Spacecraft may one day be so common
that everyone has flown on one, with thousands of new spaceships being
built every year in manufacturies dwarfing the enormous facilities we build
planes and sea-going vessels in. Building a spaceship is an enormous undertaking.
They tend to cost as much as an aircraft carrier, even though the crew and cargo are in a space
about the size of a small boat or modest yacht. They require precision-manufacturing unlike any
other vehicle, and yet, one day their manufacture might be as routine as that of an automobile.
Emphasis on routine, not simple, as there’s nothing simple about the modern car,
and I wouldn’t expect that to be true of personal spaceships either, but in many ways
a spaceship can be simpler than a terrestrial vehicle. Space is mostly empty with few
obstacles or shifts to road surface or air density to require constant course corrections.
Today we’ll be examining a lot of different paths for the future of building ships, and we’ll be
looking at everything from simple asteroid-mining ships produced in a low-tech future, to cases
where all the real work was done on the software and design end, and someone just spills some
self-replicator robots on a spare asteroid, and they replicate and spawn ships out of it, or maybe
even grow them quasi-biologically. And as we’ll see, even in these high-tech situations, there
are limits to how fast you can produce a fleet. Fundamentally, scale is what we’re looking
at today though; not building just one ship, or a prototype, but the mass manufacture
of them: starship factories or dockyards. When I was a young teen there was a cartoon called
Exosquad that is a bit of a hidden gem. It often had deep, realistic, and philosophical discussions
buried among second-rate animation and comic action scenes, but I liked it and it features
humanity at war, and the good guys forced back to Jupiter at the start of the second season,
where they’re repairing and rebuilding their fleet. They lost the flagship in the close of the
first season, and its successor was sabotaged and blown up while halfway built at their shipyard.
That made me wonder how the heck they ever found the resources to build a ship, when the
bad guys had the 3 real planets, Earth, Venus, and Mars, under their thumb, and we’re
never told that any of the others are seriously inhabited. Those ships were huge, and they showed
people in spacesuits going EVA to weld them. Automation, of course, can be a huge game changer
in how these things work, but if you think about the newest aircraft carrier we built, the Gerald
Ford, that came in at around 13 billion dollars, and if to keep the math easy we assumed everyone
in that building process and supply chain averaged 26 dollars an hour in pay and compensation,
that would imply half a billion man-hours went into construction – most of that occurring
off-site, mining and making the materials and paying the people who did that or made their
equipment – but half a billion is a huge number, and even the annual version of that,
assuming 2000 work hours a year, is still a quarter of a million years of labor.
It’s important to understand that to build in space, almost all the material needs to be up
there for it be economically plausible, and while a lot of the cost might be in hyper-expensive
tiny microchips or research, development, and prototyping down on the ground, you are building
that ship in space, and also mining, refining and smelting all the metal for it there too.
It is not hard to imagine needing a million people in space, in terms of workers and families,
to build one big spaceship, especially those mile-long ones popular in sci fi. Even with a
lot of automation, that still seems plausible. But I thought we’d begin in an era when there are
still less than a million people living in space, and look at a pair of fictional shipyards,
The George Mueller Orbital Shipyard, around Earth, affectionately known as Port
George, and the Linus Collection Point, orbiting 22 Kalliope in the Asteroid Belt,
affectionately known as the Scrapyard. Port George is named in honor of
the father of the space shuttle, and handles a number of different designs but is
best known for its tender vessels, medium-sized ships with a crew of 6-12 designed for multi-week
operation anywhere in Cis-Lunar Space out to the Earth-Sun Lagrange L-1 Point. It’s designed to be
able to rendezvous with existing space structures or satellites and either to bring them inside,
for smaller ones, or go EVA for larger ones. Port George itself has relatively minimal
EVA involved in construction, it is a huge, and mostly hollow facility, and even its areas
exposed to the vacuum of space are typically still enclosed, just depressurized, to help
protect the crew and vehicle under construction. Many of the ships they build are principally
made of aluminum mined from Luna, and so there is little concern of oxidation damage to the
outside. Even for those ships which are more sensitive to oxygen it is usually considered
safer and easier to have a thin-walled pressure chamber for the dockyard workers to operate in.
