Many people commute as much as a hundred kilometers to work every day.
Interestingly, this is also about how far it is to low orbit.
If we ever want to have a truly serious presence in space,
we are going to need a lot of infrastructure in orbit.
Today we are going to try painting a picture of what that would look like, and review some
of the challenges we have to overcome to achieve it.
Many of those challenges get far easier as you get more and more infrastructure built
up there, but one, the danger of orbital debris, increases as you get more potential sources
of debris in orbit.
You can also have a runaway chain-reaction of destruction we'll discuss later today
known as Kessler Syndrome which has the potential to cut you off from space altogether.
So how do we start building up that infrastructure and what should it be?
There are many approaches and since last time in the series we discussed spaceports in terms
of the Gateway Foundations' design, I thought we'd just pick up from there.
This concept of building in space is hardly new, but in recent years there has been an
increasing focus on the use of robots and drones for doing the work, so we'll spend
some time on that too.
Whether new plans or old, what they all seem to have in common is a big problem of what
exactly you are manufacturing up there and what the infrastructure and factories would
be for.
As I mentioned in the Interplanetary Trade episode, in the long term Earth really has
nothing it needs to import besides precious minerals like gold and platinum and rare earth
elements, and Earth's obvious exports are people and information.
The latter doesn't require much infrastructure in space.
Earth churns out endless volumes of science, movies, TV shows, etc and can export them
by simple radio or laser broadcast.
I made those statements in the context of interplanetary trade and, in so-doing, I lumped
Earth's space-based industries in Near Earth Orbit with ground-based manufacture.
Instead, today we will separate the industries on Earth from those in Orbit around Earth.
We actually don't need much infrastructure to receive precious metals from asteroid mining.
It takes a lot of fuel to move materials at interplanetary speeds so by and large you'd
do your refining on location and send it to Earth in a pod to aerobrake in our atmosphere
and be picked up by a ship.
You wouldn't send it to Earth orbit for smelting and transport it down.
However, the value of refined metals like Aluminium and Steel in LEO cannot be overstated.
The budding space construction industry created by a project like building a spaceport will
be very hungry for raw material inputs, and launching all of them from earth is expensive.
Many folks will be aware of the recent successful Falcon Heavy launch that put Elon Musk's
Tesla into a solar orbit that will pass by Mars.
I have to admire the publicity stunt of that launch, especially the inclusion of "The
Hitchhiker's Guide to the Galaxy" by Douglas Adams in the glovebox, as well as a towel
and a sign saying "Don't Panic" and, of course, the data storage device with Isaac Asimov's
classic space trilogy on it.
Importantly for our purposes today, though, the Falcon Heavy will allow us to put payloads
of up to 64 metric tonnes, a mass greater than a 737 jetliner loaded with passengers,
crew, luggage and fuel into low Earth orbit in a single launch.
It is the most powerful operational rocket by far and it can deliver more than two and
a half times the payload that the Space Shuttles could.
We could actually carry the million pound mass required to recreate the entire International
Space Station in only 7 launches of the Falcon Heavy.
We haven't seen something that powerful since the Saturn V rockets last flew astronauts
to the Moon in the early 1970s.
Unlike the Saturn V rockets, though, the Falcon Heavy is reusable and that should significantly
lower the cost of launching materials into space.
SpaceX aims to have up to 2 launches a day.
That's not to say it will be cheap, but it will mean that we can more cheaply bootstrap
a space-based industry from Earth to start mining the materials we need in space from
asteroids and the Moon.
Orbital infrastructure to accommodate those activities would mostly be places like our
Spaceport for processing all the people coming and going.
For other infrastructure, we have to ask, what you would actually benefit from building
in space.
What products are better built there rather than on Earth?
Or in-situ on whatever planet, moon, or asteroid that either needs them or extracts the raw
materials for them?
Not to mention the things than can be done in space or zero G which can't be done here
on earth.
To manufacture anything up in space implies that it has some advantage over doing it on
Earth first then shipping it to space.
