Thursday, October 28, 2010

The Industrial Scale of Space


How many people does it take to build a spaceship? The actual fabrication process might be entirely automated, but how large must a community or society be to have the productive muscle, and range of specialized skills, needed to build and operate spacecraft?

This question lurks behind the last couple of discussion threads, and many earlier ones. It is implicated in a number of classic SF tropes. How long can a crew keep their ship going before they need repairs that only a cageworks can perform? Are outpost colonies condemned to slide to pre-industrial conditions? (Or extinction, if they cannot survive without industrial technology.) Can more robust colonies maintain space fleets?

Note that industrial scale is quite different from techlevel, which is more or less whether a society knows how to build and maintain spaceships at all, and what kinds. A familiar example of industrial scale is automobiles. With a good machine shop you could build a car entirely from scratch, fabricating all the parts, but the cost in labor and shop time would be many times the cost of a production car.

A nod to Henry Ford, and once again to Adam Smith, who lived so early in the dawn of the Industrial Revolution that he only mentions the steam engine in a footnote, but who hit on the importance of the division of labor.


So, how many people does it take to build a spaceship? Certainly no more than three billion, the world population at the time of Apollo. And I would say no less than about 100 million, because France had to partner up in the ESA in order to get in the game. Even the grotesque exception that proves the rule, North Korea, has a population of 25 million.

I grant that national space programs are political entities, and an imperfect metric of industrial capacity - which is in any case part of an interdependent world economy, not neatly partitioned by borders. But it is the metric we have, and the ability to build space boosters corresponds roughly to the ability to build large commercial airframes, also confined to a few big economies.

Only a very large economy, a large industrial infrastructure, can support the web of factories and skunkworks, launch and tracking sites, academic institutes and training facilities, with their hundreds of specialized skills, that go into present day space operations.

This blog generally presumes, for the sake of discussion, that in the Plausible Midfuture the cost of space travel will be very much lower than it is today. This is further presumed to be not because spacecraft become cheaper, but because they become more productive. Today, $100 million buys you a booster good for one trip to orbit, carrying a few tons. In the PM it might buy a shuttle capable of hundreds of orbital missions, or a deep space ship making biennial Earth-Mars trips for decades.

The ships may still cost just as much, and building and operating them may likewise require an equally large industrial base. Note that the industrial base means much more than just the shipbuilding industry as such - it means the tools that build the tools that build the tools. There may be a time when spacecraft fabrication, as such, has largely moved into space, but still relies on Earth's vast and mature industry for its own most sophisticated components.

But could future techs drastically reduce the needed industrial scale, to the point where smaller communities - in the extreme case, individual households - could maintain themselves in space?

I am going to sidestep all pseudo-technical discussion of nanotech, 3-D printing, and all of that. Basically we are talking about replicators, where 'replicator' is really just techjargon for a compact super machine shop that can fabricate any desired item, including a copy of itself. Presumably all routine processes can be automated, so that the only human labor is setting up the job (the industrial equivalent of 'rules of engagement' decisions).

I can't think of any reason in principle why a midfuture tech couldn't build that capability on a pretty small scale, whether it fits in a backpack or a Winnebago.

It will not abolish costs, because its existence creates an opportunity cost: It can only make one thing at a time, so you have to choose. And for a portable home replicator to make a duplicate of itself may take quite a long time. My computer has far more speed and power than a 1960s mainframe, but it chugs away for hours on one 3-D render.

And replicators will not abolish economies of scale. Take making cars. The core replicator element might be able to make anything that will fit in its fab chamber - today a car, tomorrow a CAT scan machine. But if you dedicate it to making cars you can set up the shop floor to bring steel in one door and roll cars out the other. You can hire people who know about cars to do the detailed job specifications, so the AIs won't have to waste time on handholding.

Taken one at a time these advantages are incremental, but add them all up across supply chains and industries and they become overwhelming. Organized industries will continue to have an essential advantage over do-it-yourself, an advantage that will drive trade and economic life in general. This does not mean that people cannot live 'off the grid,' but doing so will take more work to maintain any given standard of living.


Now, to add one more chainsaw to juggle, the ability of communities to live independently in space is a matter of techlevel. A space population cannot sustain itself if the cost of keeping one person alive in space is more than one person-year of labor output. (A population can be sustained by Earth, of course, as the ISS is sustained now.)

In fact, for an economically independent society the cost of sustaining one person must be a good deal less than one person-year of output, because a society must support many people - most obviously children - who are not productive in any immediate economic sense.

Today it costs several hundred million a year to keep one person in space, on order of 10,000 person years of output. This cost can surely be reduced dramatically, let us say to 100 (current) person-years of output. Productivity has increased roughly tenfold in the 200 years of the Industrial Revolution; if it continues at the same pace, independently space-living populations become just barely viable in the 25th century, with most adults working to keep the hab going.

This still does not mean that your libertarian commune of 100 households can head off for the stars. There remains that little matter, or all too big a matter, of the industrial scale of space. If the industrial scale of technology - the advantages of scale - remain high, while all-round techlevel increases, it might take a mega-hab cluster of 100 million people to provide the range of skills and internal efficiencies needed to sustain itself in space.

To get small, economically independent and self-sustaining groups living in space you need both an increase in overall techlevel and a similarly dramatic reduction in the industrial scale required to support space operations.


For those of you who want one, which is a lot of you, here are a few escape hatches:

Most obvious and shameless, rich people, whose income is a lot more than average per capita productivity. But unless they are also making their money in space, this is merely a case of Earth subsidizing people in space, not an economically independent space population.

Another escape hatch, invoked in comment threads, is to make do with less. Science as a practice and profession is an outgrowth of Western monasticism. (Take that, mystical Eastern monks!) And there is a distinct ascetic streak in the space movement. Among the stars we can live at one with Nature, drawing on the essentials of energy and matter, unencumbered by smooth talkin' lawyers and fancy talkin' wimmin, or pretty talkin' gents as the case may be.

In practice, what doing with less - living closer to the productivity threshold - means is that people spend much of their time cleaning balky toilets, or fixing plumbing for more exotic but equally noisome fluids. That is what 'keeping spacecraft going' will be largely about, with the occasional call for a replacement part, whether you order it from stock or fab it onboard.

Faking it. Some people may seem to live independently on a small scale without really doing so, or only within narrow limits. Take one of my favorite tropes, the Serenity style space freighter. It is plausible to me that such a ship could keep going for quite a long time, perhaps many years, on just fuel and the most basic supplies. That is what it was designed to do.

Note that a ship like this probably cost more from the builder than a core zone ship that is designed for pull & replace servicing between runs. And eventually it will need to go into a cageworks for replacement or scrapping. But in the meanwhile it goes and goes.

Extend this concept a bit and you have a whole outer-fringe ecosystem that was costly to build in the first place, and would be costly to fully overhaul and restore to factory standard, but is relatively cheap to operate, and can be operated safely for decades or even generations before its service life is finally at the limit.

So this ecosystem, once built, can go for a long time as if it were independent of outside support - and can come, culturally, to take independence for granted. Though the bill will eventually come due.

The practical effect is much like Ken Burnside's 3-Gen rule, but the basis is entirely different. The 3-Gen rule argues that 'normal' human societies lack the social discipline to maintain something as complex as a hab. In my case the hab is not expected to replace or renew itself, only go a long time before either one is needed.


And there you have it. If you want small, independent, more or less self sustaining space habs in the midfuture, I have no Space Patrol, and no reason to send it to stop you if I did. Just be aware of the techlevel you are implying, 4-5 orders of magnitude improvement over what we have achieved in 50 years of space travel.

Space is hard. Even with enormous progress in both our overall capabilities and our specific space techniques it will still be hard.



This image of the ISS from Astronomy Picture of the Day deserved a reprise.

77 comments:

Milo said...

Rick:

"But could future techs drastically reduce the needed industrial scale, to the point where smaller communities - in the extreme case, individual households - could maintain themselves in space?"

I would bet against the individual households, barring magitech (like very cheap FTL drives that can be installed into a simple sealed tank, or nanotech seeds that can grow a spaceship with no human input if you just plob them down somewhere with sunlight and some good ores in the soil). Even Space Viking scenarios assume the barbarians have tribes at the least.

If individual households do manage to own spaceships, they will be able to do so only due the support of economic infrastructure beyond the household itself (just like most people don't know how to build cars themselves, let alone how to build the tools that build the tools that build the tools to build cars). However, this can still manage to support some barbarism, if the engineers are sufficiently no-questions-asked about who purchases their services. Even Vikings had specialist blacksmiths...


"It will not abolish costs, because its existence creates an opportunity cost: It can only make one thing at a time, so you have to choose."

So the question is: is it a viable tactic to choose to have the single thing you make be a rocket, and then use that rocket to steal all the other things you need?


"The 3-Gen rule argues that 'normal' human societies lack the social discipline to maintain something as complex as a hab."

The Netherlands would like to quibble with your suggestion that it is impossible to keep terraformed regions stable for centuries.

