Distributed Probing?

August 18, 2007 8:05 AM

Why hasn't anyone invented cheap space probes? Some kind of thing that could be sent up en mass to gather just base data we don't have on all the planets/moons/etc . . .?

So I've been watching the Universe series on the History Channel. One point that keeps coming up is that they don't know a lot of things about certain planets or moons because we haven't sent a probe there yet, and probably won't for a long time.

That got me thinking, why hasn't anyone, NASA or otherwise, developed "cheap" space probes to do basic information gathering? Relatively speaking of course. Instead of spending millions on a probe, maybe a few hundred thousand. Something that gets it to its destination, takes a bunch of pictures and sends those pictures home.

I'm sure I'm being naive asking this. Please enlighten me!

posted by [insert clever name here] to science & nature (34 comments total)

Well in a way we have invented them. Several recent probes are updates of old designs.
posted by Mister_A at 8:16 AM on August 18, 2007

The cost of lifting something into orbit is immense, especially given the risk involved. There is no fuel in space, so all fuel must be carried with the rocket. That adds weight to the rocket, requiring more fuel.

Given that if a rocket fails, your billions of investment and years of work are all done, you can see where the problem starts.
posted by Ironmouth at 8:23 AM on August 18, 2007

The phrase you're looking for to google is "Faster, Cheaper, Better".

To be successful, space equipment must:

get into space (very expensive)

be reliable with a low failure rate (expensive)

be hardened against radiation (ditto)

use electronics that aren't available at Radio Shack.

go through government contracts (2 screws = $10000)

posted by blue_beetle at 8:30 AM on August 18, 2007 [1 favorite has favorites]

Ironmouth has it right - even for a vanishingly small probe mass, the differential equation that describes the fuel needed to lift it grows quickly. To fake some numbers, it takes x fuel to lift the probe, and 0.8 * x fuel to lift that fuel, and 0.8 * 0.8 * x fuel to lift that fuel, to practical limits.

Once you're dealing with that much fuel, you've already bought quite a bit, so why not go ahead and buy an entire tanker instead of half of one? The marginal costs are quite small. And once you've got that extra fuel, why not send up more stuff and reduce the cost per unit mass of the final lift?
posted by TheNewWazoo at 8:30 AM on August 18, 2007

I work in the space launch business... Ironmouth has the right idea. Even if you could develop and build a probe for a few hundred thousand dollars, it'll still cost you hundreds of millions of dollars to launch it and otherwise support the mission. There's a lot of time, people, and expensive hardware involved in putting stuff in space.
posted by jal0021 at 8:33 AM on August 18, 2007 [1 favorite has favorites]

For few hundred thousand, you get something you can't launch. Accelerating a kilogram up to interplanetary speeds is your-head-asplode expensive, even after NASA became ludicrously good at orbital slingshots.

IANARocketScientist. My sense is that a lot of the expense is the launcher and launch, and rational economizing because the launch is so expensive. If the launch is really expensive, it makes sense to spend lots of money lightening and optimizing the probe so you can do more science -- if you have a choice between a $1 billion total cost probe with 5 experiments and a $1.1 billion total cost probe with 8, that's easy. Likewise, anything that costs that much you won't get to do often, so it's often sensible to spend a lot of money on reliability and testing.
posted by ROU_Xenophobe at 8:36 AM on August 18, 2007

Why do it cheap when the government will do it for free (to you)?
posted by LarryC at 8:40 AM on August 18, 2007

That got me thinking, why hasn't anyone, NASA or otherwise, developed "cheap" space probes to do basic information gathering?

I don't think that's the real question. It's more "why don't we consider these things worthy of spending money on". Taking the US as an example, we spend a lot of cash on various things, almost half a trillion on defense, but only 16 billion on NASA, most of which goes to the Space Shuttle. We have the money to spend, but clearly our priorities are elsewhere.