That could be something like a large spaceship or station that couldn't be built on Earth
and shipped up, 100 meter plus telescopes, or something like a water recycling plant,
because it's cheaper to recycle water than import it up from Earth.
Similarly, for anything wanted far from Earth, like asteroid mining equipment, being able
to make it near Earth, but not having to drag it out of our full gravity well is quite advantageous.
However, there are features of space that are beneficial to some manufacturing.
Solar Power is vastly more reliable in space, where there are never any clouds and where
night time is very short.
The higher your orbit the less of it is in Earth's Shadow, and while low orbit satellites
spend a lot of time in Earth's shadow, it's still less than an hour, not many hours of
twilight and night like on Earth, meaning you don't need as much battery capacity.
Similarly, micro-gravity has significant manufacturing advantages.
There's no pressure in the hard vacuum of space either, so many materials typically
go straight from solid to gas, and there are some manufacturing techniques that use vacuum
and would use them more if getting a near perfect vacuum wasn't such a pain as it
is here on Earth.
One example is solar panel manufacture, which often includes a step that needs to be done
in a vacuum and this is an expensive part of the process to achieve here on Earth.
Many crystals form best in microgravity too and those have applications in industries
like medicine, optics and electronics.
Another technique that is much easier in microgravity is microencapsulation, which has applications
in adhesives, anti-corrosive coatings, pharmaceuticals, self-healing coatings, DNA protection from
degradation and sample storage.
Interestingly, many of these applications will not only be useful as export products
back to Earth but will also be very useful to the space infrastructure itself.
We are already aware, though, of a technology that we can only make in high quality and
quantities in microgravity that will be the next iteration in fiber optics and lasers.
These are called heavy metal fluoride glasses, the best known of which is a family called
ZBLANs, named using the first letters of the chemical symbols for the metal atoms that
they are composed of: Zirconium, Barium, Lanthanum, Aluminium and Sodium, that last one has the
chemical symbol Na, so that's where the N comes from.
They make the silica-based fiber optics we currently use look like a horse and cart compared
to the sports cars of ZBLANs.
They would solve many of today's problems we have in spectroscopy, sensing, laser power
delivery and fiber lasers and amplifiers.
Given that they also have military applications, there is a strong economic and political push
to manufacture these.
On Earth, ZBLANs are difficult to make in high quality because gravity causes the formation
of crystals in the structure and these crystals disrupt its desirable optical characteristics.
Doing this manufacturing in microgravity avoids this problem because crystal formation is
all but eliminated.
A kilometer run of ZBLAN fiber optic cable was 3D printed on the International Space
Station last year and this was returned to Earth last month.
We will probably hear more about that industrial experiment in the coming months.
The importance of this experiment is that it is arguably the first one to test true
industrialization in space and this shows we are on the cusp of our move into space
industrialization.
We are only at the beginning of our understanding of how to do space-based chemistry and how
we can perform industrial processes in a vacuum and microgravity, but based on results so
far, there will be many more useful and innovative processes and products that can be produced
more easily in space than down on Earth.
Now I noted at the beginning that orbital space isn't much further than many folks
commute daily, and as we saw in the Orbital Rings episode it is possible to get huge amounts
of people and material up to space fast and cheap enough for daily commutes, but more
importantly the actual commute time to low orbit is about one millisecond.
That's because you don't have to have equipment being run in orbit by people on
site, and unlike robots sent to other planets and asteroids, they don't have to be smart
either.
They just need to be something a human can decently control by remote.
So you don't even have to commute to work, you just log in at home, and even time lag
when the drone is on the other side of the planet is short enough you'd barely be able
to notice it.
However if your robot has even a small amount of AI built into it, one person may be able
to control many drones, Bots or PODs performing the most simple to complex tasks.
This addition of AI also makes large scale unmanned robotic missions much more viable
than they are today, because an AI can compensate for communications lag.