Well, mostly stable. Occasional flooding disasters happen, and chances are in a setting with many dome cities occasional decompression disasters will happen as well.

Thucydides said...

While space exploration, exploitation and eventual colonization starts from a mighty springboard, I have the feeling that the people out there will be able to get away with a lot less, use local resources and strip down and simplify as much as possible.

Hernán Cortés, when faced with a humiliating setback after being expelled from the Aztec capital, rather than go home and restock, he used available materials, including local supplies of sulphur and potassium nitrate, wood (for boats) and local labour to restock his weaponry and renew the assault from a different direction.

Western settlers in both Canada and the United States (as well as the earlier Voyagers in the sub arctic made do with very little compared to the standards of settlers and citizens in the cities and along the coastal regions.

Space roughnecks might go from high tech photovoltaics which need to be made in a fab to solar thermal generators which can be made from salvaged parts and large mirrors in extreme cases, or at any rate without the expense of a fab capable of making high purity silicon wafers. Similar arguments can be made about screws and other fasteners; energy is cheap and welding or heat sealing takes advantage of all this.

Transportation costs inside a colony or habitat can be very low in theory; everything you consume becomes feedstock for everything you purchase, and only needs to move a few kilometres at most to go from you to the replicator and back again (assuming some sort of centralized system or industrial district to take advantage of economies of scale).

So to reduce costs and mass, colonists will attempt to use as little mass as possible, and to switch to systems they can make and maintain themselves. The Island Two (Bernal sphere) configuration seems to meet many of the needs due to the relatively small size and compact layout. Solar thermal spacecraft might become the standard in the inner system due to the inexpensive nature of the rocket and relative simplicity for the operator. Maybe the abacus will make a comeback compared to heavy calculators, cash registers and other appliances.

Bryan said...

It will not abolish costs, because its existence creates an opportunity cost: It can only make one thing at a time, so you have to choose. And for a portable home replicator to make a duplicate of itself may take quite a long time.

But self-replication opens the door to exponential growth. A very simple way of setting up a colony may be to simply send a replicator and some sort of mining/processing robots to the desired site. Replicator copies itself, using mined materials. New copy makes a second set of mining/processing robots, and then uses the materials they provides to replicate itself.

Repeat ad nauseum until you have a fleet of replcators, supported by a fleet of mining/processing robots. Then turn the industrial capacity to building the colony. The colonists get a home, and an industrial base, all in one.

FWIW, there is an existing project to try and make self-replicating rapid prototypers:
http://en.wikipedia.org/wiki/RepRap_Project

So far it can only replicate its mechanical parts, but it is hoped in a few generations it may be able to replicate everything - electroniccs, motors, etc.

jnutley said...

The post bases it's case on an apples (space colonization) to oranges (first of their kind developmental space craft) comparison. Also, implying that the whole world is required to perform such a task is misleading.

The appropriate bases for such an estimate are a) A spacecraft system in the mode of sustained operation and b) A national civilian transportation network. Compare the percentage of GDP used for both activities, NOT the total population, to estimate the level of comitment necessary for space colonization.

The only historical spacecraft that qualifies for part (a) is Soyuz T and TM which flew between 1976 and 2003. I have no idea where to find their percent of the Soviet/Russian budget. But that percent, divided by the GDP of the Soviet/Russian nation during those years will give you man-year level of commitment to maintain a working man-rated space system.

For part (b) we can look at the car in the United States, starting after the last interstate is in place, in 1962, or Japan's rail system after it has a certain percentage of the nation and cities served. This time consider only the percentage of GDP applied to these nets. Include the cost of road maintenance and car purchases or rail maintenance and tickets purchased by the population. This is a first order estimate of the comfortable level of commitment that a technical society can place on an overarching infastructure.

Space colony infrastructure will cover the physical plant of the colony. It will consume a percentage of the GDP similar to the examples in (b). The cost of operating and maintaining the life support system will have at least some overlap with the general catagory of agriculture, so it should not, in total, be applied to our estimate. The first colony will need a sherrif and some paramedic/firefighter types, but no standing military. A variety of things that states on Earth must fund, the first colony will have no need of, nor will the second or third initially need to consider such costs.

The utility of (a) is that it will help us determine the minimum number of people in a sustainable (for tens of human generations) colony. A calculus that takes (a) and applies it as an estimate of (b) will be a good first guess. The hope of "rocketeers" like me is that SpaceX or it's Indian shadow will show that the absolute cost of (a) is even smaller. The scale of the first colony can then be set at a clan or confederation of tribes level of population. Hundreds, not Billions, and within the reach of a near-term future prince of Davos to achieve.

jnutley said...
This comment has been removed by the author.
Raymond said...

I wish to subscribe to this newsletter.

Rick said...

If you mean subscribe to this blog, hit the Posts (Atom) link at the bottom of the main page. I think - I never subscribe to RSS feeds, so I don't know how all that works.

Yes, there is a big difference between the semi-experimental human spaceflight of today and what it would be like as a regular operational activity. That is part of what I had in mind in saying that the real cost of space travel can fall by a couple of orders of magnitude, even apart from overall tech progress.

I am also not persuaded that the analogy to surface transport systems fully holds. The Netherlands could, on a smaller scale, build rail and highway systems of similar performance, and no doubt has. But you can't scale down a space effort much before you hit granularity: It takes a big economy to develop big boosters.


The Netherlands, by the way, has no beef with me - I don't agree with Ken's 3-Gen premise. If your techlevel is high enough, I don't see why a hab couldn't be sustainable. You just need a Really High techlevel.


Did pioneers of various sorts really do with that much less? Compared to the upper classes back home, certainly, but compared to ordinary peasants and workmen? At almost any time in the agrarian age, cheap land outweighed expensive tools.

Tony said...

Milo:

"The Netherlands would like to quibble with your suggestion that it is impossible to keep terraformed regions stable for centuries."

The Netherlands are hardly terraformed by SF standards. The atmospheric composition nor the insolation were changed by anything the Netherlanders did.

As for social stability and its affect on massive land reclamation, one should note that the Netherlanders were extremely pessimistic about this. They created structures and systems that would degrade gracefully and which could go for years -- and in some instances decades -- without any meaningful maintenance.

Raymond said...

Rick:

Oh, that was just so I can get email updates for this thread, too. Blogger's annoying like that.

In general:

How much of this is about required industrial complexity, and how much is simply allocation? Aside from a few communication and navigation systems (which are largely already in place), spaceflight is a marginal activity. And as long as we lack a significant population offworld, it will continue to be so. If we find a McGuffin to justify it (or say screw it and do it anyways), and get our big spacer population, it would probably increase the proportion of the economy devoted to it. (How'd that be for a make-work project?)

As far as any offworlders are concerned, spaceflight would have an economic importance comparable to finance and automotive combined. How does that change the numbers?

Geoffrey S H said...

I'm going to play devil's advocate for once (and not post a ridiculously fantastic suggestion), and perhaps suggest that the ISS really is the furtherst we will go for a long time. If we don't get any rate of increase in launches and interest in space now, then might that mean that we are stuck with what we have for centuries? Might even a space craft capable of leaving earth orbit and perhaps going to a lagrange point etc not be constructed until passing of hundreds, even (I'm utterly serious here) thousands of years? The rate of technological increase seems to be confinedto very small scale techlologies and vehicles- to take the plunge, I'd say they won't have a serious effect on the space industry for a very, very long time...

@ Byron:

Very interesting, good to see a clanking relicator in action. Nevertheless it does seem far far to good to be true... I'll search for someindependant sources to clarify this, but I admit I did have to restrain myself from instantly dismissing it as fantastical and vapourware. Apart from the improvements described on wiki, I wonder how much such a replcator can be impoved, before molecular scale units come onto the market...

KraKon said...

Book and mark.

KraKon said...

There.

Bryan said...

@ Byron:

Very interesting, good to see a clanking relicator in action. Nevertheless it does seem far far to good to be true... I'll search for someindependant sources to clarify this, but I admit I did have to restrain myself from instantly dismissing it as fantastical and vapourware


Its bryan, not byron....

And reprap is very much a real thing - I have one sitting on a lab bench, right next to a commercial prototyper. As I said before, it can replicate all of its mechanical bits, although electronic parts currently remain beyond its capacity.

I wonder how much such a replcator can be impoved, before molecular scale units come onto the market

But is a molecular-scale replicator sufficient. Atom-by-atom assembly would be slow for large items. I'd envision that any plausable system like I describe would require both, all in one package. A large-scale device for large parts, a molecular-scale one for electornics and fine parts.

Anonymous said...

Ok, I'm going to take a stab at it. If you only concentrate on building and operating a general transport-type rocket ship to help suport the colony: they don't waste time, energy, or capital on anything that does not directly support the production, operating, and maintaining of said spacecraft and the colony and colonists themselves; that means no real stores or resturants, or any luxuries; the colonists all wear jumpsuits, eat at cafeterias,don't have movie theaters, etc... Under these circumstanses then you could get away with about 10,000 people(maybe a bit more), but only if the colonists accepted a couple of generations of hardship.

