How would you make "Sensors" for a starship?

[Crossroads Inc.]

It is a bog standard for every Sci-Fi show out there. Any ship out among the stars, sees something odd, and someone will say "Bring up Sensors!" But, when in outer space, what the hell do you even use as Sensors? If you are engaging in battle, you may need to locate and identify something at thousands, perhaps tens of thousands of kilometers. As far as I know we still use RADAR and various inferred sensors on ships and airplanes, and those have, what, a range of 20 or 40 miles?

So I ask, if you were going into space, what could you use for immense distances?

Moved up to the Library as an informative discussion of the optical limitations of telescopes in hard sci-fi.

[Starglider]

Nuclear-pumped pulse radar, e.g. the Excalibur from the Rama books. Eject a big fusion bomb in an appropriately designed housing and let it put out a RF pulse strong enough to image a whole solar system.

Actually passive IR should be very effective, since spacecraft (and particularly spacecraft drives) are so hot compared to virtually everything else in space. Get a redundant set of small-diameter telescopes with a computer-driven scanning and tracking system. You'll probably want some large antennaes for EM comms/radar/leakage detection/interception too; in zero gravity you can make huge flimsy ones for use when the ship is coasting and fold them away for heavy acceleration and/or combat.

[Darth Wong]

Reality check: when we see things far away in space, they're just dots. Literally, one dot. You actually need a surprisingly huge telescope in order to get a clear visual of anything at what we would consider an astronomical distance.

http://www.stardestroyer.net/Resources/ ... lLens.html

Beyond a certain range for a given size of object, all you get is one dot. You can't tell how big it is, whether it's one object or fifty, (as long as they're all within the limit radius), or whether it's nothing more than an infrared emitter drone.

A lot of our ideas about what we can detect (and how easily we can render stealth tactics useless) are drawn from terrestrial military research where you're looking at things which are a few km away, or maybe even a few hundred km away. Things change when you're looking at something which is millions of km away. Astronomers develop all kinds of fascinating methods for extracting as much information as they can from a dot: the brightness of it, the distance to it, the cyclical changes in its brightness, its frequency, etc. But it's still a lot less information than we typically expect to get from things at the ranges we're accustomed to on Earth.

[Darth Wong]

It is a bog standard for every Sci-Fi show out there. Any ship out among the stars, sees something odd, and someone will say "Bring up Sensors!" But, when in outer space, what the hell do you even use as Sensors? If you are engaging in battle, you may need to locate and identify something at thousands, perhaps tens of thousands of kilometers. As far as I know we still use RADAR and various inferred sensors on ships and airplanes, and those have, what, a range of 20 or 40 miles?

So I ask, if you were going into space, what could you use for immense distances?

There are basically two ways to detect anything at range:

1. Passively wait to detect emissions from the object.
2. Bounce something off the object and wait for a detectable interaction.

At much closer ranges, you can do a lot more tricks, like creating a forcefield which interferes with the object and then detecting changes in the forcefield. But that's irrelevant here, where you're stuck with the two basic principles.

Electromagnetically, the passive sensor is an optical scope or telescope, which every starship should actually have, but which you never see anyone mention in a sci-fi movie because it doesn't sound futuristic enough. The receiver in the telescope would determine which kinds of frequencies you can detect. The active sensor is radar: bounce an EM pulse off a target and wait for the reflection.

Telescopes obviously work over long ranges, but most sci-fi fans ignore their limitations and think they can detect objects at arbitrary distance, with arbitrary resolution. That's utter bullshit. Radar can also work over long ranges, but the problem becomes sensitivity: as the distance increases, you need to increase your power or decrease the width of your scanning beam in order to get useful reflections. If you use a very focused beam, you can get longer range but it also takes much longer to sweep across a useful section of sky.

[Starglider]

Reality check: when we see things far away in space, they're just dots. Literally, one dot. You actually need a surprisingly huge telescope in order to get a clear visual of anything at what we would consider an astronomical distance.