Needless to say we don't have many drones and interfaces quite up to that quality yet,
but when it comes to telepresence surgery, we have been doing that since the Lindbergh
Operation back in 2001, when a surgeon remotely removed a patient's gallbladder.
The surgeon was in New York directly controlling a robot operating on the patient in France.
Nearly two decades later, many other examples of this telepresence surgery have been performed,
but it is still not the norm because the need for it is marginal, meaning it is not widespread
or cost-effective.
Having said that, though, the technology is proven and available for when we need it.
Also needless to say that would alter a lot of Earth-based mining and manufacturing too.
Even ignoring manufacturing dangers, we have a lot of factories that would benefit from
being remotely operated in conditions that are rough on humans.
A pretty big chunk of infrastructure in space is going to be centered around humans, again
why we began our look at this concept last time with spaceports, and why we'd probably
see a surprising amount of space based hotels or even private homes.
It's not something we tend to think about much but your big issue with a private home
in space is the size needed to supply decent artificial gravity by spin, and that can be
gotten around by having many such homes on the ends of tethers like pods on a carousel
wheel.
On top of those we have 4 major sub-categories of early infrastructure we'd expect to see.
The first is power generation.
As I said, solar is more reliable in space and also very easy to beam around.
You can have panels on a given installation but you can also beam it from more specialized
power collectors as they don't need to worry about the atmosphere getting in the way and
keeping a beam on target and focused to a spot a few hundred or thousand kilometers
away in space is pretty simple.
However we can also beam that power down and out.
The down part is familiar, instead of using up useful space down on Earth you collect
power up in space and beam it down in concentrated fashion to collectors on Earth.
Likely you'd use several collectors as the satellite moves around the planet and can
re-divert the beam if you've got problems from clouds or breakdown or whatever.
The beaming out part is a little less familiar, though we've talked about that before here
and the new Project Starshot has helped familiarize folks with it too.
You can push things with light, up to truly high speeds, like the Laser Highway we've
discussed, but you can also heat or power things with such beams and that's handy
for spaceships too.
The notion of an Ion Drive is one that accelerates slowly but to a very high speed, by pushing
propellant particles out at very high speed, and that's very appealing for interplanetary
travel or moving around things already in orbit where you don't need a high thrust,
just a long slow and steady one.
It still takes fuel though, both for the propellant and whatever is giving it the energy to heat
that propellant up, and it's common to want to run an ion drive either on a nuclear reactor,
which has a fair amount of problems attached to its use, or on solar, which isn't terribly
dense, particularly as you get further from the sun, and means a ton of panels you have
to drag along and protect from micrometeors.
A beam going to a ship though, be it a ship or satellite in orbit or something moving
between planets, can provide the energy to superheat that propellant, saving you a lot
of mass, and this actually works even if you are trying to approach the beam, which is
shoving on you, because the beam carries less momentum than the propellant it will energize
does.
And while some materials are better than others, you can pretty much use anything as that propellant
if you have to.
A large percentage of the energy an ion drive uses is for ionizing that propellant, some
being much lower energy to do, so if your power is coming from an external source you
can consider propellants that are easier to obtain economically or in an emergency.
Short of us developing a viable commercial fusion reactor, solar power in space is going
to be a huge chunk of the orbital infrastructure.
Indeed it might still be preferable over fusion in many cases.
To power things in orbit, to send power down to earth, to beam power out to ships or installations
far from Earth.
It also has the dual advantage of being handy for vaporizing space junk and meteors or giving
them a shove.
Of course it's a little problematic in that you can also vaporize things down on Earth,
so you want to design all such beams to have a maximum focus spot that can't blow things
up fast down on the planet or for ones where we need a tight focus, like for hitting distant
spaceships, use those frequencies of the spectrum that are most disrupted by our Atmosphere,
or use enhanced adaptive optics to jumble the spectra when not pointing exactly where
desired.
Such solar arrays would be a huge chunk of the orbital infrastructure but look like an
even bigger portion as they'd generally be slim and wide.
