Newsgroups: comp.robotics.research
Path: cantaloupe.srv.cs.cmu.edu!europa.chnt.gtegsc.com!news.sprintlink.net!uunet!in1.uu.net!newsflash.concordia.ca!sunqbc.risq.net!hobbit.ireq.hydro.qc.ca!NetNews.IREQ.Hydro.QC.CA!mboyer
From: ion@falcon.cc.ukans.edu (Iain Shigeoka)
Subject: Re: Help Define Term Robotics
Message-ID: <MBOYER.95Jul7150318@pellan.ireq-robot.hydro.qc.ca>
Lines: 120
Sender: news@ireq.hydro.qc.ca (Netnews Admin)
Organization: La division Robotique de l'Institut de recherche d'Hydro-Quebec
References: <1995Jun30.101949.96741@kuhub.cc.ukans.edu>
            <1995Jul2.115236.96925@kuhub.cc.ukans.edu>
Date: Fri, 7 Jul 1995 19:03:18 GMT
Approved: mboyer@ireq-robot.hydro.qc.ca, crr@ireq-robot.hydro.qc.ca

> : This is a somewhat simple question, my 9-year old asked:
> : "Is the tea kettle a robot ?" 


What is a robot?  It's a good question, which unfortunately doesn't
have any good answers.  In fact, there seems to be NO universally
recognized defintion of a robot.  I hope we get a good dialog going 
here so we can add this to the comp.robotics.research FAQ and charter!

The japanese tend to use the definition:  "any device which replaces
human labour" (Soska, 1985) which seems to be the definition people
are throwing around here.  This is an extremely broad definition of a 
robot and it can easily be stretched to include almost any machine we 
make today.

On the other hand, the Robot Institute of America uses the following
definition:
  
  "A programmable multi-function manipulator designed to move 
  material, parts, or specialized devices through variable programmed
  motions for performance of a variety of tasks."  
  
                                  (Schlussel, 1983)
  
This definition is pretty restrictive.  Are robots strictly industrial
robots and other work-related manipulators?  This definition
completely leaves out mobile robots, and even robots in the Isaac
Asimov (Golden Age Science Fiction) sense of a humanoid automaton or 
android.  It would be a shame to exclude the very types of machines 
that fascinated us as children (and probably led in some way, to our 
chosen careers).

So where does that leave us?  Perhaps we should follow the Japanese
Industrial Robot Association, which classifies robots into six
categories from manual manipulators to intelligent robots (Schlussel,
1983).  I'm sure this would satisfy many of us who love "a place for 
everything, and everything in its place".  So, who's working with a
JIRA Class 4 Robot?  :)

I'd say one of the best definitions, is unfortunately one of the 
vaguest.  It comes from McKerrow (1986):

  A Robot is a machine which can be programmed to do a variety of
  tasks, in the same way that a computer is an electronic circuit 
  which can be programmed to do a variety of tasks.  

It uses an analogy that most people can understand and which, 
strangely enough, seems quite accurate.  This definition gives a sense
of HOW to go about deciding if a machine is a robot or not on an 
individual basis, rather than trying to set down rigid guidelines.  It
implies a level of sophistication and semi-intellegence that I think 
is very appropriate for differentiating between "machines" and 
"robots".


While we're on the subject, what is robotics?  This question, like 
"what is a robot" is not as straight forward as it would seem.  I 
mean, you can't really say, "Robotics is the study of robots" can you? 
If that were the case, wouldn't robotics be the same as mechanical 
design with a little computer science thrown in?  I don't know. 
Perhaps it is.

Brady (1985) would like to think that "Robotics is the intelligent
connection of perception to action".  This definition seems to very 
biased towards the computer science (esp. Artificial Intellegence) 
aspects of robotics.  It doesn't mention anything about hardware or 
physical mechanisms of any kind.  Can you have a virtual robot (one
that completely exists in software) or would that be something else?

McKerrow's definition doesn't help us out here since it states, "A 
robot is a machine...".  Machines don't necessarily need to have 
moving parts.  In fact, strictly speaking, McKerrow's definition can 
be twisted to mean "A robot is a computer running a program which 
allows it to do a variety of tasks".  I know that's stretching the 
definition but it begs the question, "Does robotics (and robots) need
to deal with physical interaction with the environment?"  I personally
believe the answer to that question is yes.  I think a robot needs to
go out and physically move something (even just itself) around in the
"real world", not just sit and passively and observe.

McKerrow (1986) once again comes to the rescue with one of the better 
definition of robotics:

   Robotics is the discipline which involves:
    
    a)  The design, manufacture, control, and programming of robots.
    b)  The use of robots to solve problems.
    c)  The study of the control processes, sensors, and algorithms
        used in humans, animals, and machines.
    d)  The application of these control processes and algorithms to
        the design of robots.



REFERENCES

Brady, M. (1985). Artificial Intelligence and Robotics, Artificial
  Intellegence. 26, pp. 79-121.

McKerrow, P.J. (1986).  "Robotics, An Academic Discipline?", Robotics,
  2, 3, pp. 267-274.

Schlussel, K. (1983). Robotics and Artificial Intellegence Across the
  Atlantic and Pacific, IEEE Transactions on Industrial Electronics.
  IE-30, 3, pp. 244-251.

Soska, G. V. (1985). Third Generation Robots:  Their Definition,
  Characteristics, and Applications, Robotics Age. 7,5, pp. 14-16.

                                                        
***************************************************     
* Iain Shigeoka:  ion@falcon.cc.ukans.edu         *     
*                 http://falcon.cc.ukans.edu/~ion *     
***************************************************


--
 *********************** (moderated) ***************************
 Submissions:                 Meta-discussions/information:
 crr@ireq-robot.hydro.qc.ca   crr-request@ireq-robot.hydro.qc.ca
