NASA and the Vision thing

Not to my knowledge. Ariane as I said represents a level playing field.
On the subject of Proton/Long March there is the question of
competition based on low labor costs. These costs are certain to rise.

It also needs either some very high bandwidth sensors or some massive
amount of processing power for AI (and the development of that
software) or both (to account for what it does during communications lag).

At LEO the obvious solution is high bandwidth with transfer function
processing only to compensate for time delays. The launching of a set
of MEO communications satellites would cost a lot lot less than the
ISS. If we are separated from where we are working with a time delay of
x we have a Nyquist frequecy of 1/2x, up to 1/2x we can compenstate
using transfer functions.

As has been said the repair to Hubble is so slow that if you had a
telepresence as far as the Moon, it would not make that much difference.

wrote in message
ups.com...
Not to my knowledge. Ariane as I said represents a level playing field.

Once again, you are evading the question. You made a point that the Ariane
lower stage could be reused. I asked you:

Exactly how many Ariane lower stages have in fact been recovered and
reflown? Verifiable cites, please.

And you replied:

The same number as the number of missions.

Are you saying that there is only one lower stage and it's being reused
every time? Please provide a verifiable cite that shows that *any* Ariane
lower stage has been reused.

On the subject of Proton/Long March

I'll be happy to discuss them *after* you have provided a verifiable
reference to the reuse of Ariane lower stages.

wrote in message
oups.com...
At LEO the obvious solution is high bandwidth with transfer function
processing only to compensate for time delays.

What sort of "transfer function" eliminates the delay caused by the speed of
light?

The launching of a set
of MEO communications satellites would cost a lot lot less than the
ISS.

Such as? Let's have some verifiable numbers.

If we are separated from where we are working with a time delay of
x we have a Nyquist frequecy of 1/2x, up to 1/2x we can compenstate
using transfer functions.

How does adding in the *additional* delay caused by the processing of these
"transfer functions" cause the signal to arrive faster?

As has been said the repair to Hubble is so slow that if you had a
telepresence as far as the Moon, it would not make that much difference.


How do you know this? Why is the Hubble repair "slow"?

In article om,
wrote:
Proton and Long March are of course even cheaper but P/LM are only
cheaper because of labor costs.

If you meant to say "low wages" -- the costs for *all* current space
launch systems are totally dominated by labor costs -- that's not the
whole story for Proton in particular. It would be cheaper even at Western
wages, because fewer people are needed to build it and launch it. The
Soviets invested heavily in setting up truly *operational* launch systems,
with hardware optimized for easy manufacturing, factories set up for
volume production with a minimum of hand labor(*), and mechanized launch
facilities that can be run by small crews.

(* A Proton goes from metal plate to finished rocket in 11 months, in a
factory that works one shift five days a week. Western launchers tend to
take 2-3 years with two or even three shifts at work. )

Proton *costs* would certainly go up if the crews were paid Western wages.
It's not clear that Proton *prices* would go up, because they are quite
blatantly set in reference to Western prices, rather than being based on
actual Proton costs, and the profit margin appears to be large.
--
spsystems.net is temporarily off the air; | Henry Spencer
mail to henry at zoo.utoronto.ca instead. |

What sort of "transfer function" eliminates the delay caused by the speed of light?

You have a Nyquist frequenct of c/2d. That is your limit. What you do
is increase the high order components up to this freqency.

how does adding in the *additional* delay caused by the processing of these
"transfer functions" cause the signal to arrive faster?

It dosn't, it can't but the fact is that the transfer function is in
fact giving you the impression of faster arrival.

wrote in message
oups.com...
how does adding in the *additional* delay caused by the processing of
these
"transfer functions" cause the signal to arrive faster?

It dosn't, it can't but the fact is that the transfer function is in
fact giving you the impression of faster arrival.

I'm certain that, as our rover is tumbling down the slope, it will have a
*real good* impression of those "transfer functions".

As I have explained it can't get rid of the time delay but what it can
do is make this delay less aparent psychologically. Let us assume we
are doing a fairly simply task. I have a spanner and I am unscewing a
bolt on Hubble. I put a torque on and as soon as the bolt moves I
reduce that torque. If I deliver x Nm through my glove sensor when it
moves I will still be appling a torque of x until it moves. I will
still feel the pressure on my hand even when it is moving. There is a
feedback. I am not (even in the case of zero delay) stopping the torque
for at least 0.2sec. If the bolt is remote sending a sharp pulse back
will give me the feel of proximity. This is the way our sense of touch
operates.

If we were being even more intelligent we would limit the amount the
spanner (on Hubble) could move until signals had been sent by me. If I
were to turn the spanner really fast however, the robot would follow
me, but not before it knoew I was turning fast.

This contains the mechanism for compensating for delay by means of
transfer functions. Clearly this method is only valid where the delays
are small (LEO and MEO).

wrote:
: As I have explained it can't get rid of the time delay but what it can
: do is make this delay less aparent psychologically. Let us assume we
: are doing a fairly simply task. I have a spanner and I am unscewing a
: bolt on Hubble. I put a torque on and as soon as the bolt moves I
: reduce that torque. If I deliver x Nm through my glove sensor when it
: moves I will still be appling a torque of x until it moves. I will
: still feel the pressure on my hand even when it is moving. There is a
: feedback. I am not (even in the case of zero delay) stopping the torque
: for at least 0.2sec. If the bolt is remote sending a sharp pulse back
: will give me the feel of proximity. This is the way our sense of touch
: operates.

: If we were being even more intelligent we would limit the amount the
: spanner (on Hubble) could move until signals had been sent by me. If I
: were to turn the spanner really fast however, the robot would follow
: me, but not before it knoew I was turning fast.

: This contains the mechanism for compensating for delay by means of
: transfer functions. Clearly this method is only valid where the delays
: are small (LEO and MEO).

The problem is Hubble's gyros are due to fail in 2007 or after. To do the
type of telerobotics that you suggest takes time to develop and time to
test, the latter presumably on the space station, which seems the logical
choice. To implement all this in the timely matter needed we'd be better
off sending an astornaut repair crew, which has already done several
similar missions.

Eric

On Wed, 21 Dec 2005 17:51:34 +0000 (UTC),
(Eric Chomko) wrote:

wrote:
: As I have explained it can't get rid of the time delay but what it can
: do is make this delay less aparent psychologically. Let us assume we
: are doing a fairly simply task. I have a spanner and I am unscewing a
: bolt on Hubble. I put a torque on and as soon as the bolt moves I
: reduce that torque. If I deliver x Nm through my glove sensor when it
: moves I will still be appling a torque of x until it moves. I will
: still feel the pressure on my hand even when it is moving. There is a
: feedback. I am not (even in the case of zero delay) stopping the torque
: for at least 0.2sec. If the bolt is remote sending a sharp pulse back
: will give me the feel of proximity. This is the way our sense of touch
: operates.

: If we were being even more intelligent we would limit the amount the
: spanner (on Hubble) could move until signals had been sent by me. If I
: were to turn the spanner really fast however, the robot would follow
: me, but not before it knoew I was turning fast.

: This contains the mechanism for compensating for delay by means of
: transfer functions. Clearly this method is only valid where the delays
: are small (LEO and MEO).

The problem is Hubble's gyros are due to fail in 2007 or after. To do the
type of telerobotics that you suggest takes time to develop and time to
test, the latter presumably on the space station, which seems the logical
choice. To implement all this in the timely matter needed we'd be better
off sending an astornaut repair crew, which has already done several
similar missions.

The 'planet finder' telescope mission will be underway in 2007, and it
will make Hubble look like a set of Opera glasses.


--

Christopher