However, if you want to maintain a standard of living higher than North Korea, than I'd say you'd need a somewhat larger population; namely, about 100,000. That gives the colonists the people needed for operating the industrial base (including mining), the standard-of-living support, the colony infrastructure, administration, etc. That is a small industrial city; I'll use the city just south of the one I live in, Pueblo. This is a good example, due to the fact that is has a steel industry and a rocket factory. Pueblo is big enough to have all the things that are needed to keep the population's standard of living fairly high, and yet small enough that is still has some of the feeling of living in a small town. Pueblo has a university, is close to agrecultural areas, as well as iron mines, and has it's own water and power sources. Move it to Mars and put it under a dome, and Pueblo would make a pretty good analogue for a typical space colony. I think that the answer to Rick's question lies somewhere between 10,000 and 100,000 with the industrial core, resource gathering, and agracultural capability of our example city of Pueblo.
It's big, but not impossibly so. I can see a scientific outpost growing, by normal expansion (and maybe a little immigration), until it could reach the level where they could support one or more domestically built cargo rocket to boost their economy and/or population.

Ferrell

Jim Baerg said...

The issues of industrial scale & the time scale for interplanetary travel make me sceptical of proposals like Zubrin's to colonize Mars while mostly ignoring the moon.

Some McGuffinite that makes development of the moon, cis-lunar space & maybe near earth asteroids profitable, puts space development where spare parts can be delivered from earth in a few days if necessary. Plausible launch costs from earth will still be high so space industry would start with extracting or making the McGuffinite that pays the bills plus material like propellant that is needed in space in large quantities at low prices so local manufacture can compete with lauch from Earth.

The McGuffinite that appeals to me as relatively plausible is extracting platinum group elements metallic meteorites fragments in the lunar regolith or from near earth asteroids. Most other elements are at least as readily available on the earth's surface as in space.

Milo said...

Bryan:

"But self-replication opens the door to exponential growth."

If you have sufficient raw materials and energy. Human population would in theory grow exponentially too, but in practice we just grow until the planet is filled up.

So the question is what a given colony's "carrying capacity" for replicators is, given a certain techlevel and size of skilled population. It can then be assumed the number of available replicators will grow fairly abruptly to reach the limit of this carrying capacity.



Geoffrey S H:

"Might even a space craft capable of leaving earth orbit and perhaps going to a lagrange point etc not be constructed until passing of hundreds, even (I'm utterly serious here) thousands of years?"

Are we talking about manned or unmanned here? Because we've sent unmanned probes to basically anywhere in the solar system already.



Bryan:

"I'd envision that any plausable system like I describe would require both, all in one package. A large-scale device for large parts, a molecular-scale one for electornics and fine parts."

That's pretty reasonable. Use a hammer when you need a hammer and tweezers when you need tweezers.

However, I wonder if molecular-scale assemblers would actually use nanobots (as people seem to like suggesting), or would use tactics like laser deposition to perform fine manipulation from a larger working platform. (This is called nanolithography, apparantly.)

Anonymous said...

For planetary colonies, whether in the solar system or beyond, it's probably necessary to differentiate between being able to build (preferably reusable) surface-to-orbit shuttles and being able to build interplanetary vessels. The former might be easier to build, since although they would need high thrust while the interplanetary craft might not, they would not need as much long-term habitability. The ability to build shuttles would probably be more immediately useful, since it would mean that the interplanetary craft visiting from elsewhere would no longer have to carry their own shuttles to transfer passengers and/or cargo to or from the surface, which would open up more opportunities for commerce.
For space habitats in the solar system, they will presumably start off dependent upon Earth for everything except perhaps electricity and food, but as the number of habitats increases and habitat populations grow, they will be able to do more and more for themselves until they become practically self-sufficient.

R.C.

stoffer said...

Why does everyone think that Viking were savages? Vikings were highly skilled in metallurgy, shipbuilding and seamanship. Their longships are still masterpieces of naval engineering. They were built light, strong, flexible and fast. The building technique was optimized to get the most of the wood used.
Also, Vikings had state structures, with kings, armies and administration. Denmark was once a Viking empire covering preset day Denmark, north of Germany, Skåne (southern Sweden), large parts of present day Norway and parts of the shore of the Baltic sea. They were not more barbarian than any other ethnic group of the early medieval Europe.

Coming back to future economy, I think that corporation-states and communist states are much more feasible than some romantic barbarians. Communist and corporate states have the necessary level of discipline to survive in the harsh conditions of outer space. Actually I think that communist states are even more probable than corporation-states. Projects like planet terraforming need a long time to complete and corporations need to show monthly profit growth to their investors. Communist states are free from that.

There might be some barbarians living from the scrap of a fallen empire, but as the example of Soviet Union shows, most of the power and wealth from the fallen empire will fall into the hands of oligarchs.

Anonymous said...

Three billion people of Earth in 1969. did not all contribute to the Apollo program. The best real results we could get by looking at the USSR in those years, because it was largely self-suficient (autarhic). Even then needed number of people to build and operate spacecraft would actualy be smaller because of growing tech level, economic return from space idustry, and especialy because USSR was very ineficient and labour intensive in agriculture and similar sectors.
English is not my mother tongue so sorry for possible mistakes.

Tony said...

"Even then needed number of people to build and operate spacecraft would actualy be smaller because of growing tech level..."

Setting the relative inefficiency of Soviet agriculture and (some, not all) industry aside, the increases in technological sophistication in some Soviet centers of excellence was a direct result of the overall size of the economy. Unless you inherit, buy, or steal your technology -- in which case you can't sustain it -- your potential for technolgical sophistication is a direct function of your GDP.

Raymond said...

I think "direct function of your GDP" is an oversimplification. Smaller states can punch well above their weight class in terms of raw technological knowledge (see Canada, Sweden, Korea) if they have access to sufficiently large markets for their high-tech industries. Larger countries can possess higher technology than is currently generally used within their borders when hobbled by legacy infrastructure (see America with regards to telecom). Somewhat smaller economies with largeish populations can produce current-generation tech while most of their population is at Third-World levels (see Brazil, specifically Embraer). It's much more complicated and tangled of an analysis than a simple formula with GDP as the only variable.

Anonymous said...

I tried to relate isolated economy(USSR) to the isolation of a colony on a distant world, note that none of the countryes you report is isolated in such a manner. In plausible midfuture the inability to transport large amounts of goods will also isolate economy in many other ways. I agree it is not depended soley on GDP.

Rick said...

Welcome to a couple more new commenters!

perhaps suggest that the ISS really is the furthest we will go for a long time

Meaning human spaceflight, this strikes me as entirely plausible. At this point the only real reason for humans to go into space is because it is cool, and to study living and working in space.

Humans in space are not needed for exploration, where as noted we have already reached every planet, and have landed on Venus, the Moon, Mars, and Titan. (And entered the atmosphere of Jupiter.) Humans are not needed to exploit Earth orbital space in the ways we are already exploiting it - Clarke assumed an orbital comm relay would need a crew just to replace vacuum tubes.

It is not even a given that humans would be needed to mine McGuffinite, and anyone out to make a buck by doing it would certainly try to avoid the expense of deep space life support.

So space may turn out to be a place where we go, but not in person.


The three billion people of 1969 Earth all contributed, in some degree, to Earth's economy. And remember that I was setting an upper bound there; my figure for the lower bound is on order of 100 million.

Smaller states can punch well above their weight class in terms of raw technological knowledge (see Canada, Sweden, Korea) if they have access to sufficiently large markets for their high-tech industries.

But that's the rub - they function in a larger world economy.

Now imagine the colony planets of New Sweden and Nouvelle Canada, with similar populations to their current day mother countries. On the one hand they have (we will stipulate) a whole planet's worth of good farmland and raw materials.

On the other hand their only external trade is whatever can pay starship freight, and their total imports probably wouldn't fill one present day cargo ship. They aren't importing much heavy machinery for their factories.

So the question is, how big must an isolated colony planet population be to maintain the sort of complex, comprehensive industrial base that can operate spacecraft?

And as RC said, woe to the colony planet that cannot maintain a surface to orbit capability. If a trade starship needs to carry her own shuttle, that is a big chunk of cargo capacity lost and a big expense added.

You could have the interesting scenario of a colony building expendable boosters to reach orbiting trade ships, because their economy and the traffic volume don't justify the extra cost of building and maintaining a reusable shuttle.


The colony of New Pueblo is orders of magnitude smaller than New Canada. I think they might be able to provide routine servicing and maintence for a shuttle, but replacing it with local resources would be a much harder task.


No one has called the Vikings savages, though early medieval Christendom called them worse. For discussion on this blog I've called them barbarians, a rather different thing. In the popular culture they are nearly the barbarians par excellence.

In another sense they are 'barbarians' in being people who lived on the fringe of an established civilization, were connected with it without being fully part of it, and raided the bejeezus out of it.

By curious historical coincidence, the last two serious Viking invasions of England happened in the same year. The one from Norway failed, the one from Normandy succeeded. (One result is that Scandinavia falls off my personal, Anglocentric historical radar after 1066.)

