In addition to lack of solar energy, I wonder if nuclear power source is necessary to keep batteries or other electronics warm enough to function properly.

Conversely, on Titan, we won't need to worry about heat sinks as everything is kept at ~94K in a bath of liquid and frozen methane. Perhaps in a few hundred years AWS's most popular region for intensive compute will be under Titan's methane sea. High spin up cost, but cheap super-conduction environment once you're up.

Interestingly they actually exploit these cold temperatures to keep things cool. The Dragonfly Gamma-Ray and Neutron Spectrometer(DraGNS)'s gamma ray detector is mounted outside.[0] In most other probes the detector would need to kept cold with a cryocooler. [0]https://www.jhuapl.edu/techdigest/TD/td3403/34_03-Lorenz.pdf

Hah, I was always intrigued by the idea of placing solar cells on the bright side of Mercury and servers on the dark side. I wonder what other good options might exist in the solar system.

Mercury's whisper thin atmosphere won't be much use for cooling anything, though possibly the surface would be. However, the real problem is that Mercury isn't actually tidally locked anyway.

In certain regions on Mercury the temperature stays within 1 degree of room temperature less than a meter underground. http://einstein-schrodinger.com/Mercury_temperature.pdf It’s actually relatively habitable. http://einstein-schrodinger.com/mercury_colony.html

> In fact, if we delay until the distant future, or even 50 years or so, such an effort probably will become impossible. This is because us humans will consume the Earth's non-renewable energy and mineral resources almost completely within the next 50-100 years, severely reducing our discretionary income for costly activities such as space travel.

I wonder when that was written...

Likely around 2013 by the Wayback Machine.

Makes me think of the Gell-Mann amnesia effect.

You could have the facility move constantly to keep the facility straddled across the sunlight line. At the equator, this would require constantly moving at about 7 mph, though you could reduce the speed necessary by moving to higher latitudes.

Appears in 2312 by Kim Stanley Robinson.

They move the whole city around the planet at 7mph!! 10/10 would recommend

Sounds like Absolution Gap and/or a less annoying version of the Mortal Engines movie (please tell me the book was better than this).

Also features in Charlie Stross, Saturn's Children.

Or another scifi novel, West of January by Dave Duncan.

Also appears in Asimov's classic "I, Robot" collection, as a Powell and Donovan story called "Runaround"[1].

[1]: https://en.wikipedia.org/wiki/Runaround_(story)

To add to the vein of sci-fi novels using this idea: Inverted World by Christopher Priest. Well, similar concept but weirdified in the 1970’s fashion.

Yeah but if you break down...

Like in the scifi books?

Reminds me of the book Absolution Gap

Mercury doesn't keep the same side to the sun all the time. So you'd need to either bury everything deep enough (obviously not possible for solar cells), or keep them such that they're always on the terminator. The latter has been explored in Kim Stanley Robinson's work, 2312.

https://en.wikipedia.org/wiki/2312_(novel)

The poles of the moon. Have 3 "legs", and the alternate being the hot and cold areas.

Yes, you need equipment that can swap being hot and cold, but it's a lot easier to get to :)

mercury-dark-1

I find it kind of funny that you think AWS will still be around in a hundred years.

This paper company was founded 771 AD: https://www.shinise.ne.jp/j_kyoto/voices/vol/48/48-d.html

It's not even the longest-lived company in the world. This construction company was founded in 578 AD and continuously operated until 2006 when it was bought by a larger company: https://en.wikipedia.org/wiki/Kong%C5%8D_Gumi

100 years ago was only 1919 -- here are some random companies from 1919 that I found on Wikipedia (ignoring for a moment that this is plainly surviorship bias):

https://en.wikipedia.org/wiki/Champion_(sportswear)

https://en.wikipedia.org/wiki/KitchenAid

https://en.wikipedia.org/wiki/Thrifty_PayLess

https://en.wikipedia.org/wiki/Tesco

https://en.wikipedia.org/wiki/Proffitt%27s

https://en.wikipedia.org/wiki/Bentley

https://en.wikipedia.org/wiki/A%26W_Restaurants

Companies surviving for 100 years isn't quite routine but it's certainly not so rare that its surprising when one does. I don't think it's at all inconceivable that Amazon could stay open for a hundred years. It's made it past the first 10 years, the turbulent start when most companies are culled. I'd wager that the biggest threat to Amazon surviving to 100 is being broken up a la Bell/Standard Oil.