But yeah, we really should have several probes around each planet, studying them. When I'm dictator, this plan will be put into place, along with no pants Fridays and abolishing Mondays.
posted by Brandon Blatcher at 8:41 AM on August 18, 2007

Freeman Dyson wrote about space science and where the money is being put and why in chapter 9 of his book "Infinite In All Directions". Written in 1985, but still applicable today. Basic message is that for pure science, money could be much more efficiently spent and that the space program as it exists is not really all about science.
posted by DarkForest at 8:54 AM on August 18, 2007

Hey, it's not like just taking photos is that easy. You have to:

1) Get the camera in the right spot. Tricky and expensive, since you need a booster from earth orbit. And you need all the equipment to know where you are and whether it's the right place. No GPS in space.
2) Point the camera in the right direction. Tricky and expensive, since you either need to do it Hubble Space Telescope style with gyroscopes, or with thrusters.
3) Get the photos back. Tricky and expensive, since you are far away, and you need a powerful enough radio and an antenna, plus you probably should point towards Earth too.

Plus you have to power this thing, which, as you might expect, is tricky and expensive. And you need to do it all in a clean room, because if your probe misses and slams into Mercury, we probably don't want to infect it with Earth germs.

Now, once you solve all those problems, it would be pretty silly not to put all the instruments you can on the spacecraft. Because all of them need those functions too, and it would be a waste to do all this just to take photos.
posted by smackfu at 9:48 AM on August 18, 2007

Once we see a working Space Elevator and the cost of transporting matter into orbit decreases drastically and isn't directly linked to its mass, we'll see all sorts of things open up.

Orbital mining, waste removal, research like you're talking. In my opinion a space elevator is going to be the next internal combustion engine, an invention that drastically changes society.

But for now like everyone else has mentioned the cost for orbital transit is extremely prohibitive.
posted by spatula at 10:15 AM on August 18, 2007

Besides which, it has actually happened.
posted by Steven C. Den Beste at 10:20 AM on August 18, 2007

R&D costs a whole lot of money. At this time, we are giving NASA less money per year then they got in 1968, not counting inflation. When we need to spend $11,000,000 an hour to keep pumping missiles into a degenerating third world country, there just isn't a whole lot to spare for 'useless frivolities' like scientific research, education, and healthcare.

If you want something cheap to go into space, the government is a terrible investment. You should start asking when Virgin Galactic and all of those people are going to be able to start up their space tours, because if anything can turn a prohibitively expensive and stagnant industry into a highly profitable escape for middle-class suburbanites with a weekend to spare (and a few hundred extra pieces of scientific carry-on luggage), it's the ingenuity and desperation of a board of trustees with an unprofitable quarterly report due.
posted by sandswipe at 11:13 AM on August 18, 2007

Most other people have gotten it right. Getting something to another planet -- or even into orbit around our own -- is ludicrously expensive (mostly because you need a lot of reaction mass / fuel, which costs a lot of money to lift). So much so, that if you're going to pay the billions of dollars required to accellerate something into a transfer orbit to elsewhere in the solar system, you might as well spend the money on a spacecraft that won't die halfway and turn it into a giant waste of time.

Hence, the spacecraft are expensive and well-engineered. They ought to be: depending on the lineup of the planets, you might only get one shot at a particular exploratory path once in a lifetime (or once in many lifetimes).

There are some relatively-inexpensive earth-orbiting satellites, built with COTS gear. The Amateur Radio Satellites are good examples of this sort of economizing. But putting something into an orbit that will get it to another planet is orders of magnitude harder than putting something into LEO or geosync; rather than just strapping a box of electronics to an old ICBM and crossing your fingers, you're building an honest-to-god interplanetary spacecraft (and doing real rocket science / orbital mechanics to get it to where you want to go). That's just not cheap.

In the future, it might be possible to build cheaper interplanetary probes if rocket engines could be avoided and replaced with solar sails or ion thrusters or some other propulsion method that avoided lots of fuel. But that's a ways out from now.
posted by Kadin2048 at 12:08 PM on August 18, 2007

Now it may be my turn to be naive.

