Robot Rules

A thread of 2 messages, 8 November 1991 to 11 November 1991, in the order they were sent.

#3093 Robot Rules

Howdy folks,

I just got a package of Robot/Cyborg design rules from GEnie, via Rob Dean.
I read through it and found that it was basically a compilation of already
published stuff collected from Book 8 Robots, 101 Robots, with some stuff
on sensors thrown in from the MegaTrav Ref's Guide.  (also the performance
stuff from the ref's guide)

But, unfortunately this did nothing to answer the questions on robot
construction I have.  Basically, all it did was to outline stuff I had been
doing already!  I did have some substantial disagreements with some of the
stuff that was new.

My questions on robots...
Pseudo-Biological Robots:  To produce a Pseudo-Bio, one makes a normal robot
design and multiplies the total weight by 1.5 and the total cost by 8.
Well, I disagree with the idea of multiplying the TOTAL by these figures.
	I can understand why this would be the case with the chassis and most
of the robots sensors.  After all, You have to fool someone into thinking that
a camera is now an eye.  A taste sensor is now a tongue.  A mechanical arm is
now a plastic covered replica of an arm with muscles and such.  A radio will
have to have it's anetenni disguised.  A radome will have to look like part of
a sophonts body.  All that makes sence.
	But what about parts that do not require access to the outside
environment?  A neutrino sensor?  A densiometer?  These don't require exterior
access.  They don't have to be disguised.  The power source?  No, the
the transmission?  Well, it has to walk like a sophont so yes, that should be
increased in price/weight.  Well all of these items are pretty much noise
level when compared to the robots BRAIN.
	The Brain:  I would assume that a robot of equal intelligence with
emotion simulation should 'act' sufficiently like a sophont, compared to a
pseudo-bio model.  Thus the brain shouldn't be subject to the cost/weight
modifiers.
	On the other hand, one could argue that a normal robot doesn't have to
act quite so 'realisticly' so a Pseudo-Bio brain would have to be more
sophisticated to deal with the task of acting like a sophont.  To this I have
to counter that the brain (if more sophisticated) should then get bonuses to
it's intelligence and education.  (logical?)  If a brain is 8 times as costly,
then it should have a modifier to intelligence.  If it is 1.5 times as massive,
then the extra processors in it should make it more intelligent.  (Right?)
	So what soaks up all this cost and weight?  The brain of a pseudo
bio is usually the most expensive thing in it.  A brain that has a good
intelligence Costs a lot.  To then multiply it by 8 with no real good reason
and no advantage to performance I can see is silly.  Opinions?

More Brain headaches:  In my opinion, there should be tech level effects on
data storage.  There are only 2 types available in Book 8: (Standard, Synaptic)
The Standard is TL 8 and the synaptic is TL 11.
There is NO diffence between a TL 12 standard and a TL 18 standard storage
unit.  This I don't like.  There ought to be an improvement of some kind.
(maybe a +1 education for each TL over 12?) Opinions?

Robot speed:  I think the max road speed of a robot should be calculated from
the same chart as the one in the Ref's manual.  Remember that is it's ROAD
speed, there are no obstacles in it's way!  Most robots end up with a high
power to weight ratio (even after you subtract the peripherals out)
You end up with a robot that should be able to run at 187 kph!  Most buildings
and starships could (at the VERY best) be considered off road.  A robot would
have to move at walking speed to go through a starship.  (Much Much slower
than it's off road speed)  {Even slower in something that looked like my
office or my room!}  So I don't see a robot with a large max speed as being
a problem.  Opinions?

Transmissions:  I think that a robots transmission should only have to deal
with the amount of power left after the peripherals are powered.  Thus if
a robot has a 20Kw power supply, and 11Kwatts are used by the brain and other
systems that are necessary while the robot is running at top speed;  Then the
transmission should be geared for the remaining 9Kw.  Opinions?

Flying speeds:  Here's an odd one, The max speed should come from the grav
thrust tables, but what about flying through a building or ship?  NOE?  well
110+kph speeds are going to be rediculous inside.

This brings on another question:  Where does NOE speed get used?  What altitude?
After all, we have TL-7 and 8, F-111's and Tornados flying just subsonic at
treetop height.  According to the ref's manual TL 8 max NOE speed is something
around 100 kph!  And here we've got them cruising along with no trouble at
nearly 10 times that!  5 times what a TL 15 ship is supposed to be capable of!
	So what gives?  If you ask me the NOE speed of a craft should depend
not just on the sensors of the craft, but the computer of that craft. Opinions?

Note:  These are my stands on how robots should be built.  So, when in the
future you see a robot design, it will probably be build with the above in
mind.  Specifically Pseudo-Bio models, and the Transmission stuff.

Scott Kellogg

#3097 Re: Robot Rules

> From: [email removed]
> Subject: (3093) Robot Rules
>
> More Brain headaches:  In my opinion, there should be tech level effects on
> data storage.

  Agreed, furthermore the power consumption for the processors and storage is
too high even by todays standards (I compared it with 1liter of Z80Ls:)

> (maybe a +1 education for each TL over 12?) Opinions?

  I'd rather take smaller units. With the size of the programs multiplying
beyond control for higher skillevels than 1, more storage is needed before
nore edu is needed.

>This brings on another question:  Where does NOE speed get used? What altitude?
>After all, we have TL-7 and 8, F-111's and Tornados flying just subsonic at
>treetop height.

  Lowest authorized altitude for the Swedish AF during peace is 20m over land
and 10m over water, no speed restrictions, so NOE has to be lower than that:)

  I suspect that Traveller NOE is the type of NOE attack helicopters use, ie
hiding behind treelines, flying under powerlines and bridges and following
roads.

> 	So what gives?  If you ask me the NOE speed of a craft should depend
>not just on the sensors of the craft, but the computer of that craft. Opinions?

  The computer is not all that influential. It is possible to fly the aircraft
version of NOE from 300 to 600 km/h without sensors (ie following roads,
under bridges, and so low that you have to pull up to avoid meeting traffic:)
*in small planes with quick reactions*. If the speed is higher, or the
reactions of the a/c slower, computers and sensors have a growing role.
I suspect that the size of the plane and its speed of reaction is the
greatest determining factor. There is a world of difference between flying a
BUFF and a small prop job at 10 meters, and a computer can compute the best
trajectory that *don't* inpact ground:)
  However, if you are planning to do it in bad weather or at night, sensors
are increasingly nessesary.

  For grav-powered vehicles, I'd suggest that they work in the same way as
helicopters do today (after all, both suspend their whole weight by thrust
alone and move about by vectoring that thrust, or by inducing thrust from
other sources). The effect for this with regards to NOE is that they can
travel arbitrarily close to the ground at high speeds, but that the maximum
acceleration limits their reaction-time and thus their NOE speed. Sensors and
computers would be useful for increasing prediction time and finding the
optimum envelope for minimum visibility.

> Scott Kellogg

- -bertil-
- --
"It can be shown that for any nutty theory, beyond-the-fringe political view or
 strange religion there exists a proponent on the Net. The proof is left as an
 exercise for your kill-file."