The Avedis Zildjian Company, now an American-based cymbal manufacturer, was founded in Constantinople in 1623. Hell, even Nintendo's 123.

https://en.wikipedia.org/wiki/Avedis_Zildjian_Company

A company lasts just 11 years on average on the S&P 500 now. Companies are surviving for shorter time periods now then they used to.

First off, what is your source for that average? This report (https://www.innosight.com/insight/creative-destruction/) has a much higher average, and forecasts a fall to 12 years by 2027.

Second, that's average time in the S&P 500 index. As noted and n the report, there are many reasons besides going under that companies drop off the index. That number doesn't really say much about longevity of the company.

Like with pre-modern human lifespans, there's a VERY long tail on that number.

For ones maybe more familiar to an American audience and/or to techies:

Bayer (1860s Germany), Johnson & Johnson (1886), 3M (1902), the Coca-Cola Company (1892), IBM (1911), HP (1939), Mitsubishi (1870), Wells Fargo (1852).

The italian bank Monte dei Paschi di Siena was founded in 1624, they even use some of the centuries old building as offices...it's not doing very well but it's still there

The oldest preserved share of any company is from 1288. It's since been merged, but it's still around.

https://en.wikipedia.org/wiki/Stora_Enso#History_of_Stora

Even after being broken up, AWS would still exist - it just wouldn’t be part of the amazon conglomerate

AWS is different from Amazon though. AWS is a cutting edge technology in a fast moving industry that is likely to be made obsolete in the next 20 to 30 years.

Heck, IBM is 100+ lol

That's a good point - the "waste" heat from an RTG (which is only about 5% efficient) is very useful in those extreme cold temperatures.

NASA makes a smaller RTGs for heating tasks, they can scale smaller if they don’t need to generate electricity.

Super fast compute, but awful ping…

To put it into perspective:

https://solarsystem.nasa.gov/moons/saturn-moons/titan/overvi...

According to this (at the time of reading) a one way trip at the speed of light is ~75 minutes, so you will have approximately two and a half hours of ping.

Ouch.

Still, it raises an interesting question: how would a hypothetical Earth-Titan network protocol work?

That protocol exists, it is called Delay-Tolerant-Networking.

http://ipnsig.org/wp-content/uploads/2015/05/Whats-new-in-IO...

Can't do much about the time it takes light to get here, so I guess the answer is we'd have to learn to love 75 minute ping :)

Alcubierre warp drive powered Sneakernet?

If you can't be with the one you ping, ping the one you're with.

The InterPlanetary File System is a thing, it just isn't interplanetary yet: https://en.wikipedia.org/wiki/InterPlanetary_File_System

Batch instead of stream I'd reckon.

Perfect for mining bitcoin

Not really; if you find a block it will mostly likely have been beaten by someone else's block by the time yours reaches most nodes...unless there is a Saturn based side chain or similar.

This will sound crazy, and I'm not suggesting the consumer would deal with it directly, but I think we'll need coins for each planetary body and another one for the solar system that has a much longer block time. More scales would be useful, but these are the minimum. Repeat for each star, with MilkyWayCoin having an absurd block time.

It depends if the "most nodes" are on Mercury or on Earth.

Might be suitable for quantum computing ... Re. thin atmosphere and low gravity, humans could probable be able to fly wearing wings. So if we put a base there it would not be boring.

Never thought about this before, in theory does this bring super conductors closer to reality?

Why do you think they launched a deep space reviving network at re:invent last year?

Well we don't want to disturb the environment too much

The data rate from the surface of titan is abysmal. Even with a relay, we won't get much video, but man, I hope we get some.

From a scientific standpoint, there's really no value in downlinking video as opposed to still images. It costs several multiples worth of scarce downlink bandwidth, for a tiny gain in what there is to see of the same rocks at slightly different angles. Same reason we didn't have New Horizons at Pluto or the various Mars rovers do any video.

There might be some scientific value for taking video. One measurement where video might be valuable is measuring the saltation threshold, or the wind speed at which grains begin to move, of the surface using rotor downwash. These measurements are important to understand the dune pattern on Titan. This could be pretty weird as the dust on titan is believe to be made of something similiar to moth balls, which could stick to each other like packing peanuts due to frictional electrostatic charging.

[0]https://www.jhuapl.edu/techdigest/TD/td3403/34_03-Lorenz.pdf [1]https://www.news.gatech.edu/2017/03/27/electric-sands-titan

Why is bandwidth so scare still? Satellites in geosynchronous orbit on earth can do terabit/sec of total capacity now.