I accept that there are ridiculous costs involved in lifting things to orbit but would there be any economies of scale?

Could you, for example, raise 100 probes for the cost of raising one spaceship and then fire them all off in a flurry of exploratory excitement?
posted by Lionel d'Lion at 12:24 PM on August 18, 2007

Sure, if the 100 probes massed less than the one spaceship. Or if you sent the 100 probes on missions that required a lot less delta-v than the spaceship's mission. The 100 would probably be less efficient though, since you'd have 100 antennas instead of one, and 100 guidance computers instead of one, and 100 of lots of other things doing boring jobs instead of just one, and so a lot less room for doing science.

And sending a kilo to Jupiter is sending a kilo to Jupiter. Orbital mechanics don't care whether it's a kilo that's part of one big space probe or a cloud of little ones. AFAIK, which is not far, about the best you can do is send more or bigger probes on a roundabout route with several slingshot moves, so it takes 10 years to get there instead of 2 or 3. We do this already.
posted by ROU_Xenophobe at 1:14 PM on August 18, 2007

The problem boiled down to its most basic variables is:

To move something from the surface of the Earth to a path which would intercept another body, you need to impart a change in velocity to that object. That change in velocity is, intuitively enough, referred to as the "delta-v", and varies by orbital maneuver. An orbital maneuver is a move from one orbital "situation" to another, e.g. launching from the Earth's surface to Earth orbit, or going from Earth orbit to Mars orbit, etc.

There's a great chart here specifying the various delta-vs involved in going from the Earth to Mars and the positions of interest between here and there.

The Tsiolkovsky rocket equation is pretty straightforward and says that the delta-v is equal to the exhaust velocity of a rocket's thruster times the natural logarithm of (start mass divided by end mass). delta-v = exhaust-v * ln(start mass / end mass).

You need to change your velocity by about 10,000 m/s to get to Earth orbit. (I use 'change' here because I'm trying to be accurate, and everything is relative, after all.)

Let's pretend for the sake of argument that we have a rocket that is a single-stage-to-orbit vehicle, i.e., no boosters, etc (which doesn't exist right now, unless you count Scaled Composites' Space Ship One, which doesn't do much but go up and come down again) that has an effective exhaust velocity of 3000 m/s (this is a bit better than the Space Shuttle's main boosters) and burns fuel at 100 kg/s (this is totally made up: SS1 burns fuel at about 28 kg/s, the shuttle's main boosters burn fuel at about 4000 kg/s).

And let's say you run the engines for two minutes, 120s, so you need 12000 kg of fuel to do that.

delta-v = exhaust-v * ln(start mass / end mass)

We want to go to low earth orbit, 10000 km/s is our delta-v.
10000 m/s = 3000 m/s * ln( (structure + payload + fuel) / (structure + payload) )

which is equivalent to

(structure + payload + fuel) / (structure + payload) = e^(10000/3000)

(structure + payload + fuel) / (structure + payload) = 28.01...

Which means, even using our nice little fake engine, your fully-fueled weight must be approximately 28 times your fully-unfueled weight.

Also note that it doesn't matter what you do, it's all down to weight. That specific number, 28, will change depending on the engine you use, but generally it sucks, given current technology.

There are no economies of scale that will make pounds weigh less than a pound. It takes massive amounts of energy to move things off the planet. Gravity sucks.
posted by blacklite at 1:15 PM on August 18, 2007

I was wondering the same thing as Lionel d'Lion. All the comments here suggest that the majority of the expense is getting something up out of the gravity well, couldn't we just stick a whole bunch of probes into the payload rather than just one?
posted by quin at 1:29 PM on August 18, 2007

That, of course, is what I get for not previewing.
posted by quin at 1:32 PM on August 18, 2007

I'm going to throw out something that I'm surprised nobody has yet mentioned, which makes me think that either I'm completely wrong or everyone is concentrating on payload issues.