Interferometry could help with that. I imagine you'd have a few small telescopes scattered around the hull of your ship operating in survey mode, then when they find an interesting target they'd switch to interferometric mode to image it. That would give you an effective mirror size (for resolution purposes) equal to the size of your ship, though at some complexity cost, and brightness might still be an issue.

[Darth Wong]

Interferometry is useful if you're trying to simulate a larger lens with a bunch of smaller ones, but even if we assume a huge lens, the limitations are much greater than sci-fi fans realize. I've seen some pretty stupid shit posted by people who think that the whole "infrared makes you stand out against space" line makes pretty much anything impossible to hide.

Concrete example: using 750nm infrared light and a gigantic 500 metre wide telescope (let's be realistic; how big is this ship?), an approaching 100m wide starship at a range of 50 million km produces a picture with a resolution of ... 1 pixel. That's just not good enough for detecting much useful information. You wouldn't be able to tell if it's a 100m wide starship or a 1 metre wide emitter drone. At 1 AU, you wouldn't be able to tell if it's a single starship or a whole flotilla.

[Starglider]

I've seen some pretty stupid shit posted by people who think that the whole "infrared makes you stand out against space" line makes pretty much anything impossible to hide.

People do take it to excess, but I believe that line usually refers to drive exhaust, and possibly cooling. The exhaust plume from a fusion or antimatter rocket is going to be radiating a huge amount of EM and the radiators cooling the drive will probably be glowing in the high IR as well. Chilling one side of the ship and using decoy drones and cold thrusters (e.g. mass drivers) should work fine as long as you know the enemy is looking for you and you aren't doing serious maneuvering, but that severely limits your tactical options.

With regard to interferometer size, it is possible to use multiple ships or drones in formation to perform very long baseline interferometry; we can already do this for astronomical purposes with satellite constellations. Obviously this would come with its own set of difficulties and limitations, but it may still be useful, particularly for static early warning arrays.

[Darth Wong]

Who says you even need to chill one side of the ship? This is space; there is no atmospheric drag. You could have an "umbrella" suspended away from your ship in order to make it less visible. Unless you do something really stupid like turning sideways relative to your target and doing a full burn, that will make you really hard to detect. Add the fact that he's severely range-limited anyway, and it's not as hard as people think to either sneak up on someone or confuse the hell out of him with decoys. After all, even if he picks you up, all he sees is a dot.

[Lord of the Abyss]

Concrete example: using 750nm infrared light and a gigantic 500 metre wide telescope (let's be realistic; how big is this ship?), an approaching 100m wide starship at a range of 50 million km produces a picture with a resolution of ... 1 pixel.

Would it be possible to deploy something like a solar sail, except shaped to work as a telescope mirror instead ? Or perhaps something like Robert L. Forward's microwave sail 'Starwisp' but as a receiver and not as a sail. Would that work to get such a large size telescope/radio telescope from a reasonably sized ship ?

[Starglider]

Would it be possible to deploy something like a solar sail, except shaped to work as a telescope mirror instead?

How are you going to make that conform to the specific micrometre-accuracy perfectly smooth curved surface required for telescope applications? The whole thing would have to be some sort of super-exotic active material that could measure and adjust its own precise shape. You could use a huge number of rigid segments, but that would be much heavier, harder to fold/unfold and is still a massive actuator and control challenge. Then you have the problem of pointing it accurately...

[aimless]

Interferometry is useful if you're trying to simulate a larger lens with a bunch of smaller ones, but even if we assume a huge lens, the limitations are much greater than sci-fi fans realize. I've seen some pretty stupid shit posted by people who think that the whole "infrared makes you stand out against space" line makes pretty much anything impossible to hide.

Concrete example: using 750nm infrared light and a gigantic 500 metre wide telescope (let's be realistic; how big is this ship?), an approaching 100m wide starship at a range of 50 million km produces a picture with a resolution of ... 1 pixel. That's just not good enough for detecting much useful information. You wouldn't be able to tell if it's a 100m wide starship or a 1 metre wide emitter drone. At 1 AU, you wouldn't be able to tell if it's a single starship or a whole flotilla.