Now you might be thinking that if we did enough of them they might block sufficient light
getting to the planet, but you would need a lot of those and you can also place them
in higher orbits where they get blocked by the earth's shadow less anyway, and you
can always have them with a counterweight so they can tilt on their axis when directly
between the Sun and Earth to let light pass uninterrupted.
Though you might not want to.
A problem every technological civilization is likely to face is their planet getting
hotter.
Not just from greenhouse gases but by sheer energy needs.
A few hundred terawatts of fusion power might be totally carbon neutral but it still puts
out heat.
So you might actually want to introduce solar shades to block some of that light, or some
of its less useful frequencies like infrared, from hitting Earth.
A millimeter-thick chunk of aluminum foil as big as a football field would not be big
enough to be seen obstructing sunlight to us on the ground, is pretty cheap per unit
of area, and could presumably be mass manufactured up there far more easily than anything else,
meter for meter.
They are also fairly immune to micrometeor damage for much the same reason it's hard
to destroy a sheet on a clothesline with a pistol, the bullet goes straight through without
really doing functional damage.
Done properly you get a mirror, with which you can bounce light into a close collector
dish which can then spit that radiation out as power too.
A solar panel doesn't have to be as wide as all the light it collects, if it's got
a mirror and dish concentrating that light for it instead.
So you can handle concerns about a warming planet while increasing your power supply
for beaming down to the planet, using up there, or beaming off to distant objects.
These same kind of giant mirrors can be used for huge telescopes too, though you'd want
to ratchet up the quality of the mirrors a lot.
I want to save ridiculously big telescopes for its own episode down the road but once
you are up in space, building in space, you can make things that make the Hubble Telescope
look like something you'd buy in a toy store, and able to directly image exoplanets.
Indeed you can make planet sized telescopes, thinner than paper, again something we'll
discuss another day.
Another big part of that early infrastructure is going to be fueling stations.
As I said last time, you can predict a lot of things you'd want on or near a spaceport
by looking at what we keep near airports, seaports, train stations, and truck stops
here on Earth.
You'd expect to see these as parts of spaceports but you might prefer to keep your fuel in
the same orbit but a couple kilometers forward or behind, and anything automated, be it automated
spaceships or just satellites with fuel for station-keeping, have no need to dock at a
full blown spaceport unless they need hands-on maintenance.
They wouldn't necessarily need to dock either, those fuel depots might have bots that could
fly out with fuel to them directly then come back.
You also don't necessarily need fuel for station keeping either, you could shove them
around with beams but we also have the electrodynamic tethering approach we discussed in the Skyhooks
episode, where by using a long tether with power running through it we can magnetically
push off Earth's magnetosphere.
All of which means you've got an awful lot of stuff in space.
Not just ships and stations but huge solar arrays and long tethers.
Space is big.
Really big.
Even just around Earth it is hard to believe how vastly, hugely, mind-bogglingly big it
is.
However that stuff in orbit is spinning around the planet several hundred times faster than
a car on the highway, and so each bit covers a lot of ground and even a very tiny bit can
smash into things with all the kinetic energy of a car on the highway.
That screwdriver you lost working on the outside of a space station, that floated away at walking
speed, can slam into another orbiting object on a different trajectory like a cannon shot,
and in the process throw out more debris that scatters everywhere smacking into more things,
which shoot out more debris and so on.
We see an example of this in the film Gravity.
If you've got a lot of material in orbit, it's potentially possible for a single small
piece of trash to set off an apocalypse in orbit.
Down on the ground too, since a lot of the bigger items we might put up there could have
bits and pieces fall down big enough to survive re-entry and leave a crater somewhere, possibly
someone's home.
So even infrastructure on Earth probably wouldn't survive unscathed but up in orbit you could
potentially have a total wipe out of all your equipment and personnel.
To make this worse, while a lot of that scattered debris would fall down and burn up, a lot
of it would persist in orbit for potentially many years.