Specialization.

I used to be a space buff as a kid, and I noticed one odd thing ... it seemed like no two space probes were perfectly identical. Could be wrong about it. Maybe they are now, but it seems like there's nothing generic about them. Mass manufacturing of identical items is what brings down the cost (aside from the payload issue). When you spend five person-months evolving a brand new low power antenna configuration, little things like that add to the cost. It's always some new bit of technology that could be added.

And that's not even counting the software, which gets patched and repatched en route. I think the cost would come down if they settled on (ha!) a generic probe: N number of solar panels, this nuclear battery, a slew of instruments. This has its own problems - you wouldn't be able to specialize the probe to look for the "conditions of interest." If something has particularly fascinating magnetic fields, either you add that to the standard array for all probes or you go without.

The research and development costs for one probe could get spread out over one hundred probes, but you'd have to resist the urge to put on this latest solar panel with a .5% improvement in efficiency. Getting each one on the right trajectory still sucks, which is why I think we need some kind of platform located at a Lagrange point just to launch probes, but that's beyond the scope of this discussion.
posted by adipocere at 2:39 PM on August 18, 2007

Here's a great excerpt from Neil deGrase Tyson's wonderful "Death by Black Hole". This book is a great read, especially for space lay-men like myself. It really puts things in perspective:

"Beginning in 1969, space probes were designed and launched that shaped two decades of planetary reconnaissance in our solar system. The celebrated Pioneer, Voyager, and Viking missions were part of this era. So too was the Mars Observer, which was lost on arrival in the Matrian atmosphere in 1993.

Each of these spacecraft took many years to plan and build. Each mission was ambitious in the breadth and depth of its scientific objectives and typically cost taxpayers between $1 and $2 billion. During a 1990s change in administration, NASA introduced a "faster, cheaper, better" paradigm for a new class of spacecraft that cost between $100 and $200 million. Unlike previous spacecraft, these could be planned and designed swiftly, enabling missions with more sharply defined objectives. Of course that meant a mission failure would be less costly and less damaging to the overall program of exploration.

In 1999, however, two of these more economical Mars missions failed, with a total hit to taxpayers of about $250 million. Yet public reaction was just as negative as it had been to the billion-dollar Mars Observer. The new media reported the $250 million as an unthinkably huge waste of money and proclaimed that something was wrong with NASA. The result was an investigation and a congressional hearing.

Not to defend failure, but $250 million is not much more than the cost to produce Kevin Costner's film flop Waterworld. It's also the cost of about two days in orbit for the space shuttle, and it's one-fifth the cost of the previously lost Mars Observer. Without these comparisons, and without the reminder that these failures were consistent with the "faster, cheaper, better" paradigm, in which risks are spread among multiple missions, you would think that the $1 million equals $1 billion equals $1 trillion.

Nobody announced that the $250-million loss amounts to less than $1 per person in the United States. This much money, in the form of pennies, is surely just laying around in our streets, which are filled with people too busy to bend down and pick them up."
posted by afx114 at 3:25 PM on August 18, 2007 [2 favorites has favorites]

... In other words, the current probes ARE cheap. Relatively, anyway.
posted by afx114 at 3:30 PM on August 18, 2007

All the comments here suggest that the majority of the expense is getting something up out of the gravity well, couldn't we just stick a whole bunch of probes into the payload rather than just one?

Sure, we could. A Saturn V or Energia can loft ~100000 kg into low orbit. If you wanted, that 100000kg could be a bundle of probes with their own final stages to get them from LEO into interplanetary trajectories.

Whether this would make more economic sense than using a greater number of Deltas or Arianes or whatever is a boring empirical question.

A sticking point with it, though, is that you might launch 20 probes into LEO now, and send 4 on their way, but the other 16 might have to wait several months or years for their launch windows in an environment that might be less friendly than a dirtside clean room.