Indeed, much of Port George is under low-speed rotation to permit about a fifth of Earth-normal
Gravity, as it was determined that a standard earth atmosphere was needed, but that the
workers mostly benefited from just enough gravity to feel like there was an up and
a down. Personal living quarters generally are kept at 80-100% of standard gravity.
One of the more interesting things about Port George is that a large portion of its
workers live down on Earth, and it would not be unusual for a work crew to consist of one or
two humans and a dozen robots being controlled by telepresence operators down on Earth - after all,
the time delay from Earth's surface to orbit is much less than a second, so, although the latency
is something that takes some getting used to – and may require some anti-nausea medication - it's far
from a dealbreaker. After some trial and error, it was determined that keeping those same operators
with those same on-site teams tended to result in far better group cohesion and quality control, so
many an on-site dockyard worker has workmates down on Earth they’re close enough with to occasionally
visit whenever they’re on shore leave. For very big ships, such as an Aldrin Cycler, Port
George used to do exterior unshielded assembly of main components, but is now using inflatable
bubbles of a thin-walled, radiation absorbent material, to allow crews to use a relatively
low-mass spacesuit with their helmet mag-locked to their thigh for if there’s a pressure drop.
This is seen as an overall safer work environment due to the physical exhaustion and psychological
issues of bulkier suits designed for extended open-space, or ‘void’ work as it is called. The
more recent efforts to systemize space debris clearance have also made such large inflatable
bubbles more long-lived. Though punctures remain common and are typically treated with a quick
patch till the entire bubble degrades, or the project is complete, and it can be recycled.
Along with being able to rely on Earth for remote workers, Port George gets a lot of state funding
and has critical but small components manufactured down on Earth. Nonetheless it still has to
be frugal about all its resources, especially those brought up from Earth, which is why salvage
and recycling are critical to their operations. Port George is also known for its construction
of the DS-12 Toy Box, a space debris collection ship with minimum bells and whistles, designed
to allow a 1 or 2 person crew to salvage damaged orbitals and control a number of drones that are
able to collect pieces of scrap with low delta-v relative to the ship. The nickname of Toy Box
officially is for all the neat drones the model comes equipped with, but everyone knows it's for
all the weird and unique garbage they collect. While the DS-12 Toybox is viewed with a certain
amount of amused contempt as a junker among other orbital spaceship crews, the handful of them
that have been transported to the Asteroid Belt are practically considered luxury yachts.
The growing Space Industry is hungry for metal but specifically for the cheapest metal, and
that’s where The Scrapyard comes in. 22 Kalliope, at over a hundred miles across in most
directions, is the second largest Metallic Asteroid in the Belt, and something of a rubble
pile including hydrated mineral and silicates, and is home to a handful of mining operations, but
Kalliope is better known for having its own Moon, Linus, which itself is 20 miles across and thus
is bigger than both of Mars’ moons combined. The Scrapyard formed not long after someone
had the idea to run a skyhook-style tether directly between Kalliope and Linus,
hanging just over Kalliope’s surface, which has a 4 hour rotation rate. Linus orbits it
twice a week a thousand kilometers from Kalliope, and which with good timing can be used to allow
ships or cargo pods to accelerate and release around the Belt or even back to Earth. This has
made it a popular port of call of asteroid miners, which the Census say now number 30,000 throughout
the system, roughly 10,000 of which are in single or two person crewed, owner operator mining
and prospecting ships, several hundred of which visit this asteroid pair for the discounted
rates of shipping cargo both to and from Earth. Many of these ships were built at the
Scrapyard on Linus, and here we see the Art of Minimalist Shipbuilding not seen
since the old days of early spaceship travel, and many of which would be illegal to
operate near Earth for a variety of reasons, ranging from worker safety hazards to some
using radioactive materials. Ultimately, every dollar spent on a ship or its crew
has to be paid for with the metals it mines, so the Scrapyard allows lone individuals or small
teams to be competitive with the larger and better equipped markets, at the price of taking some
additional risks and enduring more discomforts. Fundamentally, a spaceship is just
a pressurized box with an engine, and the Scrapyard understands this all too well.
