redford@JEREMY.DEC (John Redford) (03/05/86)
Re: Ben Bova's comment that James Van Allen should replace the grad
students in his lab with robots before he replaces the astronauts in space:
The January issue of Byte has an article about automating an organic
chemistry lab ("Automation in Organic Synthesis" by Gary W. Kramer
and Philip L. Fuchs, Byte, Jan '86, pg 263). These guys are chemists
at Purdue. They use a robot arm to
handle samples and a number of uproc-driven instruments to do things
like clean sample tubes and control experiments. The problem in this
field is that a large number of experiments must be carried out to
learn the precise effects of variables on the chemical reaction rate
and yield. The experiments don't vary much from run to run and are
tedious to do, and so are prime candidates for automation. Their
system is fairly elaborate and took quite a while to develop, but can
make organic synthesis quite a bit easier.
So robots are already starting to replace grad students. I'm sure
that grad students bored by endless experiments don't mind. However, I
wouldn't take this as a big point in favor of using robots in space. You can
only automate what you fully understand. Chemical experiments have
been done for centuries, and the techniques and problems are well known.
Very little work of any kind has been done in space. Look at how
long it took and how hard it was to repair the Solar Max satellite.
It wasn't because the astronauts were stupid or clumsy; it's just
that no one had ever done it before and no one knew what was involved.
A more serious example was Viking. An enormous amount of work went
into developing the three life-detection experiments; they were
probably the most sophisticated automated experiments ever built.
And yet the net results were ambiguous: one experiment detected life
and the other two didn't. After a lot of analysis it was decided
that the positive indication was the result of strange soil chemistry
and not of life, but it would have been a lot more reassuring to have
someone there to run still another experiment. There would have been
no lack of grad students to volunteer for THAT job.
Now, teleoperators are a great deal more flexible than robots, and
could be put into use a great deal more quickly. As Dietz has
pointed out, they are already in use for deep-sea work (there's an
article about this in the latest "High Technology"). The key problem
there (as many have pointed out) is the time lag. Even in LEO there
will be a significant lag when the satellite is on the far side of
the earth from the operator. Assuming that the signal goes through
submarine cables at 0.7c and there are no delays in switching, the
best-case delay is about 100 ms.
One answer for this might be computer simulation of the motion of the arm.
When the operator moves the arm, the computer predicts what happens
and displays that on the screen. When the real position comes back,
the screen is updated to reflect it. This will cause an occasional
sudden jerk in the position on the screen, but that should be uncommon.
This might help the problems with operator fatigue that were
mentioned earlier.
At longer ranges, though, there's no substitute for having someone
there. GEO is probably marginal, and the Moon is right out. LEO is
practically next door. In fact, ever since the Shuttle started
sending up IEEE members, the Institute has taken to calling it Area
10. If we want to do things farther away than our low-orbit backyard,
we'll have to put people there.
John Redford
DEC-Hudson
Posted: Wed 5-Mar-1986 11:51 Jerusalem Local Time (GMT+2)
To: RHEA::DECWRL::"space@angband"dcn@IHUXL.UUCP (03/09/86)
The astronauts trained for many hours on how to repair Solar Max. They knew what was wrong with it, and how to fix it. In the case of unexpected difficulties, all they could do was prepare backup plans. When the astronaut in the MMU failed to hard dock with Solar Max, they tried the arm and got it. I think this is another case where it was cheaper and faster to train humans than to design specialized gadgets to do it automatically. Dave Newkirk, ihnp4!ihuxl!dcn