Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site ucbvax.BERKELEY.EDU
Path: utzoo!watmath!clyde!burl!ulysses!bellcore!decvax!ittatc!dcdwest!sdcsvax!ucbvax!space
From: dietz@SLB-DOLL.CSNET (Paul Dietz)
Newsgroups: net.space
Subject: Re: teleoperators
Message-ID: <8602080021.AA01079@s1-b.arpa>
Date: Fri, 7-Feb-86 17:09:36 EST
Article-I.D.: s1-b.8602080021.AA01079
Posted: Fri Feb  7 17:09:36 1986
Date-Received: Tue, 11-Feb-86 04:22:44 EST
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The ARPA Internet
Lines: 68

>> Because it's hard to keep people in GEO.  Teleoperator technology would
>>extend easily, perhaps all the way to the moon.

>As far as I can see, the only difference is in delta-vee, and robots and
>teleoperators pay the same that people do.

The difference is people have to be supplied with shielding, food,
water, air, and have to be brought back after a while (so the delta-V
isn't the same).

>Granted you don't want to do much construction at 100 nm, it doesn't follow
>that you want to do all or most in GEO.  HEO, 200-1000 nm up, might well be
>preferable.

I'll buy that for things built from earth-launched materials.  At what
altitude will the space station be?

>"I don't believe...". This argument is really about marginal cost, and neither
>you nor I have the figures.  The teleoperator/robot people have an advantage;
>they have fixed NASA manned costs to shoot at.  Paul, I will happily buy your
>argument, if you can tell me, with firm numbers, what the
>teleoperator-constructed Space Station will cost.   Until then, you're
>betting on the efficiency of an untried technology, a bet few of us would
>care to make without more information.

Yet we are willing to build the space station on very vague arguments
for its utility.  There's a double standard here.

I'm really arguing plausibility.  Human EVA in low orbit is very
expensive: I estimate very roughly around $100,000/man-hour of useful work 
when and if a space station is built (I'd like to the see the NASA number).
A $10 million teleoperator system will pay for itself in well under a year,
even if it is ten times less efficient than a man in a space suit.  Away
from the station an advantage would be even greater.

Let's take a crack at estimating the cost.  A teleoperator on a space
station will have (say): two direct drive robot arms with replacable end
effectors, a carousel device for holding the effectors, several CCD TV
cameras, a communications system (fiber optic or microwave), a control
computer, a power supply (fuel cells, batteries or direct wiring to the
station), and a frame to hold it all together and hold spare parts.
Also, it would need a ground interface: computers, input
devices, video and graphics displays.  I see no reason why such a device
couldn't be built for more than a few hundred thousand dollars (ignoring
development costs).  Even if space qualification multiplies that
by ten it should still make sense for EVA in low orbit.
Development costs are harder to estimate.  

To get a firm estimate of costs we need to know what the space station
will be used for (one can't, after all, predict how much a factory
will cost if you don't know what it's making).  Given what we know
about products that can be economically made in a space station,
I'd say extremely inexpensive teleoperators could make them just as well
as humans.

>No one thinks that such machines are impossible in any physical sense. I think
>that they're probably beyond current ME and CS technology.  I could be wrong.
>But I don't think increased demand is going to solve the problem for you;
>mechanical linkages really don't have terrific economies of scale.

You misunderstand.  I wasn't arguing that economies of scale will make
teleoperators cheap; rather, I was arguing that they hadn't been
developed because it hadn't been worth anyone's while to do so
for terrestrial applications.  Also, I'm not arguing for extremely
capable teleoperators that can fully replace humans with no loss
of efficiency; rather, I want remote-controlled dumb manipulators
that may not even have tactile feedback (although I'm willing to
take any additional functionality that's feasible).