Their ships aren’t for landing on big planets or taking off through atmospheres. They don’t
need the radiation shielding the DS-12 needs, because they’re about 3 times further from the Sun
than Earth, and thus only get 12% of the radiation from the Sun, and there’s no Van Allen Radiation
Belt out in the Asteroid Belt, and also, no government inspectors pushing for worker safety.
This far out from the sun, and orbiting it, and not Earth, the majority of space
debris is traveling relatively slowly, so it is possible to armor ships against that, and
uranium is reasonably plentiful out in the belt, as is thorium. Solar power isn’t a very realistic
option here, it can be done by using large, thin reflective dishes to concentrate sunlight,
but that’s less viable while moving around, and the preferred method for mining and
prospecting is to park your ship in a deep crater, to protect it from micro-meteors and other
debris, but where it won’t get a lot of sunlight. So the Scrapyard tends to make a lot of ships
that are very like the cargo pods they shoot back to Earth; brutally simple. Those ships are
often only two rooms, a main room for living, working, and sleeping, and a smaller airlocked
room for exiting, which often does double-duty as a bathroom. They’ve got robots who make the big
metal plates, and the rest they slap together, and a lone person can easily move a multi-ton
plate in microgravity. It's a slow process of shoving inertial mass around, but
there’s just the air holding it back. That’s one of the neat things about a zero-gravity
shipyard, you don’t see many forklifts, and large items are often slowly moved into
place with a protective coating of inflatable air pillows on the side to minimize collision
damage during installation. Sometimes, one will get punctured and cause a fairly heavy object,
like a nuclear power module, to start drifting and spinning around. There are surprisingly few
injuries in the shipyards, but at least one was someone having a heart attack at the sight of
a nuclear reactor careening around the dock. They can make a ship in just a few days,
because they are not large or complex. Weld the hull together, get an airlock and engine
port on it, a window, usually on the opposite side of the airlock, so you can escape
if one side is blocked for some reason. If your ship burns something out on an asteroid
crater it can cause a venting of gas that could knock it over, and then, zero-gravity or not, it
could get stuck and wedged in, airlock-side down. The very simplest ships just have a RTG and a
device that can run metalysis on ore and produce metal plus oxygen, and they use that oxygen as
their propellant when they’re ready to leave, usually buying methane for fuel back
at the depot, shipped in from Titan, which buys pressurized pods from Kalliope.
These ships have nothing like the efficiency of the bigger and more expensive fission reactor
versions, running on more elaborate Ion Drives, but they are cheap and you can make one on your
own, often by scrounging parts from many of the wrecked and discarded bits awaiting salvage or
disposal on Linus, whose escape velocity of about 20 meters per second, or 45 miles per hour makes
it just sufficient enough to comfortably hold down cargo or salvage, as well as the crew, who might
get flung off something, or some loose debris may collide into and cause an air tank to leak.