On your future point, interesting observations! But Russian oligarchs might fit into a barbarian-warlord trope rather nicely. The Russians themselves have adopted the word 'mafiya,' which I imagine they got far more from Hollywood than from Sicily.

And in the context of a collapsed interstellar bubble, the Tsar is far away.

Rick said...

Just a note that even I am not immune from Blogger tricks. It said my last post was too long, but posted it anyway.

So if you get that warning, my advice is to open the comments in a new browser tab and see if your post got posted.

Tony said...

Raymond:

"I think "direct function of your GDP" is an oversimplification. Smaller states can punch well above their weight class in terms of raw technological knowledge (see Canada, Sweden, Korea) if they have access to sufficiently large markets for their high-tech industries. Larger countries can possess higher technology than is currently generally used within their borders when hobbled by legacy infrastructure (see America with regards to telecom). Somewhat smaller economies with largeish populations can produce current-generation tech while most of their population is at Third-World levels (see Brazil, specifically Embraer). It's much more complicated and tangled of an analysis than a simple formula with GDP as the only variable."

Maybe it's a bit of an oversimplification for the general case. But with spaceflight we're talking about either a large enough market or a big enough government program to pay all the bills. Without those, the tech level for spaceflight is simply not going to be there. Even in cases of more portable, more generally applicable technologies, it takes large economies to make them go at the higher levels. Canadian/Swedish/Korean/Finnish expertise is not funded if the markets don't exist for their wares. Likewise, emerging industrial states buy up already developed expertise from large markets. Brazil (and hispanophone South America), India, Arabia, and large parts of East Asia buy all (or major parts, in the case of India, China, and Japan) of their space access in the three established space access markets (US, Russia, EU) with very minimal marginal participation of other space access providers, including their own national programs (where they exist). What's the common denominator of these markets? Large indigenous high tech industries supported by large GDPs.

Raymond said...

Rick:

Nouvelle Canada, if it sufficiently resembles the original, would have mature industries of aerospace, nuclear power, conventional power, mining, refining, heavy construction, rail transportation, robotics, industrial agriculture, hydroponics, telecommunications, computing, processor design and fabrication, film, music, theatre, and burgeoning expertise in nanotechnology, biotechnology, solar and wind power, and even quantum computing. Plus, we're used to large, unpopulated hinterlands and strings of cities tenuously connected at long distances, and we have one of the better educational infrastructures in the world, quite capable of maintaining expertise in the relevant areas.

Did I leave anything relevant out?

Tony:

That analysis may hold for the present milieu, given the marginal nature of space travel at present. If we're talking about a space colony, a much higher portion of GDP will likely be allocated to space access, and less to consumer goods or exporting resources. And as far as technical knowledge, it can be shared (or bought) at interplanetary distances. The key requirement is not developing all your technology yourself from scratch, but having the educational resources to integrate and utilize the technology available (which is achievable at much lower populations).

Rick said...

Canada has a population of 34 million people, which on a broad order of magnitude scale is not far below my 100 million threshold.

I should maybe clarify that I am not claiming great precision for this figure. The question isn't something that lends itself to precision. But I think for example that a for colony of one million a native launch capability would be out of reach.

Milo said...

Stoffer:

"Coming back to future economy, I think that corporation-states and communist states are much more feasible than some romantic barbarians. Communist and corporate states have the necessary level of discipline to survive in the harsh conditions of outer space. Actually I think that communist states are even more probable than corporation-states. Projects like planet terraforming need a long time to complete and corporations need to show monthly profit growth to their investors. Communist states are free from that."

So you're saying that decades after the collapse of the Soviet Union, they'll still have the last laugh in the Space Race?



Rick:

"Humans in space are not needed for exploration,"

Not needed, no. But do you think NASA doesn't want to use human exploration, if they could? Just look at how much headache the Spirit rover is causing over a problem that could easily be solved by a human researcher walking over and giving it a yank. Look at how scientists are poring over readings to guess whether the Huygens probe landed in muddy clay, or displaced a pebble. A human researcher would surely know what he's standing in! And would be able to continue making reports for longer than 90 minutes.

We're using robots because humans (or rather the craft to carry them) are prohibitively expensive. If we could afford human exploration, we would, because humans have an important feature robots lack: they're their own fault tolerance system. Humans are much better than robots at improvising in unexpected situations. And what is research if not going and seeking out unexpected situations?


"Now imagine the colony planets of New Sweden and Nouvelle Canada,"

I am fairly certain there are more people with English as their first language in Canada than Sweden.


"And as RC said, woe to the colony planet that cannot maintain a surface to orbit capability. If a trade starship needs to carry her own shuttle, that is a big chunk of cargo capacity lost and a big expense added."

Note that you don't necessarily need the capacity to build shuttles. If you have the capacity to dock, launch, and refuel shuttles, then that is sufficient for a young colony, even if the craft themselves (or their most sensitive components) are imported.

This is, of course, assuming that you do not expect your shuttles to get shot at much.


"By curious historical coincidence, the last two serious Viking invasions of England happened in the same year. The one from Norway failed, the one from Normandy succeeded."

I consider the Normen to be ex-Vikings. They were too deeply assimilated into French culture, and they came to England as unified conquerors in the name of their duke, not as swarms of small raiding parties. (Of course, by this measure, the Mongols under Genghis Khan wouldn't quite count either...)

Raymond said...

Milo:

The francophone population would bitch and moan and threaten to leave to form Nouvelle Quebec. The name of the colony would be a concession to keep them grudgingly within the fold (along with promises to locate a certain amount of the rocket-building in the Montreal Sector).

The anglophones would just roll their eyes and go along with it, then just call them some very English names behind their backs.

Thucydides said...

I would have thought the very visible example of the implosion and collapse of the USSR and all it's satellite empire or the current condition of the DPRK would have cleared everyone's head of the notion that a communist, fascist, socialist or other command economy variation would succeed in space when none have on Earth.

Command economies can devote massive amounts of resources to whatever the ruler(s) deem important; there is nothing to stop this until the resources run out, but misallocation of resources, the inability to solve or even deal with the "local knowledge problem" and a brittle social structure with limited ability to deal with unanticipated situations all conspire against long term success.

Now if you treat a colony or base as a military installation, you still have the issues of resource allocation, and military facilities do not operate in an autonomous fashion from the larger economy. (Ones that do have two problems; they are devoting their internal resources on becoming self sufficient, reducing their utility as a military installation, and they are positioning themselves to be in revolt from their civilian masters. Future Space Navies or Aerospace Forces might have a commissar position to prevent just that since by definition space forces must be self sufficient).

Looking at the problem from the market, the issue for any colony is to minimize costs. Since importation is going to be extremely expensive, only the most critical high value items will have a market. Minimizing opportunity costs will be next on the list, your replicator cannot be tied up on any long term projects. Creating rolls of aluminum film might be the priority task since it can be used to form mirrors to concentrate thermal energy on ore to bake, boil or vapourize valuable materials out of the raw ore, generate electrical energy with a thermal generator
(or MHD if you have a really big mirror) or heat remass for local space propulsion, as well as any other structural uses aluminum film might have in space (multiple layers of aluminum foil like material made up some of the structure of the LEM, for example).

Since labour and man made resources will be in short supply, there will be a space industrial revolution as people learn to adapt to these conditions.

Tony said...

Raymond:

"That analysis may hold for the present milieu, given the marginal nature of space travel at present. If we're talking about a space colony, a much higher portion of GDP will likely be allocated to space access, and less to consumer goods or exporting resources. And as far as technical knowledge, it can be shared (or bought) at interplanetary distances. The key requirement is not developing all your technology yourself from scratch, but having the educational resources to integrate and utilize the technology available (which is achievable at much lower populations)."

Even if you have Great Patriotic War Soviet Union levels of societal regimentation, I don't think you can possibly muster the range of skills necessary to keep a space civilization running with very much under several millions, perhaps tens of millions, of people. In order to have enough economic degrees of freedom and surplus to maintain all of the necessary skillsets, you need a large base of, for lack of a better term, mundane professions and people.*

*Which is why I don't look down my nose at what my fannish friends call "mundanes". If you do useful work, no matter how routine and and seemingly simple, you're an important person, IMO.

Milo said...

Thucydides:

"by definition space forces must be self sufficient"

No more so than the old kind of navy. Nuclear submarines can stay out for months at a time, but they are by no means self-sufficient.

Thucydides said...

Except for orbital forces or ASAT batteries, space forces must be self sufficient to a much greater degree than any previous military forces.

Perhaps the best analogy would be the age of sail, where ships could be gone and out of communication for years at a time, and did have the option of pulling into even remote headlands to collect wood and other materials to make some repairs (think of a SN vessel pulling up to an uninhabited NEO to extract water for remass, or more futuristic ones ramscooping the atmosphere of a gas giant planet). This implies some sort of ability to process materials on board the warship, something that most modern ships do not have. To prevent contamination of the computer networks running the ship, there may be an embargo on exterior communications, or at least a severe limit to what can be received and sent.

