[ExI] Engineering

[Adrian Tymes]

On Mon, Dec 17, 2012 at 7:53 AM, Keith Henson <hkeithhenson at gmail.com> wrote:
> On Mon, Dec 17, 2012 at 4:00 AM,  Adrian Tymes <atymes at gmail.com> wrote:
>> Right, but do *you* have the specific blueprints that they used?
>
>> Could you bend metal to put that particular thing together?
>
> Why put the burden on me?

Because you are the one who is making the proposal.

Even if you hire someone else to do it, they're still going to have to
prove out the systems.  This is so ingrained in engineering startups,
that practically all sources of financing for them require that such
iterative steps be followed.

Proposing to just put the satellite up without these steps is as
unworkable as proposing to design and make an all-electric
supersonic airliner that will take paying customers without any
flight testing.

Or to design and make a new class of rocket that will take
payloads to orbit, without testing any of the subsystems.

Or to design and make a new automobile for sale to the general
public, without putting it through the usual array of safety tests.

You complain about the difficulty of getting all the funding
needed to do this thing.  I am telling you how to change your
plan to reduce this from "impossible" to merely "very hard".
Yes, it's going to add more steps.  You'd have to do them
eventually.

> Do you know how to refine and shape the
> aluminum in your soda can?

As it happens, I do.  Though I am not proposing to make soda
cans, so this is irrelevant.

> The problem with power satellites is *not* collecting power in space
> or beaming it to the ground.

Actually, it is *a* problem.  There may be larger problems, but
that doesn't mean the smaller ones can be ignored.

More importantly, demonstrating that you can solve what you
are certain you can solve, directly helps with the bigger problem:
getting the money to put the large satellite into space.

> And for fundamental physics reasons you can't start small.

Not to get useful amounts of power, true.  That doesn't mean
you can't prove the systems without spending a lot of money -
which proof you will need, in order to get the money to build
the full-sized version.

http://en.wikipedia.org/wiki/Pilot_plant

Pilot plants exist for a reason, even if they likewise run into
constraints of physics that keep their output to "measurable"
rather than "useful".

> 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.
(At least at first: costs come down the more financiers can
be certain that not only will the idea work, but that you - or
whatever specific persons are doing the asking - can in
fact pull it off.  This is why startups and large companies
make such a big deal over who their CxOs are - you and I
might see it as overhyped, but the people who provide the
money think differently than you and I.)