One of those issues with EVA, for an asteroid miner or dockyard worker, is that if you forget to
tether yourself or any gear, it’s generally going to get lost. Fetching drones are very popular
but not cheap. Kalliope’s has a annual festival, which is held every 5 Earth years, as Kalliope’s
orbit around the Sun is 5 Earth years, and it has many contests, one of which is who can untangle
tool tethers fastest while in a full EVA suit, and the current champion, four times running,
is the current elected mayor of the binary asteroid group. She ran on untangling the complex
regulations for shipping back to Earth. Other contests include sealing a leaky compartment,
finding a leak on a wall covered in consoles, and various feats of dexterity and acrobatics
in microgravity while wearing a spacesuit. This episode isn’t about the lives of asteroids
miners or shipyard workers, but the folks at the Scrapyard show us a future in which you could
have a thriving shipbuilding and asteroid-mining economy a billion kilometers from Earth, and yet
needing only modern technology and automation. Let’s consider the other extremes of that, like
ultra-automation and self-growing ships, and begin by considering the nanotech self-replicator
option. We tend to have this assumption that we can make tiny little machines that can make
copies of themselves and perform some other task, general or specialized, and we assume if something
the size of a biological cell can do this, we should be able to make something as good
or better, and about the same size or smaller. Such being the case, one designed to live in
a vacuum and take apart local regolith seems plausible enough, and they could have their own
equivalent of DNA and an extra separate one for the processes or items they were supposed to
make. Like turn on, make a copy of itself, build one of object A, make another
copy of itself, build another object A, etc. That would be a specialized
version, an Object A might be a specific object like a chunk of metal plate, or
a paperclip, or maybe even a whole spaceship. Now, this is popular in sci fi. As are
the nanobots running wild as Grey Goo or Hegemonizing Swarm. However, in practice we
would borrow further from nature by having layers of ecosystems, and then further depart from
nature by not feeling obliged to have an organism replicate itself specifically. So instead
of having one species of self-replicator, you would probably instead have a few hundred, or
million, each designed for some specific tasks and built along different lines and scales. This one
seeks metal deposits, this one builds kilns for melting metal, this one makes the rivets
for use in Airlocks, this one makes wire, this one makes killer drones that seek out
any faulty or mutant bots and kills them. And none of them self-replicate, rather some big
drone arrives and builds ten smaller drones meant for building each of ten other smaller drones
who each have a specific model of yet-smaller drone that they make. And you probably have
some control variable the machines can’t make or get that limits reproduction – this might be
some specific rare element or some manufactured black box widget some of the bigger replicators
need, or even something akin to Bitcoin. One of the replicating layers has to mine codes
before reproducing, and this limits its reproduction once it begins getting in excess
of the estimated number needed for the project. So, your replicator arrives at a spare
asteroid and turns it into a ship, or ships, and this is simple enough conceptually but its
sheer simplicity tends to make folks assume it’s the end of all normal manufacturing.
In practice, this is not an insta-ship, even if your drones have a magical infinite
power supply, they can only work so fast without overheating themselves and their surroundings
till they get melted by their own frenetic work. Also, they break very easily, and the sturdier
you make them, the more replication time they need and the more energy per task they need. We
talked about this more in our Self-replicating space probes and Santa Claus Machine episodes,
and there are a lot of limitations in how fast you can do 3D printing, nanotech, or Star Trek
style replicators. Odds are there are more problems we don’t even know of yet too.
The Clanking Self-Replicator is often a better option, and this is more the assumption
that you’re not going for tiny little robots, but a bunch of big drones and factories.
Big animals and plants, not microbes. At its core, imagine a factory for making robots
that could make any of the robots needed to perform most of the functions in that factory, to
supply that factory, or build another copy of that factory. This is going to use a huge, insulated
cauldron for making its metals because this is more efficient than a bunch of tiny bots. We don’t
really use these on Earth though because it’s so much easier to employ a human for any of the parts
that aren’t easily roboticized or repetitive, and we have a lot of humans who need work.
It's hard to predict how labor will go on Earth; where we have a plentiful supply of people,
and there’s good reason to think those people actually require more effort and resources if
you don’t keep them occupied usefully, so we will likely never have totally automated factories.
But in space, we may have to, especially with interstellar efforts. More likely though, we
would engage in Human-Machine Teaming, which at least in recent decades has tended to prove far
more useful than machines alone or humans alone. So, our spaceship factories in space,
where manpower is likely to be limited, will probably tend to be in this semi-autonomous
clanking self-replicator situation where some large but not truly automated chunk of the
supply chain is being partially run by humans. Maybe maintenance, maybe they are the ones
doing final assembly and quality control, and the overall administration and
management might be jobs for them too. This would seem to work better for very big
projects like those mile-long battleships sci fi loves, or those even bigger Arkship
colony vessels we love to discuss on this show for interstellar generation ships.