This Collision cascade, or ablation cascade effect, is also known as Kessler Syndrome,
for Donald Kessler who noticed the possibility in 1978.
Forty years later it's still a serious concern.
Right now there are a couple thousand satellites in orbit, and nearly a million pieces of space
junk a centimeter across or larger floating around.
That junk kills about one satellite a year.
The more objects you add, the higher the probability of a strike, and the higher the number of
secondary debris you'd get from that collision.
At a certain density you can get a feedback runaway process that ends in everything up
there obliterated.
When it's over, everything is wrecked and you can't put anything new up for a long
time, as all that debris is still up there.
Low orbit, the place we most like to put stuff and have to pass through to get to higher
orbits, would have the most initial debris but also the shortest dwell time as the very
thin atmosphere up there causes enough drag to slowly de-orbit debris to burn up in the
atmosphere.
The higher you go, the thinner that is, and the longer it lasts, potentially persisting
millenia in the higher orbits.
You can't launch through that, it would be like jogging through a minefield.
So until it clears out, either naturally which could take generations, or artificially, your
planet is grounded, nobody is going anywhere.
Now if this happens, there are some approaches we can take to speed up clearance of debris.
You can also potentially armor your spaceships up like tanks but needless to say that takes
a lot of mass, something that's always to be avoided with spaceships, especially ground-to-space
varieties as opposed to space-to-space.
However ideally you want to be able to stop the cascade before that domino effect begins.
There's many ways to do this, either clearing it afterward or stopping it before the runaway.
Figuring out the best approach based on your technology is going to be a high priority
for any spacefaring civilization, be it us in twenty years or some Kardashev 2 civilization
building a Dyson Swarm, for which Kessler Syndrome at a solar-system wide scale would
be absolutely terrifying.
This starts with detection.
You have to start by figuring out what the minimum size is that you need to intercept,
based on your standard shielding and armor and how much pounding your paper thin solar
shades can absorb.
Now in space you can use pretty much any wavelength you want for radar, so you want something
most material is going to reflect and has a wavelength smaller than that minimum size
you need to intercept.
Then you need to flood orbital space with that wavelength with a ton of emitters and
receivers picking up every little fragment.
As I mentioned there are quite a few ways to deal with space junk but one of the preferred
ones proposed is called a laser broom and this can be deployed in space or from the
ground if you are cleaning up the mess afterwards.
A powerful laser could flat out vaporize the debris but the laser broom just heats one
side of the object up.
This causes ablation, where the material evaporates a bit, which produces a little thrust.
Knocking the material out of its current somewhat stable orbit into one which decays faster.
That's a good way to deal with a cloud of debris, post runaway collisions, but you can
also use it to nudge a bit of debris on a collision vector with an object, preventing
the runaway effect in the first place.
By embiggening the laser, we could provide a bigger shove or outright vaporize it; it's
not a good and efficient approach when you have a ton of debris but good for stopping
that from happening in the first place and another reason to have lots of power collectors
and power beaming stations in space.
It's also a good reason to include point-defense systems on a station.
Other approaches for mass clearance would include everything from shooting debris with
aerogel bullets to de-orbit them or launching a big nuke up to explode and clear a sector
right before a ship passes through, but ultimately we want to prevent the cascade from ever starting.
The general approach would be to recycle everything you can, blow what you can't into a fast-decaying
orbit to burn up, tether every piece of external equipment and tool to the ship or station,
and have little drones to run off and grab things that get lost.
One of the Gateway Foundation designs is just that, the Errant Object Retrieval Program
or EOR.
The current design for the ISS has an Observer Drone that heads over to an active work area
when an EVA begins, and is packed full of object recognition software and knows to look
for telltales of lost gear, like when it begins spinning around its own center of mass.
There's no air in space so stuff almost always tumble when it gets loose.
Later generations of Observer Drones in many different sizes will see extensive use in
space construction offering numerous viewpoints to Construction Control.
Small Observers will get in tight to inspect welding and big ones will survey the whole
worksite.