The ship, laser/kinetic star or constellation would eventually need to come into some docking station for repair and refit, but the duration of a patrol would be measured in years unless torch drive or some other magitech solution was available.

Tony said...

Thucydides:

"Except for orbital forces or ASAT batteries, space forces must be self sufficient to a much greater degree than any previous military forces.

Perhaps the best analogy would be the age of sail...

The ship, laser/kinetic star or constellation would eventually need to come into some docking station for repair and refit, but the duration of a patrol would be measured in years unless torch drive or some other magitech solution was available."


If we're talking about the plausible midfuture, without magitech. Then there won't be patrols or unpredictable (beyond a reasonable contingency allowance) mission durations. There won't be patrols, because patrols are an artifact of horizons. Those don't exist in space. You simply watch the other guy's installations and send your forces out when it makes sense for you to do so. Those forces have specific missions and enough delta-v and other consumables to do those missions. Dinking around looking to extract minerals and volatiles from asteroids just uses up men and equipment, and probably puts them out of position when they're needed.

Milo said...

Thucydides:

"To prevent contamination of the computer networks running the ship, there may be an embargo on exterior communications, or at least a severe limit to what can be received and sent."

Networks in space are going to be at no more risk from "contamination" than networks today, and militaries still have sophisticated communication. Seriously, you're reaching.

Communication across interplanetary distances will be hampered by lightspeed lag and by occasional occultations, but this still allows you to communicate to even the solar system's far reaches in hours - far better than anyone had during the Age of Sail.



Tony:

"There won't be patrols, because patrols are an artifact of horizons. Those don't exist in space. You simply watch the other guy's installations and send your forces out when it makes sense for you to do so."

The only kind of "patrol" that makes sense in space is deploying a ship into orbit around an enemy's planet or moon as a spy satellite. Since this requires you to be able to hold at least high orbit against the enemy's defenses, I see it as more of a siege action than a patrol action.

There are no deep space patrols.

Albert said...

I think in the blog post Rick makes a little confusion between the cost of "going where no man has gone before" and the cost of "going where lots of others have gone before".

The first case requires huge amounts of research, cadres of reserachers specialized in multiple disciplines (loads of ancillary population to support them), nation-busting amounts of money, and people specialized in building/synthetising things (chemists, engineers and mechanics) to have the final product.
That's Apollo situation.
You cannot say Apollo was an example of "maintaining the same techlevel" in spacecraft design.
It was a "breaking all records" operation, and should remain out of this discussion.

The second case requires just enough expertise to keep building the same machinery others already designed.
So you will need only the people specialized in building such stuff.
Engineers, chemists, technicians.
There will still be a little tinkering, but much more conservative and limited, since those are builders, not researchers.

This will mean that while a colony may very well be able to maintain itself indefinetly, it will most certainly lack the people and the resources to advance technologically enough to improve itself or its spacecraft to any significant degree.

Every isolated colony will tend to be a "still" of the techlevel of the parent nation at the moment of launch/foundation.

Since I like trading, I think this may be a good reason for trading.
If each colony tries to specialize a little its very limited research budget (people and resources) to one or two topics at most, they can then sell their know-how (like training engineers/researchers from other colonies) or finished products.

Each colony will still mantain enough people and knowledge to keep its current designs and structures operational indefinetly (eventually building copies when the old ones reach the end of service life), but if it wants something better either asks Mommy (Earth) or trades with other colonies that are specialized in that topic.

I think this applied to tribes of "barbarians" of the old times too. They specialized in something and traded their excellence with other's excellence.

Also a reason why Earth will be able to pwn any and all colonies for a long time, far beyond PM.

"Depart NOW from a green open round land with a blue sky and a sea, where science advances by big leaps! You will live your whole life closed in a smelly rotating tin can where you will struggle to keep everything online and make sure your offspring has the right skills so your whole settlement doesn't die horribly! You will be seen as backward at best or primitive at worst in a few decades! Even by joining efforts with the whole colonial community you won't be able to keep up with the scientific advancement of Earth people!"

Sigh.

Tony said...
There won't be patrols, because patrols are an artifact of horizons. Those don't exist in space. Well, patrolling also shortens reaction time. And when even getting to Mars takes half an year, you would like to have your forces get there in a few months.
And when enemy knows that your reaction time is smaller will have to act faster or don't act at all, and this is generally something you want.

-Albert

Rick said...

No, Apollo certainly wasn't 'maintaining the same techlevel!' But I only mentioned it in the context of Earth's population in 1969 as an upper bound.

My 100 million population estimate is based on current launch capabilities around the world. Japan has a native launch capability; France had to partner in the ESA.

If space capability were a national priority comparable to defense, an economy of about 10 million could probably do it, and that might be a fair threshold figure for a colony planet.

Also I'm talking here about broad spectrum capability - the ability of an economy to design and build its own space boosters to the current state of the art. There is a whole range of more limited capabilities down to providing basic ground servicing.


I tend to agree that on an interplanetary scale there is no reason for patrols. Peacetime flag showing is a different matter.

Deep space military operations in general will be rigid, because orbits and delta v are so constrained.

Byron said...

Its bryan, not byron....
That made me laugh. I get Bryan all the time. It's the first time it's ever happened to someone else because of me.

Replicators will still take human operators. Even the most automated is still going to require human engineers and mechanics. I've done enough CNC to know that the first is a lot harder than it looks.

Space navies also will have the issue of mass. If you wish to carry processing equipment, that's less of something else to carry. And that something is likely to be very valuable. Such as a bigger engine to get there faster. And the computer problem is simple. Set up a separate system to handle communications, not connected to the rest of the ship.

Byron said...

Didn't we already have the whole "space patrol" discussion? Because of how empty space is, and how long the transit times are, and how few the valuable places are (planets, etc) if people want a presence, they'll send a permanant station. They won't send a cruiser by Mars every year to check up on it. It ties up too many ships. Set up a few corvettes at Mars, and keep a cruiser on standby.

Jim Baerg said...

"Every isolated colony will tend to be a "still" of the techlevel of the parent nation at the moment of launch/foundation."

Actually, for the 1st few decades the colony will be Schizo Tech mix of whatever high technologies are easy to bring on the colony ship & lower technologies that are easy to duplicate on the new world.

As the colonies population increases it will be able to duplicate more of the latest tech available on other worlds & eventually start its own R&D to advance technology.

Even an intestellar colony founded by STL travel will be able to build radiotelescopes to maintain communication with the home world(s), so it will have the knowledge of the latest tech, but gradually build up the ability to duplicate it.

If instead you are talking about orbiting habitats in cis-lunar space, they will import all the lightweight high tech stuff they can afford, with the possible exception of high tech items which zero gee &/or extreme vacuum make easier to manufacture.

Milo said...

Albert:

"I think in the blog post Rick makes a little confusion between the cost of "going where no man has gone before" and the cost of "going where lots of others have gone before"."

A very good point.


"Every isolated colony will tend to be a "still" of the techlevel of the parent nation at the moment of launch/foundation.

Since I like trading, I think this may be a good reason for trading."


Unfortunately, all you really need to do is to trade information at light speed. There's no reason for anyone to actually hop into an expensive ship.


"Also a reason why Earth will be able to pwn any and all colonies for a long time, far beyond PM."

That depends on how fast those colonies can grow.

Population isn't really an issue beyond a couple centuries, since humans can breed like rabbits when we want to. A bigger issue is how quickly you can expand the life support infrastructure of a dome city, and if you have the resources to do so.


"Well, patrolling also shortens reaction time."

Already being in orbit around the enemy planet shortens reaction time even more.

Still being in orbit around your own planet also shortens reaction time by still having full fuel tanks rather than having wasted it on some meaningless notion of "patrolling" that doesn't even work in rocket mechanics.

Milo said...

Rick:

"My 100 million population estimate is based on current launch capabilities around the world. Japan has a native launch capability; France had to partner in the ESA."

So, just a question...: had to? They thought it would be better to, but are you sure they actually couldn't have done it on their own? Partnerships can often be very useful even when they're not strictly necessary.


Byron:

"They won't send a cruiser by Mars every year to check up on it. It ties up too many ships. Set up a few corvettes at Mars, and keep a cruiser on standby."

That depends on if they can resupply at Mars. If Mars can't or won't resupply them, then you need to send regular supply craft from home, which isn't going to be much easier than just sending the ships themselves back and forth (plus there's a limit on how long your crews will be willing to stay in the ship - unless they spend their downtime mingling with locals, which may threaten their loyalty).

If you actually have enough of an interest in Mars to build your own base with all the necessary services to keep your ships running, then sure. You might not even need the ships anymore at this point. But now you're pretty much building your own colony.

Thucydides said...

Interestingly enough for fans of 3d printing and replicators, there is this:

http://nextbigfuture.com/2010/11/additive-manufacturing-company.html#more

Entire cars and aircraft subassemblies created by the 3d printing process speaks of a certain level of sophistication, and perhaps the time/cost advantages of assembly line mass manufacturing might not be quite as great as has been assumed, at least in some cases.

Tony said...