Obviously latency is going to be extremely high, but I can't really understand why data rates are still so low?

Inverse square law. For the same transmit power, your received signal strength diminishes with the square of the distance

That received signal strength is what determines channel capacity (C) which is the upper limit on the bit rate you can get.

see Friis and Shannon's channel capacity

[0] https://en.wikipedia.org/wiki/Friis_transmission_equation

[1] https://en.wikipedia.org/wiki/Shannon%E2%80%93Hartley_theore...

Yes, I get that. But it seems that performance has barely improved since the 1970s despite the massive gains in telecommunication efficiencies since then. Or is the SNR so poor that even 1970s era technology was able to be close to the limit?

These are physical restrictions, not technological.

What about the laser technology NASA is developing? That should be a lot faster than current radio based technology?

Titan has smog in the visible spectrum. Details matter! The technology is also not that mature, yet.

Focusing a laser at that range requires a rather large diameter heavy set of optical components.

You will need a blinding-strong laser.

Because the signals have to travel so much farther and are so much weaker when they arrive. It's amazing that any data at all can be extracted from a radio signal that travels a billion miles.

Spacecraft transmit power and antenna size is limited. This means the amount of power that arrives at earth is rather low because microwaves spread out quite a bit with distance. The other issue is that NASA only has so many big radio dishes in the deep space network[0] to gather this small amount of power and extract a signal. Technology has improved, but in order to get much higher bandwidth we'd need bigger dishes. We also need at least three around the globe to get continuous coverage. Although perhaps by the time this mission lands we could have multi-kilometer aperture receivers on the ground or in space. Laser communication could allow for much higher bandwidth, because lasers have a shorter wavelength they don't spread out as much with distance. However, laser space communication is in its infancy and would not work well on Titan due to the smog.

[0]https://en.wikipedia.org/wiki/NASA_Deep_Space_Network

Maybe workaround transmission power by reflecting and modulating a signal over an existing power source. Like a very big mirror with an LCD pixel in front of it.

"one over r-squared"

Latency is high, but the power you have to pump through an antenna / laser to get enough watts on the receiver goes up really fast with distance.

Here's a summary that touches on it if you'd like to learn more: http://www.qrg.northwestern.edu/projects/vss/docs/Communicat...

And here's a detailed derivation for voyager. https://space.stackexchange.com/questions/24338/how-to-calcu...

I don’t think you realize just how far away the outer solar system is.

Low transmit power. Long distance. Very low SNR at receiver.

Don't get your hopes down! It's still very possible that we'll get some good, if not real-time, video out of this!

Ah, we likely will not get real time video. Maybe APL has some tricks I'm not privy to. Source: I asked a DSN coms engineer.

There will be at least an hour delay. Saturn is pretty far

10 seconds of video from Titan would change the world. It's shame we don't have this from Mars already. So it would take a few weeks to transmit - we got time.

Now figuring out which 10 seconds to select and send to earth... that'll be interesting. An on-board cache of x GB data storage, with key frames taken every 30 minutes, would probably be able to select some pretty compelling 10 second segments.

What about the world would be different after getting a 10 second video from Titan?

It would be inspiring for me at least.(Assuming something was moving)

Send them all, just pick a likely-nice one for sending while the eyes of the world are on Mission Control.

Caching everything then sending it at leisure is working nicely for New Horizons.

No 4K flyovers for Instagram?

I mean sure, eventually, over many down link cycles. But not real time.

Space travel just makes it utterly clear there is on such thing as real time. Computing is constantly running up against the speed of light.

It's not about latency, it's about power required to transmit. From far away, you need to push lots of power through to get a readable signal on the receiver side. It's kinda-sorta like seeing a flashlight.

In a sense tho, at-the-speed-of-light is real time, as it's the maximum speed of anything.

If, as clock A nears the speed of light it appears that stationary clock B is slowing down, doesn't that mean that at the speed of light is no time? Does a photon in a vacuum arrive instantly at its destination? If a photon looks at another photon, does the other photon appear not to be moving at all, or... Ouch my brain.

> If, as clock A nears the speed of light it appears that stationary clock B is slowing down, doesn't that mean that at the speed of light is no time?

If clock A could reach the speed of light (which it can't), then, yes, clock B would appear to be 'frozen in time'.

> Does a photon in a vacuum arrive instantly at its destination?

