Hacker News Reader

Hacker News

Comment by chris_va | Hacker News Reader

Full thread

chris_va

The real problem with the space economy is that there are really very few business plans:

- Space Tourism

- Imagery Acquisition

- Communications

- Science

- Astroid Mining (pending)

... And it turns out that none so far have been good businesses, and the small sat form factor is really limited. I really really love space stuff, but I don't see any of these creating explosive growth in the short term. Eventually the costs will come down, but it will probably take a long time.

*EDIT (as pointed out by pjungwir):

- Zero G manufacturing

View on HN

I keep daydreaming about what a killer business might be. A few other ideas that probably won't pan out but are interesting to think about:

- Since nuclear reactors are dangerous, what if we build them on the moon instead? Is there any way to generate energy there and get it back to earth?

- In The Moon is a Harsh Mistress the moon-dwellers fight earth by hurling rocks (maybe just threatening?), which strike with more force than nuclear weapons. What if you could use the gravity gradient to generate electricity instead? Instead of windmills we could drop rocks. :-) Of course if this actually worked we would solve global warming just in time to discover global massing.

- Or if we had a way to transfer energy, we could just put big solar arrays up there.

- I keep hoping that something will turn out to work much better in zero-gravity, like manufacturing or farming or energy production or some medical procedure.

- Since superconductors require low temperature, what if we put them in space? What if our biggest supercomputers were in orbit? What else might superconductors be good for? What if quantum computers are easier to build in vacuum?

As soon as something gets a toehold, then there will be economic incentive for mining, so that you don't have to launch all the stuff the other industries need.

> Since nuclear reactors are dangerous, what if we build them on the moon instead? Is there any way to generate energy there and get it back to earth?

Yes, but it has its own drawbacks: https://en.wikipedia.org/wiki/Reason_(short_story)

- Zero g manufacturing (good edit, has some potential)

You have something with the spaced based solar (SBSP). The unsolved problem there is the microwave power beaming, which has some technical hurdles (synchronizing microwave frequencies across multiple dishes is apparently hard, but necessary for power beaming, and there is an ozone problem). Any other space based power system has this same transmission problem (people have proposed attaching conductive tethers to orbiting NEOs to drag through the magnetic field).

The "nuclear reactors are dangerous" bit hasn't really been true in a long time (Chernobyl was the worst disaster, but killed a lot fewer people than hydro/coal/etc).

Space is actually not great for computers. Dielectrics change in vacuum, and cooling is actually a big problem (turns out space isn't actually "cold" in the traditional sense, it is insulating, and heat tends to build up). Quantum computers are probably a red herring, there is vey little difference on an atomic scale in/outside of micro-gravity.

Ha ha I knew someone would point out that vacuum isn't "cold", because lack of matter means heat can't transfer away from hot things. My comment about superconductors and computers tossed a bunch of stuff together I probably should have separated. I agree space wouldn't help for cooling conventional CPUs. Would it still help for superconductivity? (I have no idea.) Also for quantum computers I was wondering if vacuum would have any benefit for decoherence.

Even if you solve the synchronization problem, you still have the main problem with power beaming of the massive sizes needed of the transmitter and receivers at space scale distances, thanks to diffraction.

Microwaves need prohibitively large transmitter/receiver arrays if you go beyond LEO (drag is an issue in LEO). While putting a kilometer^2 scale transmitter in orbit isn't a problem, putting the similarly sized receiver for it on the ground is. At optical or smaller wavelengths, the arrays can be much smaller, but they will still be large structures.

Disclaimer: I worked with a company that does terrestrial optical power beaming.

> putting a kilometer^2 scale transmitter in orbit isn't a problem, putting the similarly sized receiver for it on the ground is

Wow, I would have guessed the exact opposite. What's the reason(s) for the higher difficulty on the ground? Planning permission? Cost of the land? Needing multiple sites in multiple countries for non-geosynchronous orbits?

> Disclaimer: I worked with a company that does terrestrial optical power beaming.

That sounds fascinating. Could you elaborate? What's the power level and the application?

Building the structure itself isn't too hard as it will be mostly empty space. For microwave it will be an area packed with odd looking antennas, optical is basically an off the shelf solar farm (though the cells should be tuned for the application rather than general sunlight).