I assume you can use the sensors from an entire fleet to create a very large interferometer telescope though? I believe there's on operating right now that combines ground telescopes with an orbiting one for a ludicrously long baseline, though I think all the very long interferometer telescopes are radio scopes.

[Darth Wong]

I think it needs to be clarified that an interferometry telescope is not actually completely identical to a physical telescope of the same size, particularly if the ratio of the small scopes is very small compared to the virtual large scope. It's not as if you can take a pair of 4 inch lenses, put them 10 km apart, and get the same pictures you would get from a 10km wide lens.

[aimless]

I think it needs to be clarified that an interferometry telescope is not actually completely identical to a physical telescope of the same size, particularly if the ratio of the small scopes is very small compared to the virtual large scope. It's not as if you can take a pair of 4 inch lenses, put them 10 km apart, and get the same pictures you would get from a 10km wide lens.

I'm having trouble finding that difference quantified anywhere on the web, I guess because interferometer telescopes are so damn complex and it varies so much, but it would be nice to get some really basic rules of thumb for the comparative effectiveness. Don't suppose you know where to find it? :p

[Morilore]

Would it be worthwhile to fire small sensor probes like bullets at enemy/unknown "dots" to radio back more detailed sensor information? How accurately could you fire a bullet or a missile at a "dot" 50 million km away, and how small could you make sensors, passive or active, powerful enough to gather information such as the rough size of the ship, the number of ships, etc. at that range? Is it realistic for such things to be able to move fast enough to get information in a useful time scale (whatever that happens to be)? If they had their own engines, would their exhaust necessarily mark them as targets and get them shot down before they get close enough to gather any kind of information? If they did not have their own engines and had to be accelerated entirely at the launch point, how fast could you realistically get something that size?

[PeZook]

Would it be worthwhile to fire small sensor probes like bullets at enemy/unknown "dots" to radio back more detailed sensor information? How accurately could you fire a bullet or a missile at a "dot" 50 million km away, and how small could you make sensors, passive or active, powerful enough to gather information such as the rough size of the ship, the number of ships, etc. at that range? Is it realistic for such things to be able to move fast enough to get information in a useful time scale (whatever that happens to be)? If they had their own engines, would their exhaust necessarily mark them as targets and get them shot down before they get close enough to gather any kind of information? If they did not have their own engines and had to be accelerated entirely at the launch point, how fast could you realistically get something that size?

The drones could simply punch themselves to a trajectory which would take them roughly into the area where you expect the enemy to be, and then coast the rest of the way on a timer which would activate them at the end of the run so that they could pepper the ship with active emissions and radio back the results before being blown up.

Actually, these detection problems give us an interesting reason for planets and moons to be important strategic locations: even a small airless moon could house gigantic sensor arrays that could provide significant advantages to their owners when it comes to control of space. You could harden them, too, far better than any starship - an attacking fleet would be pretty much screwed if the defenders calculate their trajectory early - that means they'd come under fire before being able to employ their own weapons against the defenders' ships.

[aimless]

Would it be worthwhile to fire small sensor probes like bullets at enemy/unknown "dots" to radio back more detailed sensor information? How accurately could you fire a bullet or a missile at a "dot" 50 million km away, and how small could you make sensors, passive or active, powerful enough to gather information such as the rough size of the ship, the number of ships, etc. at that range? Is it realistic for such things to be able to move fast enough to get information in a useful time scale (whatever that happens to be)? If they had their own engines, would their exhaust necessarily mark them as targets and get them shot down before they get close enough to gather any kind of information? If they did not have their own engines and had to be accelerated entirely at the launch point, how fast could you realistically get something that size?

What's a useful timescale? The fastest you're going to get a probe nearby to something 50 million km away is several hours, and that's a superduper torchship probe accelerating at something absurd like 10 g's. If you just railgunned a probe with a very optimistic projectile velocity of 20km/s and your ship wasn't moving towards theirs...well that would take a month.

Some small sensor ship between those extremes could easily be useful and under most realistic conditions would be able to gather data outside of effective weapons range. This is all so fuzzy though without defining a lot of stuff.

