Newsgroups: comp.robotics
Path: brunix!uunet!think.com!spdcc!das-news.harvard.edu!cantaloupe.srv.cs.cmu.edu!news
From: fitz@frc.ri.cmu.edu (Kerien Fitzpatrick)
Subject: Re: How to Explore Mars
Message-ID: <C14942.n2u.2@cs.cmu.edu>
Sender: news@cs.cmu.edu (Usenet News System)
Nntp-Posting-Host: dirt.frc.ri.cmu.edu
Reply-To: fitz@frc.ri.cmu.edu
Organization: Field Robotics Center, Carnegie Mellon University
References: <58691@dime.cs.umass.edu>
Date: Tue, 19 Jan 1993 19:36:00 GMT
Lines: 117

In article 58691@dime.cs.umass.edu, connolly@pooh.cs.umass.edu (Christopher Ian Connolly) writes:
> In article <1993Jan11.170735.20763@miki.pictel.com> wpns@miki.pictel.com (Willie Smith) writes:
> >I think everyone is setting up straw robots for the other side's
> >viewpoint.
> 
> I think this hits close to the heart of the problem.  The big/small
> issue is simply a tradeoff which is dependent on the problem to be
> solved and the available technology.  The apparent split in the
> robotics community over the issue seems to be artificial & largely
> political.  This might stem from a lack of precision in the problem
> statements.  For example, the term "planetary exploration" is still
> extremely vague, i.e., exactly what are we trying to automate?
> 
> -- 
> 	-	-	-	-	-	-	-
> Christopher Ian Connolly			connolly@cs.umass.edu
> Laboratory for Perceptual Robotics		wa2ifi
> University of Massachusetts at Amherst		Amherst, MA 01003

Some people are a bit sensitive.....a familiar phrase is "why are politics in
academia & research so vicious.....because the stakes are so low."

The arguments flying back and forth aren't really focussed on the issues -
more of pissing matches.  Let's look at nature (please refrain from 
religious issues) - after millions of years it looks as if a large variety
of sizes/shapes is the way to go.  It really does focus on what you want
to accomplish.

I've been at the FRC since its inception back in 1983.  Each project goes
through many phases - the first of which is to determine what is really
necessary.  The senior people (certain others posting here are not in
this group) do not have preconceived notions about how large to build
robots.  We maintain a mix of basic and applied research - if these things
never leave the lab then what good has come of it.  I am fascinated by
space research yet have little desire to involve myself with it.  I
prefer projects which will benefit people on Earth in the next 5-10
years - this is where I live.

We (the entire robotics community) have the ability to build capable,
redundant robots dependent on the mission objectives.  If the mission
objective is determine if there is water present on the surface of Mars
then thousands of small robots seems best to me.  On the other hand, if
you want to determine whether or not water exists, map the surface more
accurately, gather soil samples, ..... - then a handful of larger robots
seems best.  Everything depends on the complete mission objectives.
Large is a relative term....look at mammals on Earth.
It costs $$ to control large mass and it costs $$ to miniaturize.

Just so everyone understands - Gerry does not speak for the entire FRC
organization...just for himself.  The same goes for me and most likely
for Erann Gat.  True researchers have no bias and can recognize the 
appropriateness in all technology....not just that suitable to their
work.

The Erebus project was not something that fell within FRC's strategic
goals - it was an opportunity that was presented by others.  The senior
people within the FRC discussed this long and hard as to whether or not
we should do this.  Less than a year is incredibly ambitious considering
the amount of money available for a project of this scope.  Since
project goals were being discussed - there were two sets.  One from
NASA - prove that technology (mechanism, perception, & control)
can function in the Antartic (Mars) environment.  Then since Antartica is 
controlled by NSF (National Science Foundation) we had to incorporate
their own goals - provide information about the gaseous emanations
from the lava lake.  The robot developed formed from mountain climbing
experience - the incline of the inside of the crater varies, but 
typical range is 60 to 90 degrees.  Thus a walking mechanism was
incorporated into a rappelling scenario.  A tether with a kevlar
member would help hold the robot while it walked over boulders, etc.
To help limit the size of a robot - all power generation was done off-board 
and passed over the tether.  The onboard tether experience problems due
to manufacturing errors.  The kevlar member wound around the fiber and,
tensioned, degraded things such that the digital data channels would
not lock up reliably....to much db loss.  A backup fiber was installed
and would just drape.

The backup fiber was damaged somehow (haven't had a chance to talk with
someone that was actually there...) and was kinked.  This degraded the
signals such that the digital channels became flaky.  At first people
believed the electronics were at fault, but found the severely-kinked
fiber and attempted to un-kink which caused it to break.  FRC wanted
a fifth technician down there (the electronics guru), but we were
not allowed to have five.  Since all the electronics/tether hardware
was working fine up in Pittsburgh we hoped this would not become
a problem.  Whether or not a fusion splice could actually have been
made in the field is speculation - we were heartbroken that an event
such as that would bring down the project.  An analogy to space-flight
events would be that of mistakenly sending the command that turns the
antennae away from Earth.  Did the robot fail or did the support people
fail?  Should we really care?  Should we let concerns such as that 
inhibit us from taking the risk in the first place? - No.  Those of
us at the FRC knew this project was extremely risky and knew that
others would take every advantage of a failure.  No one can put aside 
that a capable robot was designed and fabricated.  Perception and
scientific sensors were integrated and functional.  It was transported
to an environment simulating that of Mars, reassembled from pieces,
and started up.  All systems were operating and the robot walked.
It did not complete the full mission and that indicates a mission
failure, yet there were many sub-mission successes.  If we don't fail once 
in awhile then we don't learn as much.  The next time we tether a robot -
(this was not our first...but the first optical-fiber) you can
rest assured that a tank will be able to park on the tether.

Hopefully, others out there can differentiate between constructive
criticism and criticism that is only offered to support another belief.


---
Kerien Fitzpatrick			Pittsburgh, PA 15213
Field Robotics Center			(412)268-6564
The Robotics Institute			Internet: fitz@frc.ri.cmu.edu
Carnegie Mellon University