Thucydides:

"Entire cars and aircraft subassemblies created by the 3d printing process speaks of a certain level of sophistication, and perhaps the time/cost advantages of assembly line mass manufacturing might not be quite as great as has been assumed, at least in some cases."

It's not the entire car, it's the body. The aircraft parts under discussion are prop blades and wings, both of which are textbook examples of low volume, high precision parts.

In the case of the car body, I'm see this as more of a (relatively) low cost-of-entry play, rather than a radical new manufacturing method. If you have the infrastructure of a high volume automobile manufacture, it's hard to see how body parts are more efficiently created by deposition than they would be by stamping or molding.

Likewise, in the case of wings and propeller blades, this is just an alternate way of doing something that is already slow and expensive.

Rick said...

Well, the French 'had to' in the sense that they chose to, at the cost of some bragging rights, which have been a significant factor in the history of space programs.

But on the lines you mention, I think the ISS is not just post cold war politics, but an implicit acknowledgment all around that maintaining a space station is so difficult that it is better to do as a combined effort.

Having said that, a smaller economy could probably do it if it were a priority equivalent to national defense, perhaps 10 million - the equivalent of one metropolitan area and its region.


On manufacturing, even if there is a dramatic change in manufacturing tech, I don't think it puts an end to factories. Even if you are printing up cars, you will optimize the printer heads for each part of the job, and organize the work flow.

When the subject is replicators, I think people tend to picture a technology that not only radically reduces industrial scale, but fabulously increases productivity, a machine shop that can make a duplicate of itself in an hour, not a year.

At that techlevel the game really changes, but that is a very high techlevel.

Raymond said...

Tony:

"In the case of the car body, I'm see this as more of a (relatively) low cost-of-entry play, rather than a radical new manufacturing method. If you have the infrastructure of a high volume automobile manufacture, it's hard to see how body parts are more efficiently created by deposition than they would be by stamping or molding."

With modern cars from almost all manufacturers, you have so many models, sub-models, variants and options that easily more than half the parts on the car (including body and chassis) have multiple versions. (I should know - I work at a Porsche dealer, spending most of my time juggling this sort of thing, and the rest fighting with the logistics.) It's fairly troublesome for manufacturers to keep all the necessary molds and stamps for each piece available throughout the model's sales cycle, much less afterwards. Most OEMs have pared their supply chain to the bone with just-in-time delivery, even on the factory floor, and the contortions they have to go through to get it that way are a large part of the reason everyone was so scared of the Big Three going under during the credit crunch. Then you have the spare-part supply chains on top of the production ones, which have to maintain a certain level of production for as long as the model is on the road. Add in the non-trivial (as in fairly huge) aftermarket industry, and the efficiencies of the assembly line get stretched pretty badly, eaten away by the costs of the required logistics (and profits - most small cars are merely loss leaders for the parts).

Reasonably efficient 3d printing would very likely revolutionize the auto industry's production processes, starting with spare parts and the aftermarket, and probably bleeding onto the factory floor.

Rick:

"On manufacturing, even if there is a dramatic change in manufacturing tech, I don't think it puts an end to factories. Even if you are printing up cars, you will optimize the printer heads for each part of the job, and organize the work flow."

As an addendum to above, I think you're partly right. Assembly of a car is still a factory job for large volumes, just because of the speed. 3d printers, however, do open up a lot more in the way of customization (and we've got a pretty broad capability for that even now), and give the quickly-expanding custom and tuner shops even more to offer. I could even see a hybrid production model, where certain elements of the vehicle are pre-assembled, and then shipped to the customer's dealer/tuner/whatever for addition of custom parts and final assembly. This would be a good sight more efficient than the current system, where the car is delivered stock, and then the aftermarket guys have to tear it apart again to add in their modifications.

Don't ask me how warranty would work, though - the new car warranty system of any of the big manufacturers barely functions as it stands.

"When the subject is replicators, I think people tend to picture a technology that not only radically reduces industrial scale, but fabulously increases productivity, a machine shop that can make a duplicate of itself in an hour, not a year."

If I were a colony, give me a factory that can duplicate itself in a year, pretty please with sugar on top.

Anonymous said...

I think that replicators (or fabricators, if you prefere), will fall into just a few types for the foreseeable future: one that produces small mechanical/electrical devices and/or components; another type that produces electronic parts; another that will produce large structural or engineering components, like plates or pipes; and, finally, a type that systhisizes chemical compounds, like medicines or complex solvents, for example. That these are several years or decades away doesn't matter; you can still see them in the distance. Put several of each kind into a single building and arrange for raw materials and the only thing you need is to underbiding your compitition to your clients... This doesn't replace factories; it updates them. A small, agile, flexible manufacturing facility will be needed (at least for the first few decades) of off-world colonies, no matter whether they are on Luna, Mars, an asteroid, an orbital spacehab, or a moon of a gas giant. When it takes a year to get Fedex'd a part for your lifesupport; I'd like to make it myself, rather than sufficate, thank-you-very-much! I'll let the grandkids work on economic solutions to local manufacturing at my colony...

Ferrell

Tony said...

Raymond:

"Reasonably efficient 3d printing would very likely revolutionize the auto industry's production processes, starting with spare parts and the aftermarket, and probably bleeding onto the factory floor."

I don't think it would impact first production runs at all. It certainly wouldn't impact industry standard parts and assemblies like fasteners and bearings.

Maybe working in the Porsche supply chain may have skewed your view of this somewhat? I've bought a lot of spare parts for Toyotas, Dodges, and Fords and I think the only nonstandard one was a water heater pipe for a 22RE engine. Even the Dodge Colt drive shafts were pulled out of local inventory at a non-chain auto supply store.

Having said all of that, for out-of-stock and slow-moving items, deposition manufacturing makes sense. There are still a couple of caveats:

1. A lot of parts, particularly in the automotive industry, require special post-fabrication treatments or coatings. So setting up with just a 3D printer is going to severely limit the range and quality of parts that can be supplied.

2. Getting accurate specifications for the parts is going to be an interesting legal question, to say the least.

Raymond said...

Tony:

"I don't think it would impact first production runs at all. It certainly wouldn't impact industry standard parts and assemblies like fasteners and bearings."

I'd agree about first-run production, which is why I think it'd start with aftersales, and it would be most useful to initial production for custom parts or uncommon options. Caveats, though:

A) I think you'd be surprised at the bewildering variety of fasteners used even on a single model. They may still be best served by traditional manufacturing processes during production, but for spare-parts usage there's nothing worse than waiting on a big job for lack of a very particular bolt.

B) Bearings frequently come as part of their housing, which is application-specific. The ball bearings themselves may be standard, but laser-sintered races and housings could be useful.

"Maybe working in the Porsche supply chain may have skewed your view of this somewhat? I've bought a lot of spare parts for Toyotas, Dodges, and Fords and I think the only nonstandard one was a water heater pipe for a 22RE engine. Even the Dodge Colt drive shafts were pulled out of local inventory at a non-chain auto supply store.""

For the record, I've worked aftermarket (with a huge portion being domestics), BMW and Porsche. (Yes, I like me some German automotive engineering.)

Generally speaking, no matter what brand you're talking about, the newer the car, the more varied the parts. This is partly to more complex systems, with more options and variants to fit a wider array of models and sub-models, and partly because modern engines have more space and weight limitations given the advances in baseline performance across the industry, so systems like cooling and accessories require more model-specific alterations.

That said, engine and chassis parts are the most likely to be re-used across models, due to the development costs more than anything. Body and interior parts are the most likely to encompass a bewilderingly huge list of possibilities.

"A lot of parts, particularly in the automotive industry, require special post-fabrication treatments or coatings. So setting up with just a 3D printer is going to severely limit the range and quality of parts that can be supplied."

Most of those treatments are used in engine, drivetrain and suspension components, which are also the most commonly shared. Body parts are usually just sheetmetal of one alloy or another, plus paint. A lot of heating/cooling parts are pretty simple, acutally - just rubber and aluminum. Electronics are a whole different story, and don't get me started on how little standardization there is...

"Getting accurate specifications for the parts is going to be an interesting legal question, to say the least."

Actually, not really. OEMs can obviously use their own specs. Aftermarket suppliers license the specs from the OEMs as it is - often buying the molds from the factories originally used to produce the parts in the first place. Printer-based production would just buy a license to the CAD file.

Tony said...

Raymond:

I really do think your work in German car supply chains is skewing your viewpoint. To stick with the examples I've already put forward, bearings I've bought from third party parts suppliers generally came in sets with races. Scratch that -- they all have, in my experience. If you think about the types of failures that cause bearing replacement, it only makes sense.

The same thing for fasteners. Sure, there are specialized and proprietary fasteners. But the vast majority are SAE standard -- nowdays I bet even in cars built in Stuttgart and Munich.