And of course, if the Saturn V or Energia fucks up, you've lost a whole bunch of probes.

Specialization.

Yeah, but is the cost of the probes, even including the R&D, a significant part of the cost of the mission? What fraction of Viking or Cassini is the cost of the probes themselves?
posted by ROU_Xenophobe at 3:33 PM on August 18, 2007

And of course, if the Saturn V or Energia fucks up, you've lost a whole bunch of probes.

Which has happened. Globalstar lost 8 satellites one time when the Russian booster which was carrying them blew up.

There's a reason for the term "rocket science". It ain't easy.
posted by Steven C. Den Beste at 4:51 PM on August 18, 2007

Just in passing, we no longer have the ability to build the Saturn V. Nor, in fact, would we really want to. I personally think NASA should shut down the shuttle program and invest that money in a new series of single-use heavy boosters, but it would take years.

In the mean time, we actually use a lot of medium-lift boosters: Titan series and Atlas series.
posted by Steven C. Den Beste at 4:55 PM on August 18, 2007

Just in passing, we no longer have the ability to build the Saturn V.

Of course we do. It would just be more expensive than a new design, which is a far cry from losing the ability. There's no relevant technology that we've forgotten. It would be silly to have people make 1950s and 60s era electronics, but you could do that if you really wanted to. It would be silly to spend jillions of dollars to have machine shops make bespoke washers matching 1960s ones, but you could do so if you really wanted to. We could build exact replicas of the Saturn V if we wanted to, but it would be cheaper to make a new design around modern electronics and modern part lists.

Are you thinking of the old canard that the blueprints had to be destroyed to get the shuttle approved? That's simply untrue. A moment's casual research will show that the blueprints are stored at Goddard on microfilm, and suppliers like Rocketdyne retain their own copies of their contributions.
posted by ROU_Xenophobe at 6:38 PM on August 18, 2007

Think of it like a cluster bomb:

Load a good handfull (hundreds, thousands?) of super cheap, small (softball-sized) non-mobile (think glorified wireless webcam) into the payload of your vessel and scatter into the atmosphere of target planet.

Assume a high failure rate but that's cool 'cause they're cheap.

So working ones scatter across a swath of target planet and passively take picture and otherwise collect data then transmit.

Makes more sense than putting all your funds into some super expensive, brittle, mobile and non-redundit rig like a Mars Rover.
posted by sourwookie at 10:20 PM on August 18, 2007

redundit=redundant.
posted by sourwookie at 10:21 PM on August 18, 2007

SCDB: I personally think NASA should shut down the shuttle program and invest that money in a new series of single-use heavy boosters, but it would take years.

It's precisely what NASA plan to do once the Shuttle fleet is retired in 2010 or so - the Constellation project comprising ARES-I and ARES-V rockets.

ARES-I will be a man-rated launch vehicle for the ORION crew vehicle, while ARES-V is a heavy-lift cargo vehicle developed from the Shuttle's SRBs and Main tank stack. Future missions e.g. to the moon will launch the main Lunar stack into orbit on an ARES-V, which is then joined by a manned ORION module atop an ARES-I in a later launch.

However, this doesn't stop you using ARES-V for general heavy-lift duties between lunar missions.
posted by Nice Guy Mike at 1:49 AM on August 19, 2007

sourwookie: Each softball probe would need a high gain antenna, a power source, and a positioning system to point the antenna at earth. That's a lot of glorification. Assuming one of these babies clocks out at a kilogram (a generous estimate, I think), and we declare support structure (orbital entry, dispersal, etc.) directly equivalent to the Mars Rover project, and say that all of them land safely, that only gives us about 180 photos of our target.

Alternatively, you could send along an orbiter to babysit them, but if you're doing that, why bother with the softballs?
posted by zamboni at 6:57 AM on August 19, 2007

Now this thread really has me thinking. Is there really no way to get something into orbit aside from large, heavy, expensive rockets? No possibility of a space "ultralight" that would take far less fuel to break orbit? I'm sure there's not, but it seems like the answer would be smaller, lighter to break orbit.

