Space & Astronomy/Space and Astronomy (October 1993).iso/mac/TEXT/SPACEDIG/V09_4/V9_495.TXT

Internet Message Format | 1991-07-08 | 12KB

Return-path: <[e-mail]>
X-Andrew-Authenticated-as: 7997;andrew.cmu.edu;Ted Anderson
Received: from corsica.andrew.cmu.edu via trymail for +dist+/afs/andrew.cmu.edu/usr1/ota/space/[e-mail] (->+dist+/afs/andrew.cmu.edu/usr1/ota/space/space.dl) (->ota+space.digests)
          ID </afs/andrew.cmu.edu/usr1/ota/Mailbox/YYaolay00UkV4:qE5p>;
          Sun, 18 Jun 89 03:16:55 -0400 (EDT)
Message-ID: <[e-mail]>
Reply-To: [e-mail]
From: [e-mail]
To: [e-mail]
Date: Sun, 18 Jun 89 03:16:48 -0400 (EDT)
Subject: SPACE Digest V9 #495

SPACE Digest                                      Volume 9 : Issue 495

Today's Topics:
             Re: Space Station Computers
             Chaotic Pluto == no Nemesis?
NASA Goddard to open Space Robotics Advanced Technology Facility (Forwarded)
            RE:  Addition to mailing list
----------------------------------------------------------------------

Date: Thu, 15 Jun 1989 10:57-EDT
From: [e-mail]
Subject: Re: Space Station Computers

> Why not use an off the shelf computer?  Because space is very different than
> a computer room, or even an office desk.  There is a lot more free radiation
> to scramble chips, and although a crash on Earth is often annoying,it can be
> fatal if it's attached to your air supply (and even more dangerous things!).
> The closest to "off the shelf" that might be reasonable would be something
> that DOD uses, which is "rad hard", or at least mil spec... The
> 80386 stuff is a leap forward, there.  People can at least write some of the
> code on affordable, off-the-shelf '386 PC's.  Note that disk drives have
> never been fully qualified, so big virtual memory machines are certainly
> not simple to build.

There are some life critical support functions that may need extra
protections. But there appears to be a continuation of aircraft design
mentality here. In an aircraft things happen very quickly and need
instant response. In an orbiting shack there is very little that can go
wrong that doesn't give you hours if not days to deal with. So you lose
all power? Go fix it. You won't starve. You won't asphixiate for hours,
and if there are manual overrides on air supply, you won't asphixiate
at all. Your temperature might get unpleasant, but with 45 minutes sun
and 45 minutes shade, you should at least reach a balance.
So the O2  control craps out? Crack valves by manual and replace the
bad one. Meteor hole in a compartment? Get out, seal it off and go back
with a suit and patch it. I'd love to know exactly what a computer
could do that could cause harm faster than a human can react to shut it
down and correct the mistake. Even a bleed to vacuum is not going to be
that fast. And airlocks will be mechanically interlocked so that won't
happen either.

Lets leave the aircraft industry mentality in the air where it belongs.
A shack is a shack is a shack. If it's broke, go fix it. Maybe sit down
and think about it for an hour over a cup of tea first. No sense rushing.

> Why not use off-the-shelf unix?   Do you want to bet your life Berkley unix
> won't crash?  NASA is pretty fanatic about software verification,which means
> having source around, another problem.  My guess is that the planned flight
> system involves some kind of Posix compliant OS,preferably government owned,
> that integrates well with Ada, since everything is supposed to be in Ada.

Once again, so what? There are multiple machines. There are redundant
mechanical back ups. There are PEOPLE on board for chrissakes. If one
crashes, then reboot.

Besides, there are new UNIX like systems like CMU-MACH. Give it a few
years and I think it will be as solid as any operating system needs to
be. And it doesn't have many of the internal flaws that the standard
Unix has to live with. And who knows? Maybe they'll go public with it.

ADA is probably a mistake. I don't think I've met more than two people
who have written significant amounts of code in it. Object oriented
languages are the way the (non-government) world is going. This is just
another way in which the station is separating itself from the
mainstream of commercial technological development. Of course the use
of a UNIX clone will at least guarantee that the astronauts can smuggle
STARTREK on board. And I wonder if they will be on the INTERNET? That
could make for some interesting dialogue on Space(Digest) 1999...

> Why so many?  A lot of them will be used as distributed
> controllers.... Instead the data would be collected
> and distributed by yet other processors.

That is why an occasional crash is not serious. And SEU's ** can be
handled by parity memory, which most machines have.

> The '486 isn't hopeless, either.  That would be the logical upgrade, even
> if no plans are in the works yet.The station will be there for thirty years,
> so swapping out computer boards along the way is pretty much inevitable.

In thirty years the computer will walk on board by itself.*  Remember, 30
years ago computers were still discrete transistor and diode logic with
massive core memories of sometimes as much as MEGABYTE of memory. But
only the REALLY big ones. And they operated at hellacious speeds of a
few MILLION instructions per second! Some of them had high tech disk
drum memories that stood as high as a person. And there were probably
HUNDREDS of them in the US!!!

* You might mistake it for a CRAYnefly, so don't allow any flyswatters
on board :-)

** In line with the request for explaining such things, "Single Event
   Upset"

------------------------------

Date: 15 Jun 89 16:29:00 GMT
From: apollo![e-mail]  (Steve Rehrauer)
Subject: Chaotic Pluto == no Nemesis?