From the perspective of any possible observer, no, it always appears to be traveling at the speed of light. If an observer could travel at the speed of light (which it can't), then, yes, it would appear as if it arrived instantly in the sense that clocks at both its source and destination would appear to be not running at all.

> If a photon looks at another photon, does the other photon appear not to be moving at all, or... Ouch my brain.

Yeah, this is tricky. Photons can't look or see – what could that mean as looking/seeing involves detecting photons (or, in the sense that something like echolocation is 'seeing', detecting a pattern of matter, e.g. sound)? In a sense, not moving at all or both moving at the same maximum speed don't seem to be much different, from the 'perspective' of a photon.

Somewhat related, this video describes how, if one could travel on a vehicle that could accelerate at a constant rate continuously for decades, one would eventually see the cosmic microwave background radiation as a rainbow ring because traveling closer and closer to the speed of light would shift the apparent frequency of the radiation first into the visible spectrum (and then beyond it).

My comment was mostly tongue in cheek, I know photons don't have eyes. :)

Not completely related, but you reminded me of something interesting I remember hearing about neutrinos. Apparently (HN physicists please correct me if I'm wrong), neutrinos have an extraordinarily tiny objective lifetime, and the only reason we can detect them is that, because they travel so close to the speed of light, they experience almost no passage of time in while in flight.

I kind of wish it had an airship counterpart. Inflating a dirigible there would take advantage of the same atmosphere + gravity benefits with the added bonus of not needing constant power to move around.

Good point: I seriously wonder if they considered a dirigible design. Buoyancy + directional propulsion does seem simpler in my space-novice opinion compared to a quadcopter. A blimp would be slower than a quadcopter, but nobody's going to be piloting the thing in real time, so that shouldn't matter.

Yes, that was a mission proposal years ago, a blimp and quads. It was not selected then or rehashed this time. In addition, a blimp only concept has been floated (ha!) as well.

Blimp only wouldn't be able to take ground samples easily

How does one test on earth, a drone for a different atmospheric and gravitational environment? Some great drone simulation techniques are going to come out of this project.

No problem. Large atmospheric chamber or controlled wind tunnel. Pump the pressure up or down, change the gas mixtures, whatever. Gravity, partially support the drone's weight. NASA is good at this stuff.

Yes that's how they tested the Mars drone.

https://www.jpl.nasa.gov/news/news.php?feature=7417

Will this be the first-ever instance of nuclear-powered flight?

No.

>> The Pentagon is aware of a Russian test of a nuclear-powered cruise missile but the system is still under development and had crashed in the Arctic in 2017.[17][18][19] A RAND Corporation researcher specializing in Russia said "My guess is they're not bluffing, that they've flight-tested this thing. But that's incredible."

https://en.wikipedia.org/wiki/Nuclear-powered_aircraft#Russi...

https://en.wikipedia.org/wiki/9M730_Burevestnik

I hadn't heard of the 9M730. That page links to one for the US Air Force's Supersonic Low Altitude Missile (SLAM) or "flying crowbar" of the 50s and 60s, which makes for fascinating reading: https://en.wikipedia.org/wiki/Supersonic_Low_Altitude_Missil...

This case is different though, using the fission reaction to generate electricity which in turn powers the electric motors among other things.

The nuclear powered missiles use the fission reaction to heat incoming air similar to how a jet engine works.

No fission reaction, just radioactive decay in Dragonfly's case. It will use an RTG (radioisotope thermoelectric generator) not unlike Curiosity and most missions to the outer planets to date.

Just a note - radioactive decay is is fission, since it involves a nucleus splitting apart.

No, not quite. One of the types of radioactive decay is spontaneous fission, but it's not relevant in the context of RTGs. Alpha decay (the emission of a 4He nucleus) might in principle be called "fission" but in practice is not. Beta decay involves no splitting whatsoever.

RTGs are heavy; I suppose this means that it'll park itself for a long time to charge its batteries and then do a quick flight and repeat. Which sounds like it could work pretty well for this use case.

On the other hand, a hot-air balloon kept aloft by excess heat from radioactive metal might be a more attractive option if we were selecting proposals based on awesomeness.

Does that mean a nuclear powered cruise missile could stay airborne until mechanical failure? That would make future warfare... interesting.

It isn't all that much of a game changer. Mid-air refueling means that bombers can stay airborne a long time. Four or five bombers rotating back and forth can maintain one "on station" continuously. So the capability is already there. This tech would just make it more scifi.

