[ExI] Engineering

[Adrian Tymes]

On Wed, Dec 19, 2012 at 7:27 AM, Keith Henson <hkeithhenson at gmail.com> wrote:
> On Wed, Dec 19, 2012 at 4:00 AM,  Adrian Tymes <atymes at gmail.com> wrote:
>> On Tue, Dec 18, 2012 at 11:58 AM, Keith Henson <hkeithhenson at gmail.com> wrote:
>>> OK, if the people who can raise the money say they have to demonstrate
>>> things like power beaming before they can raise the money, fine.  They
>>> can raise the money to do whatever they think is needed to raise the
>>> rest of the money.
>>
>> Now you're asking the money-raising people to do engineering:
>> for them to demonstrate beam tracking and so on.
>
> No, all they have to do is raise the money.  Engineers can demonstrate
> that microwave beams can be pointed the right direction.

You miss the point.  The money raisers and the engineers need to
work hand in hand: the engineers get a little farther, then the money
raisers can get a little more, then the engineers have funding to get a
little farther - that's how engineering projects on this scale work.

> If it takes
> a $150 B dollar demonstration before you can raise money, there is no
> point in raising money at all.

Surely you don't think it takes $150 B to demonstrate beam tracking.

That said, it may well take $150 B in separate demonstrations - one,
then the next, then the next, raising $1 B or less at a time - before
you can raise the $150 B in one chunk you'd need for the full-up solar
satellite.

If there is no other path to that end - and there might not be - then the
point of raising the demonstration money is to get to the full-up solar
satellite.  Just because you can not jump everything in one giant leap
doesn't make the path not worth taking.

>> You need to demonstrate beam tracking before your project can
>> have the full budget raised for it.
>
> Given communication satellites, I don't think so.  But let me ask, how
> much of a demonstration do you think is needed?  Full scale, full
> power?  Reduced scale, reduced power?  Exactly how much reduction?

The demonstration needs to produce measurable results.

Moreover, you probably need to demonstrate many different aspects.
Beam tracking is just one.

Is there an existing rocket class that can take the payload you
envision?  If not, that will need to be developed and tested separately.
(This is probably WAY too expensive, so you need to design to use
some heavy lifter that is already in use - not just that someone else
used once, but that you could go out and purchase today if you had
the money.)

What about the rectenna?  You can probably demonstrate this quite
cheaply, for less than $1M, but there are not existing rectennas that
you can just buy, so you need to demonstrate that you can make
them.

Or sun tracking?  Yes, sun-tracking solar on roofs exists, but only
to a certain point: those systems are interrupted when the Sun goes
below the horizon each day.  Maybe make some sun-tracker that
works around the clock for an above-Arctic-Circle (northern Alaska
or Canada) summer?  (You wouldn't even necessarily have to go
there yourself: just hire a local, ship the unit up, and have the local
send you back the test data showing that it worked.)  Once you
have that, it's trivial to show that you can make one for the full-size
unit if you had the money.

...and that's the trend, really.  Identify all the components of your
system that have any significant engineering work that would need
to be paid for out of your budget.  Find ways to do the development
of them for less, such that you can show would-be investors that
there is as little "technical risk" (that is, as little risk of unanticipated
design flaws in the blueprints you make*) as possible.

* This means you also have to actually make blueprints for the
system, listing all the parts and preferably multiple vendors you
could buy them from.  This is another demonstration bit you can do
for cheap.

Put another way: investors in projects this large generally demand
that the project be reduced to "just insert money" before they will
insert their money.

Note that having a parts list - and preferably, a solid estimate of how
much you'll need to pay for various kinds of labor - will also give you
much better justification for how much this thing will cost.  You may
find that your initial estimate was either high or low.  This reduces
the "financial risk": the chance that the amount you're asking for is
not, in fact, how much it will actually cost.  Investors dread the
prospect that they'll put in $150B, you'll do some work, and then
you find out it'll cost another $300B.  They demand that you prove
that won't happen, at least to a very good likelihood.  If you can't -
well, others who want this much can, so those others get the
investment instead.

> Because at some point the demonstration costs as much as the full
> operating project.

Or more, when you add all the demonstrations together.  The point
is, though, that's not all in one go: the smaller chunks *can* be
raised separately, *and only after* they are all done can you raise
the big chunk you need for the full satellite.

>>>>> it's cheaper to set up the parts pipeline for hundreds of them
>>>>> than it is to build just one.
>>>>
>>>> Actually, when you factor in the cost of capital acquisition
>>>> on this scale, it might prove to be cheaper to build just one.
>>>
>>> Let's put numbers on it.
>>
>> Only if you include the cost of capital acquisition.
>
> Much unmarked sniping above,

I snipped out the rest because you were ignoring the part of
the budget I had just pointed out.

> what is "the cost of capital acquisition"
> that you are talking about here?

Raising money itself costs money.  People who raise money
don't work for free.  They need salaries while they're going out
and getting it: they know it'll be a long, hard slog that might not
pay off, and they need to pay their bills while they're working
on it.  Anyone who would raise money for commission only is
either inexperienced enough they're guaranteed to fail, or
scamming you, or both.

Further, you also need to give them a small cut of the money
raised (see the "Lehman Scale", for example).  Again, anyone
willing to work without this is scamming you - likely planning to
sit around and work for others who are giving a commission on
top of salary, while pocketing your paycheck to "work" without
results.

Further further, when you get to this scale of money, there are
inevitably kickbacks, compromises, and other things you have
to accept in order to get the money.  (You hear about the
cases where this comes back to bite the raiser.  You don't hear
about the cases where the money was raised free of these
things, because they practically don't exist.)

For instance, if you get most of your budget from the US
government, a standard approach is to spread the labor & vendors
around the country, so that many in the House of Representatives
can point out to their constituents the part of the deal that
benefits their district - not "because all of the US benefits", but
rather, which local suppliers get a piece of the action.

This can be inefficient - but if it's 1.5 times as expensive but 15
times as likely to get approved, what do you care so long as it
gets the satellite built and launched?  But you do have to budget
for this, and this is unlikely to be as much a part of the second
satellite's cost (since by then you can cite cheaper costs for
anyone who wants to go into space - and that's much easier to
spread around).

>>> There I agree with you.  One high powered CEO required.  What would
>>> you think of David Petraeus in that role?
>>
>> Not a bad first thought, but insufficient experience in technical
>> projects.
>
> I think he would be fine for this project.
> http://en.wikipedia.org/wiki/David_Petraeus#Education_and_academia

His experience is in politics and military.  He might be useful for
fundraising, but it takes a certain mindset to drive engineering
projects, and I see no evidence that he has it.

> If you don't like him, who would you suggest?

You.

>> What would you think of Keith Henson in that role?  ;)
>> Upgraded with more skills, perhaps - but skills can be learned.
>
> Not a chance.  I never been MBA material and don't have the reputation
> for leading large organizations.

MBAs can be gotten.  Rep for leading large organizations is gained
through first leading small organizations, and that is gained through
doing.  This is another reason I suggest mapping out a development
path and starting small: by the time you're asking for the full budget,
you'll have the rep you need.

> At 70, I am not likely to live long
> enough.

You've likely got at least 30 years left, the potential of cryonics aside.
Do you think this can not be accomplished, at least through first
full-up satellite launched, within that time?