BTW, I was thinking of licensing CAD files when I said there will be legal issues with 3D printing in the aftermarket. But there's a twist here that I don't think a lot of people consider. When you make out-of-stock or slow-moving items relatively economical to supply, why license it out when you can make the money yourself? Part of delivery on a future part order to a supplier might be the inclusion of digitized specs. (For parts the manufacturer doesn't design itself.) Then future orders could be serviced in-house, rather than passed on to the original supplier or simply refused. Why give that revenue stream to an outside job shop for a license fee?

Raymond said...

Tony:

WRT bearings, I think you're referring mostly to wheel bearings. Older styles tend towards standardized sizes. Recent ones are somewhat less so, usually because of the hub and carrier design of multi-link suspensions (which are much more common on newer cars, even domestics). And yes, bearings typically come with races when bought off-the-shelf. Production-wise, the race sizes are usually customized to a particular application (or selected from a range of sizes offered by the part manufacturer).

The kind of housings I'm speaking of are for things like camshafts, intermediate shafts, transmission input shafts, and other bearings in which the race is part of the housing (or must be adapted to it). Crankshaft main and rod bearings aren't even ball-type, and are absolutely application-specific, but as with a lot of engine parts, are common across one or more model ranges.

As far as fasteners go, they're all SAE, but there are huge variations in not only sizes, but head type, shoulder type and size, grade and material. Trying to keep all of them in stock is not feasible for anything past the factory.

Digitized designs do pose an interesting open question regarding the aftermarket. There are still locality costs and opportunity costs to take into account, though. If the fabrication methods we're discussing become widespread, there would still be advantages to local supply over more centralized production - unless fabs are small enough and cheap enough for each dealer to have one, some of the supply chain issues will remain. And there's still the issue of production capacity, which in the case of car manufacturers will by necessity be heavily slanted to production and support of newer vehicles over old. I think the opportunity to outsource support for old models and free up capacity for the newer ones would be well-received in some quarters (especially the lower-volume manufacturers without factories all over the globe).

Thucydides said...

Patrolling in space makes sense based on the various space warfare threads given the absolute predictability of spaceflight. If the Space Navy is held in orbit around the home planet, then you can calculate the arrival time of the enemy constellation with a high degree of precision, and since there is no stealth, you will know exactly when they launch.

While you have the same issues with patrolling craft, they add elements of complexity and a degree of uncertainty in your calculations; the patrol will certainly arrive at a different time, and enough ships on different patrols could saturate your defense systems much like a cavalry caracole could fix and defeat an infantry formation in the 16th century.

The alternative is not to have a Space Navy at all, but a Strategic Missile Force which can launch ultra high acceleration ORION missiles across the Solar System (at 100+ g). From an industrial point of view, this might be a far better allocation of resources, but from a "Rocketpunk" POV this is just awful.

WRT the use of specialized parts in cars (and presumably any other piece of high to mid tech machinery), I am willing to bet any space civilization will strip down considerably in order to free up resources (replicator or otherwise) given the limited manpower and industrial resource base they will be starting from (and for centuries to come). Screw fasteners might be replaced or supplimented by welding or heat sealing, or something we would not even recognize (manipulating the van der Waals force, perhaps?). Similar patterns of standardization will also happen with other items from chips (or photonics) to ball bearings.

Other things can be scaled quite easily, breaking off chunks of asteroid, surrounding the chunk in a plastic or metal "bag" and placing it in the focus of a solar mirror is a much easier means of mining and extracting material than almost any earthly process (evaporating brine pools to extract salt excepted).

This might lead to an economy where pure raw materials are very cheap (a 30m mirror in cis lunar space can vapouorize copper, so you could literally condense metals out of the process stream), while finished products are very rare and expensive. If your product does not use the very limited number of standard chips, parts or fasteners, then it is essentially hand crafted.

Byron said...

Your argument makes no sense. Enough ships on patrol to fix your defenses requires a lot more ships than you have. Say, two orders of magnitude. If the enemy has that much of an overmatch, you're doomed to start with.
Plus, as Tony said, patrols are based on the concept of a horizon. That's not the only thing. They're also based (in the sense of a forward-deployed unit) on the concept of staying time. Space is huge, and planets are tiny. On a given orbit, you'll spend 99.9% of your time too far away from anything to be of any use. A carrier battle group can park itself off someone's shore and stay there. A spacecraft can't.
Also, any enemy can game the patrol schedule. The absolute predictability of spaceflight is a disadvantage. Wait until one ship passes, then launch whatever. This means that the entire patrol is mostly useless. All it does is slightly limit their timeframe. While your navy is spread out on patrol, mine is concentrated around my home planet. If trouble comes up, I send it all out. The arrival times aren't that much longer, and I have a lot more forces when I need them.

Even if you somehow ignore remass restrictions and break into orbit, individual patrols are still a bad way to go about keeping a presence. You can't keep your crew aboard ship forever, and with a couple months travel to and from, you don't have much time on station. You can make the ship more habitable, but only at the cost of combat ability.

The only reasonable way to keep a presence at another planet is a patrol station. You set up a base with accommodations and recreation for the crews. Make it a couple-year posting. Have them bring their families. Rotate the crews, instead of the ships. I'd guess that with proper design, ships could be on station for ten years or more between refits. Even if you have to supply it from your home planet, it's still far cheaper than moving the ships all the time. Plus, the ships can be designed without as much in the way of habitation features. (It doesn't have to go long distances with more than a skeleton crew, except in wartime.)

Milo said...

Tony:

"2. Getting accurate specifications for the parts is going to be an interesting legal question, to say the least."

No more so than getting accurate blueprints for a conventional factory, I'd think.

Anyway, looks like we've found a niche for space pirates! :)


"When you make out-of-stock or slow-moving items relatively economical to supply, why license it out when you can make the money yourself?"

Because I'm on a different planet from my customer?



Thucydides:

"While you have the same issues with patrolling craft, they add elements of complexity and a degree of uncertainty in your calculations"

Not really. There is still no stealth in space, and that applies to "patrolling" craft too. I can tell exactly where the "patrolling" craft came from, I can tell exactly where the "patrolling" craft is going, and I can make a pretty good estimate when it'll get there.


"enough ships on different patrols could saturate your defense systems much like a cavalry caracole could fix and defeat an infantry formation in the 16th century"

Attacking one by one? That just puts them on the bad side of Lanchester's law. My planetary defenses' firing rates are too high to be defeated by such tricks.

Also, this isn't a patrol. It's just a bunch of attacking ships on slightly different vectors. They still came to my planet with the clear intention of attacking, not peeking around.


Byron:

"A carrier battle group can park itself off someone's shore and stay there. A spacecraft can't."

It can, if by "shore" you mean "orbit".

Byron said...

I already dealt with that, and the answer wasn't good. You double your remass requirements, and don't get that long on station. Presence is part of a naval patrol. It isn't inherently part of a space one.

Tony said...

Milo:

"No more so than getting accurate blueprints for a conventional factory, I'd think."

Which is not the easiest thing to do. Remember, suppliers compete for parts contracts. A lot of the time, they design the parts they build, going off of performance and envelope specifications. For a lot of things, vehicle manufacturers just pick out of catalogs and integrate those parts into their final designs.

So, even if you've got the capacity to build replacement parts of a given type, do you know who to go to for the drawings? If you do, will they license out manufacturing to you? If you're trying to set up in business for multiple part types for multiple makes and models, how many OEMs and parts suppliers do you have to go to to get all the drawings you need? Are the people that know who you need to go to willing to tell you? Why would a parts supplier give their proprietary technical information away to a potential competitor, even for a price? What kind of noncompete agreement could you darw up that they would sign?

See, it's not a straight licensing of a whole technological package, like H&K licensing G3 rifle manufacturing to Turkey and Pakistan. It's a multitude of suppliers with proprietary designs and technologies to protect, that not all of whom -- maybe not even a few of them -- are willing to license out.

"Because I'm on a different planet from my customer?"

Plausible midfuture, Milo, plausible midfuture. Part of the purchase agreement for anything that might need replacement on Mars is going to include replacement manufacturing data up front. It's going to be totally out of the normal market.

Rick said...

This is yet another case where I think uneasy peace is more interesting than all out war.

In interplanetary war there is pretty much just strike and counterstrike. In uneasy peace you probably will have patrols, for flag showing and close up observation, and sometimes policing.

The ships for that mission won't have much use in an actual shooting war, when pretty much all that matters are killer buses and laser stars.

Geoffrey S H said...

... but if there can't be a patrol, or a show-the-flag trip past a territory... you can't have that either...

tkinias said...

On the question of patrols, I think there’s a place for them in the plausible midfuture, even given the lack-of-stealth issue. But they obviously will work differently than horizon-limited air or naval patrols.

Consider Cold War–era recon satellites. You knew exactly when they were coming overhead, so they were not going to catch you by surprise; they did put limitations on what you could do in the open, though. Maybe I know that the during first weeks of February, May, August, and November my rival’s frigate will be making a “courtesy visit”; that makes it easy to hide some things from view during those visits, but it also makes things that can’t be made to disappear much harder to do.

Rick said...

Stay tuned ...

tkinias said...

(continuation)

In my own midfuture setting, major powers also tend have military spacecraft en route to and from various places of interest in the solar system, especially in low-energy transfer orbits not too distant from the gaggle of transports which left with the monsoon winds (er, departure window).