This strikes as your most likely setup in part because it resembles modern manufacturing, but
is likely to have a lot of very simplified supply chains, in locations if not steps. You’re always
able to move faster and better in manufacturing if you’re dedicating a specific spot to making
a specific item, not the whole device. For us, it isn’t the lone bot or factory that
self-replicates, or even the lone human or couple, but rather the civilization itself
that self-replicates, and that’s very needed where an arkship is considered because you are
not just building a big metal cylinder. You have to set up and decorate or equip each compartment,
some of which need to be elaborate ecosystems for organisms you’re taking with you, others would
need complex factories for gear you need to repair and replace, and yet more would be schools,
parks, farms, gardens, and entire communities. So we can imagine tiny little robots or
entire automated facilities doing this, maybe even building critters and people from
DNA templates and education and care archives, but there are so many hurdles in the way of
this process – what we call seed ships or data ships – that it would seem hard for it to go fast.
Indeed, when you think about it, while we can repair things hyper-fast NASCAR style, or build
houses or bridges overnight. Generally speaking, we tend to move bigger things, and more slowly
than in prior times. Houses don’t get built overnight or in a few weeks, it’s a process of
many months or even years for bigger buildings, and it produces superior work, at least for
whatever quantities you’re optimizing around, be it time, labor, money, minimal
paperwork, minimal local disruption, etc. I think this will be the case more in the future
too, but in the context of this larger organism doing replication, we might consider mitosis as
an option - that our starship factories of the future might be starships themselves. We have
contemplated this in the case of our Gardener Ship; a large interstellar generation ship that
needs to be able to manufacture any component the ship or colony will need in case it breaks,
since in multi-decade or multi-century journeys, anything can break. Thus they can manufacture
an entire new ship so long as they’re given the time and enough raw materials, and
they have a lot of time. They have decades between nearest points in which to
do little but build and breed more people. So, arriving at a colony destination, it
would seem very likely that many people who spent their whole life on that giant ship wouldn’t
necessarily want to become a planetary colonist, and the population would have risen during
that voyage, so some stay on the old ship, some go to form the new colony, and some go
on the new ship to a new destination after they help the colonists start up and refill
their raw supplies. They just keep doing this, growing their numbers and restocking as they
journey toward the galactic rim, see our Gardening the Galaxy episode for a look at that life.
But critically, it offers us the option of mitosis for ships, or colonial fleets, as they
might be a small fleet rather than a lone ship, and might add ships as they go, building them
from raw materials acquired at each stop, and occasionally divide that fleet like a cell
would. Or a Spaceship might flat-out grow longer over time like a big, long worm, then, unlike
actual worms, those types of ships could divide and form new ships by cutting themselves in half.
Or they might even grow themselves in a DNA helix style and unzip the ship into two identical ones.
We’ll be exploring more of the ideas of truly complicated mass manufacturing, and dedicating
entire large asteroids to industry – or entire moons or planets – this weekend in our Scifi
Sunday episode: Forge Worlds & Industrial Planets, but as we’re seeing today, it is very
likely that your true starship factories, those making big ships, along the lines of
colony vessels, not merely little shuttles, are more likely to be entire cities or nation states
in space, not some small dockyard of hundreds, and in the future, a project of that scope might
even be accomplished largely by automation. We’re also seeing this automation is likely to
involve many layers, from the classic welder, to entire biosystems of nanotech, to gardeners
and biologists, and even more obscure things like hospital administrators, investors and
recruiters, because those big ships aren’t just a gun and engine, they’re a small
civilization – or even a large one. And you need experts in everything, and storage
space for everything, up to and potentially including antimatter fuel or weapon storage,
or space for the guys who handle installing the micro-black hole in the spaceship’s basement.
Indeed, you might need a whole world to really do it right, especially for big ones, and it might
be that you do build them down on the ground, then winch them up to space by tether between
a pair of orbital rings, but again we’ll examine that option this weekend in Forge Worlds.