I seem to recall a thread here some time ago about the "Nemesis" theory
of a large solar companion out in the hinterlands.  If my oft-faulty
memory serves me this time, I seem to recall that the supporters of the
Nemesis idea have used "otherwise unexplainable" quirks in Pluto's orbit
as circumstantial evidence to support it.

Well, I was interested to read in the May/June issue of _The_Planetary_
Report_ that there is observational evidence that many small bodies in
space -- e.g.: Saturn's moon, Hyperion; many asteroids and comets; and
_Pluto_ -- actually have chaotic orbits.  That is, "chaotic" in the sense
of the "new science of chaos" (of which I understand nothing) as opposed
to the sense in which "chaos rules my desk" (which I understand far too
well for comfort :^).

If true, along with the somewhat disquieting notion that "the eternal
clockwork" of the heavens is ultimately NOT very eternal, it would seem
to chop off one of the thickest legs the Nemesis people presently stand
upon.  Perhaps Pluto's orbit is quirky simply because it IS.

Would anyone knowledgeable about such things care to comment?  Is this
old-hat to sci.spacers?

--
>>> "Aaiiyeeeee!  Death from above!" <<<  | Steve Rehrauer
    Fone: [phone] x6168             | Apollo Computer, Inc.
    ARPA: [e-mail]             | (A subsidiary of Hewlett-Packard)
"Look, Max: 'Pressurized cheese in a can'.  Even _WE_ wouldn't eat that!"

------------------------------

Date: 15 Jun 89 19:51:28 GMT
From: trident.arc.nasa.gov![e-mail]  (Peter E. Yee)
Subject: NASA Goddard to open Space Robotics Advanced Technology Facility (Forwarded)

Mark Hess
Headquarters, Washington, D.C.                      June 15, 1989

Michael J. Braukus
Goddard Space Flight Center, Greenbelt, Md.


RELEASE:  89-96

NASA GODDARD TO OPEN SPACE ROBOTICS ADVANCED TECHNOLOGY FACILITY

     NASA's Goddard Space Flight Center, Greenbelt, Md., will
open its new state-of-the-art robotics facility with a ribbon-
cutting ceremony on Tuesday, June 20, 1989.  The facility will
play an important role in developing a space robot to do things
never before done in space.

     The ribbon-cutting ceremony will begin at 10 a.m. EDT at the
Building 11 facility.  A demonstration of the facility's robotics
capabilities will follow the ceremony.

     The highly sophisticated facility will be used by Goddard's
robotics team to create, test and evaluate new robotic
technologies to support Space Station Freedom.

     As part of the Freedom project, Goddard manages the
development of the Flight Telerobotic Servicer (FTS), a robotic
device that combines teleoperation -- the using of a human
operator to direct the machine -- and autonomous capabilities for
performing tasks by itself but supervised by an astronaut.
Martin Marietta Space Systems Co., Denver, Colo., was selected by
NASA to develop the FTS.

     "The FTS will allow astronauts to accomplish routine work in
space," said Ronald Browning, Goddard's deputy director for Space
Station Freedom.  "FTS is designed to complement astronaut
extravehicular activity and will be used both with and without a
crewman."

     The new facility contains a gantry robot 40 feet wide, 60
feet long and 20 feet high with six degrees of freedom, capable
of lifting up to two tons of payload and applying 4000 foot-
pounds of torque.

     Suspended from one mast of the gantry will be a set of
teleoperated industrial arms, which will be used as an FTS
operational simulator.  Another mast carries a grapple to emulate
Freedom's remote manipulator system, which primarily will be used
to transport payloads to and from the work site.

     The facility also includes an operator work station
installed in a mockup of the Space Shuttle's aft flight deck.
This simulator will permit teleoperation of the robot, providing
valuable information about operating the FTS in the constrained
environment of the Shuttle.

     Located in a glass enclosed mezzanine, overlooking the
gantry robot, is what David Provost, head, robotics data systems
and integration section, calls one of the unique technologies
being developed by Goddard for the FTS project -- the Graphic
Robot Simulator.

     "This computerized simulator uses animated graphics to
determine such things as the robot's reach capability and
collision avoidance information," said Provost.  "It allows our
engineers to use engineering and design concepts to evaluate what
would be seen at Freedom Station 6 or 7 years from now.

     "The simulator is a very cost-effective system.  It reduces
the construction costs considerably for major spacecraft and
instrument subsystems and makes results available in a much
shorter time."

     Also located in the facility is a robotics test bed for the
evaluation of various computers, software, mechanical arms and
control devices.

     Commenting on the utilization of the robotics facility with
the FTS project, Stanford Ollendorf, chief, telerobotics
engineering office, said, "Goddard has been given a technical
challenge to build a robot to do things which have never been
done before in space."

     "With this facility and the team of specialists from
universities, industry and other NASA centers, Goddard will have
a positive impact not only on Freedom Station and the nation's
space program but also on the U.S. ability to compete in world
markets through technology transfer to private industry,"
Ollendorf continued.

------------------------------

Date: Thu, 15 Jun 89 09:05:16 CDT
From: [e-mail] (Jay Labhart)
Subject: RE:  Addition to mailing list



Would you please add this address to your mailing list.


In addition I would like to find out about a space
program called

  generic vhsic space born computer (gvsp?)


If any of your readers has any information concerning this
program it would be greatly appreciated.



Thank you

Jay Labhart
Merit Technology
[phone]

------------------------------

End of SPACE Digest V9 #495
*******************