The hard bit getting all the regulatory ducks in a row, as you will need ground->space no fly zones, environmental approval, not be sited near anything else in case of alignment/focus/astigmatism (side lobes for rf) issues, etc. Generally out in the middle of nowhere there is some endangered animal (or its migration pattern) that will cause no end of issues regulatory wise for something this big/dangerous to wildlife unless you keep the energy density down below that of the sun (at that point you might as well use normal solar). You will also need controls to rapidly shutdown the system (or shunt to new location) in case an aircraft violates the no fly zone or another satellite is about to pass between the transmitter and receiver.

> That sounds fascinating. Could you elaborate? What's the power level and the application?

What they did when I was there was fairly diverse. The biggest was delivering 1kw at a distance of 1km with ~5-10% plug to plug efficiency. The primary application for that system was inflight refueling of electrically powered drones. They normally do much smaller/simpler (10w over dozens of meters) systems for oddball applications (eg: sensors that are sitting on something charged to 500+kv, etc) and power over fiber-optics.

> Since nuclear reactors are dangerous, what if we build them on the moon instead? Is there any way to generate energy there and get it back to earth?

Any way of bringing down lots of beamed energy from the Moon could be also used as a weapon of mass destruction[0], so I imagine we first need to sort out terrestrial politics before this has a chance of catching on.

[0] - possibly quite surgical, as WMDs go.

I think it's housing, although there's a big hurdle to make it valuable since you need stuff for people to do first.

> - Since nuclear reactors are dangerous, what if we build them on the moon instead? Is there any way to generate energy there and get it back to earth?

My pet scifi idea about it: turn Mercury (rim) into a giant powerplant, use it to create tons of antimatter, ship it to earth. Containment might be tricky tho :)

> Is there any way to generate energy there and get it back to earth?

Not only nuclear, but space is a great place for solar as well - 24-hour sunlight, with no clouds or rain. That said, it's pretty prohibitively expensive unless we can build them out of silicon-based asteroids.

Tourism hasn't really paid off. Imagery seems to be doing OK, but is a fairly small business. Science isn't doing much for business at the moment, and asteroid mining is still theoretical, so I generally agree on those.

But communications? There are how many hundreds of commercially operated communications satellites in orbit right now? From a quick perusal of SpaceX's launch history, communications makes up nearly half of their business so far. Seems like pretty good business to me.

The problem with communication satellites is that the technology isn't as good (throughput, latency, cost) as a ground based system, except for handling coverage gaps.

Coverage gaps on the ground are slowly going away, so eventually this business will likely decline.

Now you're changing from "so far" to attempting to predict the future, which is rather different.

As far as the future goes, throughput and latency both improve dramatically with low-orbit systems. Cost is high (see the massive failure of Iridium to see how that can go) but is coming down pretty rapidly right now.

And given how many satellite dishes I see on buildings in cities, it seems that there are use cases beyond handling coverage gaps.

Mostly fair, but...

Latency to space is always going to be slower, as you are going to add a few hundred miles, and power requirements for high-bandwidth applications require large solar panels which are incompatible with low-altitude sats (too much drag). Then there is also a problem with non-geo sync orbits and directional antenna tracking. This is actually a huge problem. Current small sats point directly at their ground stations as they go over, which means they are only talking to one receiver at a time. There are some meta-material antennas that I have seen that might solve some of these problems, but not fully.

Any satellite dishes you see on buildings, right now, are pointed at a geo-sync satellite instead of LEO. Half of those are now pointed at a ground-based microwave receiver instead, and the current trend is away from satellite.

Anyway, time will tell, I guess.

Lower launch costs make drag less of a problem, since you can give the satellites more fuel for reboosts. A few hundred miles of latency is no big deal. 200 miles adds one millisecond of latency. The speed of light is noticeably slower in fiber, too, so that balances out once you get past a certain distance.

Talking to one receiver at a time is clearly not an inherent limitation of low-orbit satellites. Iridium satellites can handle multiple simultaneous connections without directional antennas on the ground side, and with a phased array on the satellite side.

I'm well aware that dishes are pointed at geosync satellites. The fact that there are so many in areas with good coverage from terrestrial radio and cables means that there are applications beyond filling gaps, even with the latency that geosync implies. I don't recall ever seeing a satellite dish pointed horizontally, and certainly all the little ones I see on houses aren't, so I completely don't believe that "half of those" figure.

Sorry, didn't mean to imply that your argument was wrong.