One last option for acquiring your spaceships is not to manufacture them, but to instead steal
them from others, and space piracy might be a fairly complex process, especially as it’s often
likely to be data and designs you need to steal, not just raw materials or ultra-expensive computer
chips. However, it is always possible we might be able to make ships that could travel to other
universes if the Multiverse idea is true, in which case you might be able to make your
starships by manufacturing one, then stealing copies from those adjoining multiverses where
an almost identical spaceship just got produced. Of course, that implies you’re violating the first
rule of warfare, never try picking on someone your own size, since they presumably have identical
technology and infrastructure to yourself. It’s also a good reminder about scale, because
while those massive mile-long ships might be popular in sci fi for their sheer appearance
of immensity, we have good reason to believe that bigger will be better for a lot of aspects
of ship efficiency, speed, or combat prowess, and that space faring civilizations can get away
with both quantity and quality, fielding huge numbers of huge ships which are each incredibly
sophisticated, multipurpose capable, and powerful. And so, while we might see small ship factories
like The Scrapyard, turning out tiny boats, or others making small personal yachts, like
we discussed in Your Own Personal Spaceship, there is definitely room for bigger starship
factories, and they might need that room too, as some might be as big as entire
continents, or encircle entire planets.
I always enjoy writing these episodes with
the narrative formats to them as it seems like stories are just such a good approach to teaching
science. If you’ve done much STEM teaching you probably know how hard it is to teach some
concepts, like electricity and circuits, to even smart adults or teens. It takes a
unique mixture of hands on experience and narrative to really make it stick. This is why I
was fascinated by our newest sponsor, Upper Story. Their newest game Spintronics, has number of
example puzzles for you to build woven inside a beautifully illustrated graphic novel.
So instead of boring instructions you get a steampunk-themed story of a young clockmaster
learning an alternative technology, electricity, and using gears to build mechanical circuits
and teach about both them and electricity. Spintronics does an amazing job showing all the
electronic components in a hands-on way that is fun and instructive. I loved it and found it quite
stimulating but more importantly my kids loved it, everything from experimenting with different
resistor combinations to building chain linkage, and I had to use ice cream to lure them away.
So as we go into the holiday season, if you’re looking for a fun and hands on educational gift
for your family or friends, something that can be played in a group or solo, use the link in the
episode description upperstory.com/spintronics to learn more, and don’t forget to use coupon code
ISAACARTHUR at checkout to save 10% on your order. So back in August we began circulating a petition
to save the New Horizons Space probe out in the Kuiper Belt, to keep the research team for
the billion dollar probe billions of miles from home running, and I’m very glad to say that
was a success. We got thousands of signatures, and the bulk of them, especially the early
ones so critical to getting it to snowball, come from this audience. As did the donations
that helped promote the petition for others to see it. So if you signed that petition, shared
it, donated to it, or wrote your congressmen, yes, you absolutely saved the New Horizons
probe so it could continue the process of expanding our horizons. Thank you so much. We
literally could not have done this without you. I did also want to thank the rest of the
leadership at the National Space Society for their support for this, particularly
Executive Vice President Hoyt Davidson, who did so much of the heavy lifting on the front
end of this, and the Beyond Earth Institute and Space Frontier Foundation for co-signing our
letter to NASA and congress, along with many other groups like Space.com and Universe Today who
helped raise awareness or circulate the petition once we got it going, and it just helps show what
amazing things we can do when working together for a shared greater future. So again on behalf of the
National Space Society, thank you, and Ad Astra! Speaking of building the future, this
weekend on Scifi Sunday we’ll be continuing our factory theme by contemplating
entire planets devoted to industry, in Forgeworlds & Industrial Planets. Then
next Thursday on October 19th we’ll discuss if Life Extension Is Ethical. Then we will look at
another type of dedicated planet, Fortress Worlds, on October 26th, and in two weeks we’ll finish
October, with our monthly Livestream Q&A, on the 29th. Then on November 2nd we’ll
ask if the Rebel Space Colonies we often contemplate in scifi and futurism might occur
and what they’ll be like, before we release our big-3 hour long updated and extended edition
of the Fermi Paradox Compendium on November 9th. If you’d like to get alerts when those and other
episodes come out, make sure to hit the like, subscribe, and notification buttons. You can
also help support the show on Patreon, and if you want to donate and help in other ways, you
can see those options by visiting our website, IsaacArthur.net. You can also catch all
of SFIA’s episodes early and ad free on our streaming service, Nebula, along with hours
of bonus content, at go.nebula.tv/isaacarthur. As always, thanks for watching,
and have a Great Week!