If the Australians launch a fleet from near Earth against the American settlement on Ceres, there’s no hiding it and patrols don’t make a difference. But if there’s an “incident” between Australians and Americans somewhere on the way from Earth to Ceres, the USSF might want to have a military vessel nearby to “encourage” a favorable resolution. Nearby in that context means “in a similar orbit”.

Remember that in the plausible midfuture, interplanetary spacecraft probably spend a lot of time coasting in transfer orbits...

Tony said...

tkinias:

"Remember that in the plausible midfuture, interplanetary spacecraft probably spend a lot of time coasting in transfer orbits..."

In that same midfuture, there won't be resources for dedicated warships. I think Jerry Pournelle got it right when he suggested in Exiles to Glory and "Tinker" that a common means of exerting power would be to load a ship up with cops and just take over a settlement when you got there.

Byron said...

Yes, the ships will spend a bunch of time in transfer orbits. But the ships on said orbits will likely be unmanned, and certainly unarmed. Passenger-carrying vessels (anything with a large crew, including most warships) will have different orbits than cargo vessels. And we usually don't worry about there being incidents between supertankers that require the military to resolve (except for piracy, but that's different).
And all the interesting stuff will happen around settlements. It makes a lot more sense to have ships there than in transfer orbits.
The point of a patrol wouldn't be to see what was happening. That can be accomplished with drones and spy ships. That will happen, but it will be different from full military patrols with warships.

tkinias said...

Tony:

In that same midfuture, there won't be resources for dedicated warships.

I don’t believe that necessarily follows. In the setting I’m working in, for example, Australia has a defence budget of around a trillion (present-day US) dollars, and around a third goes to the space force. Four other states have space-force budgets over $100 billion. It’s very much a midfuture setting, though, it most respects.

Byron:

Passenger-carrying vessels (anything with a large crew, including most warships) will have different orbits than cargo vessels.

Whether that is true would depend on the economics of propulsion systems. It’s quite possible that low-cost passenger transport would seek to minimize delta-v and would thus find itself in similar orbits to the robot freighters.

And all the interesting stuff will happen around settlements. It makes a lot more sense to have ships there than in transfer orbits.

Only having military/police presence around settlements would pretty much guarantee that “interesting stuff” would tend to occur elsewhere, wouldn’t it?

The point of a patrol wouldn't be to see what was happening.

I didn’t mean to suggest that; the comparison with recon satellites was just to illustrate that a predictable patrol is not ipso facto useless.

Another analogy: compare the police department setting up a CCTV camera watching my driveway and front door with their paying a personal visit once per month. The former is like the observation you can do remotely, the latter is like sending a patrol spacecraft.

Thucydides said...

Passenger ships will travel as fast as possible, simply because keeping lots of people alive will be expensive, and that the more trips you can get with paying passengers, the better off your shipping company will be. An Earth-Mars spaceline using massive VASMIR engines advertised to make the trip in 39 days may not be JetBlue, but will have certain advantages over the cut rate line that takes 180 days...

Freight, OTOH, will be tossed on transfer orbits in the space equivalent of 20' shipping containers via mass driver or some low cost propulsion system like a magnetic sail.

Thinking about this and looking at the comments, "patrols" can happen if the ships are automated busses, sheathed in metamaterials and cold launched using momentum tethers. Orbital adjustments and acceleration would need some very low signature drive (magnetic or electrostatic sails would be my guess), and they would swing by undetected to make observations or suddenly unmask and fire their KKV's in a spasm attack to overwhelm defenders.

More cruise missile than Horatio Hormblower, and alas, not a Rocketpunk future either.

Tony said...

tkinias:

"I don’t believe that necessarily follows. In the setting I’m working in, for example, Australia has a defence budget of around a trillion (present-day US) dollars, and around a third goes to the space force. Four other states have space-force budgets over $100 billion. It’s very much a midfuture setting, though, it most respects."

In the setting you're working in...I see. A trillion USD...gotcha.

You did get the memo about this being the plausible midfuture blog, right?

Jim Baerg said...

Thucydides: "Orbital adjustments and acceleration would need some very low signature drive (magnetic or electrostatic sails would be my guess)"

Would that actually be 'low signature'? Those would generate thrust by deflecting the charged particles of the solar wind, that acceleration of charges should involve some generation of radio frequency EM radiation.

Once again - no stealth in space.

Thucydides said...

No stealth in space, but much harder to spot than blazing plumes of plasma from nuclear or fusion engines. These busses might also be dropped just before or after a ship makes a burn, to drift on a different trajectory.

I would also guess that the Admiralty would set things up to minimize patterns so the RF energy would be lost in the static or masked by natural sources of RF energy. The Imperial Jovian Space Navy will be swamped with targets as other Space Navies use the magnetosphere of Jupiter to mask their activities...

Rick said...

I can imagine plausible midfutures where Australia's defense budget is more or less equivalent to the present day US defense budget, on order of $1 trillion. Let massive immigration raise Australia's population to 100 million, and put 15 percent of GDP into the military versus a few percent, and you are in the neighborhood.

The plausibility strain isn't Australia in 2200 having the military budget; it is them spending a third of it in deep space.

But we are all guilty of that one!

Anonymous said...

fioricet specialsfioricet experience

Also, there was also any delivery on the legislature of the idf.

Anonymous said...

Rick said:"The plausibility strain isn't Australia in 2200 having the military budget; it is them spending a third of it in deep space."

Depends on whether they think their stake in the PGM mines on Ceres is worth the trouble of maintaining a constellation of combat spacecraft to enforce the 'hands off' signs slapped on the cargo containers.

Ferrell

jollyreaper said...
This comment has been removed by the author.
jollyreaper said...

Here's a potential game changer -- right now, it takes years of effort for a human being to become useful at a given skill. Just think of the amount of effort it takes to maintain a community at roman levels of tech, then consider 19th century wild west tech, then consider that it only gets worse the further into the future we look. Modern day? Far more complex. A hundred years from now? Possibly more simple. Huh? How?

I'm thinking of that nice little scene from the Matrix. "Do you know how to fly this helicopter?" "No. Load the flight program. Now I do." Hotswappable skill sets. We're seeing the first inkling of this with augmented reality maintenance systems the military is researching, the sort of thing that walks you through the maintenance process. Loosen this bolt, pull this assembly. Will we be able to eventually upload these skills so you know kung-fu? Unknown but it would completely change the dynamic we're operating under now where each specialty requires a lifetime of dedication to mastering.

Now if that's not a potential solution, how about from the computer vision/manipulation end? My roomba can vac the house but it needs me to clean itself. It needs me to empty the dustbin, me to unscrew the brush deck and clean out the old grease and hairs, apply fresh grease, clean the brushes, etc. A modern warship may be automated as far as the firefight goes but humans need to push the buttons and humans need to grease the gears. The ship can't maintain itself. But we're already seeing computer vision systems paired with amazingly dextrous hands. What happens when computerized hands become cheaper than paying humans to do that part of the job?

I'm not sure how far off a self-perpetuating, automated industrial base is. I have a sneaking suspicion we'll have something like that before we see proper space colonies, the same way I think we'll see convincing androids before we see mars colonies.

Milo said...

This is where the distinction of knowledge and understanding comes into play. It takes more than just knowing a set of facts to truly develop an intuition for a skill.

Of course, if either strong AI or mind uploading becomes practical, then you could just train one brain to do something and then make a lot of copies of that brain.

jollyreaper said...

Yeah. I have a feeling that this sort of speculation is beyond plausible mid-future and is approaching woo-woo territory. It could happen or maybe not. But I think we're at the cusp of the next revolution in automatic manufacture.

George said...

Only an awfully massive economy, an oversized industrial infrastructure, will support the online of factories and skunkworks, launch and tracking sites, educational institutes and coaching facilities, with their many specialized skills, that get into a present-day space operation.

https://www.suncoasttools.com/products/carbide-end-mill-cutting-tools

Saint Michael said...

"I'm not sure how far off a self-perpetuating, automated industrial base is. I have a sneaking suspicion we'll have something like that before we see proper space colonies, the same way I think we'll see convincing androids before we see mars colonies." = jollyreaper

Depends on what is meant by "proper" space colonies. If it means a totally self sustaining single habitat, that would likely have to wait until well beyond the PMF. Even then there's volatile losses to contend with, though they might make their own mining drones.

But creating a multi-community distributed infrastructure not at all reliant on Earth could plausibly be late in the PMF. A century or two from 2021? Just as no one nation on Earth is totally without import or export, most space habs would most probably come to rely on each other's economies, populations, ecologies, and industries to form an interlocking network to produce what is needed.

Since 2010 a lot has changed, including crewed space launches getting cheaper (and looking to get even more so soon) and AI/robotics becoming more capable though hardly to self-repair capacity. See also Lights-Out Manufacturing. LOM only has taken place on Earth of course, within an already existing planet-sized infrastructure and full sized factory buildings, but it's a start. Baby steps, baby.