For Iridium, the total bandwidth per satellite is really really low (like 28.8? Newer ones might be higher), and they had to put them slightly higher up (500 miles, I think?) to avoid drag issues since they needed bigger solar panels.

The latency thing is more compared to microwave beaming, which is how most cell backhaul is done now, but you are right that the extra few hundred miles isn't critical (my head was still in geo-synch).

The current Iridium satellites are ridiculously slow. They were designed in the 90s for voice, and having any data capability at all is sort of an afterthought. One "call" is a whopping 2400bps. Each satellite can support 1100 calls simultaneously, but even if you ganged them all together that's still only 2.6Mbps total.

Iridium NEXT is going to start launching in September and will have vastly better capabilities, up to 8Mbps per connection. Still not great compared to a terrestrial wired connection or even modern cellular connections, but a lot better. If launch costs drop significantly, you could presumably do a lot better still.

>As far as the future goes, throughput and latency both improve dramatically with low-orbit systems.

The problem is that they still pare in comparison to ground based systems. On LTE, with aggregation, you can get something like 100 Mbps out of a cell phone on about >90% of the US's land area with Verizon's network. Irridium Next which doesn't even start launching until September will be limited to 8 Mbps and won't be completely ready for another couple years. Two-way LEO satellite communication will only be worth it to people on the oceans very soon. The satellite dishes on buildings only work with one way communication.

I would agree this is generally true. There isn't anything that is easier to do in "space" than it is on the planet. But I think that might not be as true for other planets.

You could, for example, put radioactive waste in an open pit on the moon. Nobody is going to "accidentally walk by it unprotected" as the radiation levels on the surface of the moon are "naturally" higher than anything we've got in our waste pile due to a lack of atmosphere between the Sun and the Moon.

Now the first thought would be "Gee Chuck, sending a Kg to the moon is pretty freakin' expensive." But if you pause and consider the money already spent on Yucca mountain and other projects, and the fact that it still isn't operational, and the on-site nuclear waste storage issue continues to make nuclear plants riskier rather than safer, then is is too freakin' expensive? Remember you don't have to "land it gently on the moon" you could literally drop it from orbit if the impact plume was pointed away from the earth.

Solves a real problem perhaps? Perhaps not. Hard to say definitively.

If you can get it to he moon, may as well put it in the sun :).

There is a slight problem with the x% chance the rocket will blow up at max-Q while still in the atmosphere.

Plus, all nuclear waste ever generated would fit in 1 football stadium. The rest of the spending is just due to political infighting that looks good to voters.

Yes, there are risks. Just saying its a defensible "permanent" solution to disposing waste with a very long half life.

As far as using the Sun as an alternate, it takes less delta-v to get to the moon than it does to get to the Sun.

Further, the overall effect of adding nuclear waste to the Moon is completely predictable in its impact, but that is less true of adding it to the Sun.

I think this isn't a viable option because you would need 100% certainty that your flight will not fail. An explosion in the atmosphere with nuclear waste involved sounds pretty devastating. This is the same reason that nuclear power should be avoided right now. Even if the chance of disaster is tiny, the consequences are far too large.

It seems paradoxical. Access to amounts of matter and energy that would be absolutely transformational to the world (solar system) economy... but no economic case to go and get them?

Perhaps this is something with a huge payoff that doesn't make investing sense on anyone's time horizon because it's longer than all our lifetimes.

Still, long-lived entities (such as nation states) do exist - depending on political conditions, projects on multi-lifetime timescales can be viable. One would like to think that our far-sighted leaders would make some investment, given there's an awfully big prize up for grabs.

Nobody has yet mentioned Helium-3, though it's related to your asteroid mining point. Helium-3 is used in fusion and seems to be the "free energy" replacement for petrochemicals that we need. China and the USA have both stated that they have sights set on Helium-3 mining on the Earth's moon, though China appears to be much more serious about it than the United States. I personally believe that Helium-3 will be the thing that causes a gold rush into space.

There is only one business plan for the 'air industry', "transportation", which was totally unforeseen in 1905, yet it seems to be quite large.

Don't get me wrong, it will get there eventually, but the current market economics aren't exactly primed for explosive growth.

Asteroid mining is a support business, not an end goal. It's more like selling stoves, it may be lucrative and a big field, but won't make people go to the middle of nowhere by itself.

Some day we'll probably start to send supercomputers into space, but there isn't really any kind shortage here on Earth that this would solve right now.

Reply on news.ycombinator.com