The second part of that is the amusing but aptly named "Frog's Tongue" Drone because
it roams around the station hull with the EVA astronaut and extends an armature to knab
loose gear.
This wily but versatile drone will later become the mule that carries astronauts tools and
parts in tool boxes to a work site.
Right now an Astronaut has all this gear attached to their spacesuits, very cumbersome.
And last of this trio is the Retriever Drone which has its own jets for chasing after something
that's gotten out of range of the Frog's Tongue lash, be it lost gear or even an Astronaut.
These specific designs are intended for the current ISS but represent the basic concept
of drone we will want with all our orbital infrastructure to minimize space junk, and
of course with more sophistication they could form automated repair for smaller installations,
either remote controlled or via AI.
As we do more construction up there we will need to worry about entire structural segments
getting lost too, if we're talking about a piece of truss like from Tethers Unlimited's
SpiderFab and Trusselator or Gateway's Segment Assembly Line, you don't want a big truss
whirling through orbital space, those don't even need to be going orbital speeds to ruin
your day, and you need to be looking at a large detection and retrieval or elimination
system spanning orbital space if you're doing a lot of construction and manufacturing
in orbit.
With space manufacture though we are worried about more than just lost or broken tools
and equipment causing debris, we also have to worry about space pollution.
The problem is that if we allow say, water, to escape, it could freeze into hypersonic
deadly hail.
The same is true of welded metals.
The technology needed to capture that debris at source is probably relatively simply but
has yet to be tested and it is probably well past time that we got going on preventing
this type of problem before it gets to Kessler Syndrome levels.
Another option is to reuse your old infrastructure by recycling it.
3D printing will form a big part of space infrastructure for several reasons.
Firstly, it is easier and considerably cheaper to ship a printer and the raw materials up
into space and have it print parts on demand than it is to ship up every conceivable item
you might need.
Secondly, and more importantly from the Kessler Syndrome point of view, 3D printing materials
generally allows them to be recycled.
This is being taken seriously by the various space agencies and is actually being tested
early this year.
The Refabricator Experiment is a 3D printing experiment scheduled for immanent launch to
the International Space Station.
It will process plastic feedstock through multiple printing and recycling cycles to
evaluate how many times the plastic materials can be re-used in the microgravity environment
before their polymers degrade to unacceptable levels.
Skip on any of these preventative measures and your very expensive and valuable orbital
infrastructure could go crashing and burning down into the atmosphere.
Needless to say, intentionally spraying a planet's orbit with a billion pea-sized
metal spheres is a pretty good way to cripple their orbital infrastructure and defenses,
something we will talk about more when we get to Planetary Invasions later this year,
in continuance of our series on Space Warfare.
First though we will look at Interstellar warfare next week, and contemplate how you
might actually go about fighting engagements that could take centuries and some of the
over-the-top weapons systems Kardashev 2 civilizations might employ against each other.
The week after that, we'll contemplate the end of civilization when our Planet and Sun
die, and how to avoid it, in Civilizations at the End of Time: Dying Earth.
And the week after that we'll contemplate circumventing space travel entirely with teleportation,
and ask if such a technology could ever exist and how we might do it.
We'll also look at some creative approaches and examples of its use in fiction, such as
our book of the month, Richard K. Morgan's novel "Altered Carbon"
For alerts when those and other episodes come out, make sure to subscribe to the channel.
I also wanted to thank the Gateway Foundation and Orbital Assembly for their help with this
episode and the last one in the series on Spaceports, they have a kickstarter for their
drone development just starting up, and are looking for volunteers to help on their projects.
So if you'd like to contribute some funds or time to help them out, or just learn more,
I'll include a link to that in the episode description below.
If you enjoyed this episode, hit the like button and share it with others.
You can also join us at our Facebook or Reddit Groups, Science and Futurism with Isaac Arthur,
to discuss these topics more.
Until next time, this is Isaac Arthur, saying thanks for watching, and have a great week!
Không có nhận xét nào:
Đăng nhận xét