Why bother? Rocket fuel is cheaper than nuclear fuel and rocket missiles can already hit anywhere.

Because a nuclear-powered missile can in principle fly in circles at low altitude off your enemy's shores for months at a time without landing or refueling, ready to be called down at any time with no detectable warning.

Well, if I am your enemy, I sure as hell won't let your missile cruise around my shore for a month

The difference between rocket fuel as a power source and nuclear radioactive decay as a power source is like comparing C4 explosive and a log fire in your fireplace. Once it great for providing a huge, fast release of energy, the other is great for providing a slow, steady release of energy.

Cheaper, easier to mass produce, doesn't need an unusually special launcher, hard to distinguish from air to air and other defensive missiles, easier to disguise the number of warheads waiting in readiness, can take an indirect/stealthy course, hard to shoot down, can include smart tracking for moving targets like carrier groups.

Maybe so it can double as a shitty nuclear bomb. I'm gonna crash my small nuclear reactor into you.

NASA has been using nuclear powered probes for 50 years. It is called a redioisotope thermoelectric generator (RTG).

https://www.space.com/13709-plutonium-shortage-nasa-planetar...

Unfortunately, there is a shortage of plutonium-238 to run these RTGs (see above story). I guess that will get resolved somehow.

The current Mars Curiosity rover is powered by an RTG. Since this is the first flying (post-landing of course) probe I suppose it will be the first nuclear powered flight of a planetary probe.

DOE has restarted Pu238 production. They're only making about a pound a year but they're still scaling up

America and Russia have both put proper fission reactors into space too.

Depends on your criteria, I suppose. https://en.wikipedia.org/wiki/Operation_Plumbbob#Propulsion_...

Probably a dumb question, but how does the nuclear source work? Nuclear powerplants work by heating steam to turn a turbine. This drone won't have heavy water reservoirs so how does it convert the fission heat into electricity? Stirling engine?

Most nuclear power sources for space stuff works by Peltier effect. See Radioisotope thermoelectric generator.

radio-isotope decay -> heat -> Peltier stack -> electricity.

Advantages: no moving parts, no vibration, infinitesimal mass change, exceedingly reliable. 'Small'

Disadvantage: Efficiency kinda crummy at ~5%. Slowly loses output as fuel decays and the Peltier stack degrades.

Probably an RTG [1], it's what NASA normally uses. It's thermoelectric [2]: Heat (in this case from radioactive decay) is converted directly into electricity without moving parts.

[1] https://en.wikipedia.org/wiki/Radioisotope_thermoelectric_ge...

[2] https://en.wikipedia.org/wiki/Thermoelectric_effect#Seebeck_...

Just like nuclear sources work on sattelites. You produce heat with fission or nuclear degradation and then thermoelectric conversion.

Why arent we just mass launching probes to all the planets?

Serious question: Why is it not OK to have a standard sensor array on every planet in the solar system?

To the second question: since each mission is expensive and time-consuming, scientists reasonably want the best sensors for each mission.

That's both the tech upgrade during the n-year coast phase, and the investigation upgrade as they work out new things to investigate.

It's not as straightforward as spamming the system with cameras. That's bound to happen more, after the MarCO cubesats had success flying along with Insight.

> Why arent we just mass launching probes to all the planets?

Because we don't live in a fictional sci-fi universe where launching things into space doesn't cost a ridiculous amount of money, time and effort, and can't only be done at certain times from certain locations in order to get the orbits right, or where years of effort and millions of dollars (or whatever currency) don't also go into building the things to be launched into space. The world has better things to spend its money on.

>Why is it not OK to have a standard sensor array on every planet in the solar system?

What is a "standard sensor array?" We've already sent a lot of probes to a lot of places[0] around the solar system. But read the rest of this thread - it's a lot more complicated than it seems. Also, remember how hostile and difficult the solar system can be. Jovian planets don't even have a solid surface to be "on," for instance, and Venus has a pressure of 90 atmospheres and it rains sulfuric acid.

[0]https://en.wikipedia.org/wiki/List_of_landings_on_extraterre...

What about hydrogen fuel cells? Isn't the whole place basically nothing but fuel?

it's all fuel, but there's no oxidizer

Those work by combining H with O. Not much available oxygen there.

is there no wind on Titan? What if there's a windstorm and it crashes the drone? At least a vehicle could resist crashing, unlike a drone.