And for those in the thread suggesting the multiple probes per launch; yes, that's exactly what I was thinking. Something small, light, and inexpensive you could either send en mass at one planet, or all directions to various planets. You send enough of them out there to get basic data, and even with some failure, you'd come back with basic data that we only have theories about right now. And maybe stir public interest enough to get more funding.

Size and weight: cameras are smaller. I don't know about Radio Transmission, but dollars to donuts, that's smaller too. Computing is minuscule. All that means scaled down size and weight; its just the fuel that might pose a problem, right?

Which leads to, is fuel for the probes themselves such an issue? I thought that once in space, you really just needed a nudge to get going in a certain direction. So why would the fuel payload be that big of an issue? Is it just because the distance is so great? Or am I misunderstanding how objects are propelled through space?
posted by [insert clever name here] at 7:52 AM on August 19, 2007

Is there really no way to get something into orbit aside from large, heavy, expensive rockets?

Sure. If you could get the materials down, you could build a space elevator or orbital tower. In theory, you could build a Lofstrom loop. For unmanned projectiles, you could shoot them into orbit with a whacking great gun. You could use a scramjet that doesn't carry all of its own oxygen, saving some weight. You could use balloons to lift rockets above most of the atmosphere to save a bit on drag.

However you do it, though, you have to impart a whole damn lot of kinetic energy to whatever you want in orbit. To get something to stay in orbit, you've got to give it *googles* 32 MJ/kg, which is a fair amount. Geosynchronous orbit is about twice that. Rockets shove that energy in in a few minutes by shoving mass out the ass end, which means that you have to actually expend a lot more energy to lift the reaction mass, and you have to lift the reaction mass to lift the reaction mass, etc.

If you had a space elevator, you could shove that energy in over a couple of days, and you could use electricity. It would still take ~5800 MJ to lift 100 kg of person to GEO. But 5800 MJ is 1600 kilowatt-hours, which costs about $120 when it's electricity.

In general, anything that means that you're not hauling your own energy and propellant around saves a lot.

No possibility of a space "ultralight" that would take far less fuel to break orbit? I'm sure there's not, but it seems like the answer would be smaller, lighter to break orbit.

Space launchers are already ultralights, built no heavier or stronger than they have to be to survive a few minutes of thrust.
posted by ROU_Xenophobe at 8:30 AM on August 19, 2007

Which leads to, is fuel for the probes themselves such an issue?

It's not. It's the fuel in the launcher, and the launcher itself.

I don't know about Radio Transmission, but dollars to donuts, that's smaller too.

Shouldn't be. It's not like spacecraft designers include overly large antennas and powerplants for kicks. They're probably already nearly as small as they can be made to be to reliably transmit and receiver over those distances.

I thought that once in space, you really just needed a nudge to get going in a certain direction.

Depends on the orbital transfer you want. Some changes in orbit require very high delta-v, others don't.

And that first part, getting into space, is a real bitch, and where most of the expense lies.
posted by ROU_Xenophobe at 8:48 AM on August 19, 2007

Firstly, science is expensive. And the most expensive science is the stuff that's never been done before, and is operating in a non-standard environment. All of a sudden, pens don't work, because there is no gravity. Most of your off the shelf equipment will assume earthlike temperatures and pressures, a Nitrogen/Oxygen atmosphere and 9.81 m/s/s gravity, and so is unlikely to work reliably in space.

Also, each of these probs will be going to places with differing temperatures, pressures, atmospheres and distance from the sun, and will therefore have different requirements. Something that works on Mercury won't work on Jupitor, and good luck getting something to survive on Venus for more than a couple of hours.

IANArocketscientist, that's just my two cents. Would be cool though.
posted by kjs4 at 7:18 PM on August 19, 2007