrial robots at least) is productivity.  If you
can increase performance for a reasonable cost, you have a real-world
application.  Even in space if you can speed up tasks (rather then waiting
30 seconds after each move of the RMS for the vibrations to die down), you
will be saving lots of time and money.

>--
>Chad E Bouton
>bouton@iastate.edu
>

Roger Hertz

-- 
hertz@ecf.toronto.edu
University of Toronto                               Ph: (416) 667-7744
Institute for Aerospace Studies                    Fax: (416) 667-7799