Actually, these detection problems give us an interesting reason for planets and moons to be important strategic locations: even a small airless moon could house gigantic sensor arrays that could provide significant advantages to their owners when it comes to control of space.

Until someone definitely proves to me that fleets can't arrange themselves into massive interferometer telescopes I'm declaring the advantage nullified

[PeZook]

Until someone definitely proves to me that fleets can't arrange themselves into massive interferometer telescopes I'm declaring the advantage nullified

A moonside array could get superior results by virtue of having a much better ratio of actual emitters to empty space for a given cost. And, of course, you can harden everything from emitters themselves through command centers to data cables.

[Morilore]

What's a useful timescale? The fastest you're going to get a probe nearby to something 50 million km away is several hours, and that's a superduper torchship probe accelerating at something absurd like 10 g's. If you just railgunned a probe with a very optimistic projectile velocity of 20km/s and your ship wasn't moving towards theirs...well that would take a month.

OK, so a 50 million km range is definitely absurd for such a thing. Math-fu tells me that to accelerate something the size of Voyager 2 through a 10 km-long mass driver to 20 km/s would require 140 GW and would take more than five hours to get from Earth to the Moon, so this whole idea, at least the way I wanted to envision it, is already looking silly.

[PeZook]

OK, so a 50 million km range is definitely absurd for such a thing. Math-fu tells me that to accelerate something the size of Voyager 2 through a 10 km-long mass driver to 20 km/s would require 140 GW and would take more than five hours to get from Earth to the Moon, so this whole idea, at least the way I wanted to envision it, is already looking silly.

It's a decent idea, just not at this kind of range. Sensor pods/probes could also be a way to build huge mobile VLAs in an attempt to gain the upper hand on the enemy. Or, perhaps, one could scatter such devices around approach orbits (there's fewer of those than people usually think, unless we assume ridiculous amounts of delta-v for the attackers) so that you know early that something is coming, and can point your absurdly huge arrays in the correct direction for a more exact fix.

[GrandMasterTerwynn]

I think it needs to be clarified that an interferometry telescope is not actually completely identical to a physical telescope of the same size, particularly if the ratio of the small scopes is very small compared to the virtual large scope. It's not as if you can take a pair of 4 inch lenses, put them 10 km apart, and get the same pictures you would get from a 10km wide lens.

I'm having trouble finding that difference quantified anywhere on the web, I guess because interferometer telescopes are so damn complex and it varies so much, but it would be nice to get some really basic rules of thumb for the comparative effectiveness. Don't suppose you know where to find it? :p

The difference is quite simple. A ten kilometer wide interferometer comprised of a pair of one meter telescopes will have the resolution of a 10 km telescope, but the light gathering capability of a pair of one meter telescopes. In short, the angular resolution of the interferometer is dictated by the diameter of the array, but the amount of light it gathers is constrained by the size of the individual scopes making up the array.

[aimless]

I think it needs to be clarified that an interferometry telescope is not actually completely identical to a physical telescope of the same size, particularly if the ratio of the small scopes is very small compared to the virtual large scope. It's not as if you can take a pair of 4 inch lenses, put them 10 km apart, and get the same pictures you would get from a 10km wide lens.

I'm having trouble finding that difference quantified anywhere on the web, I guess because interferometer telescopes are so damn complex and it varies so much, but it would be nice to get some really basic rules of thumb for the comparative effectiveness. Don't suppose you know where to find it? :p

The difference is quite simple. A ten kilometer wide interferometer comprised of a pair of one meter telescopes will have the resolution of a 10 km telescope, but the light gathering capability of a pair of one meter telescopes. In short, the angular resolution of the interferometer is dictated by the diameter of the array, but the amount of light it gathers is constrained by the size of the individual scopes making up the array.

I guess the quantification I'm looking for would be in the difference in effective range. So let's say with a Really Big Scope I can barely see a target at 1 pixel resolution. I break up that area amongst 10 ships and spread them out to boost my resolution. Now with this diameter I can theoretically get a 1 pixel resolution much farther out, but are you saying that I won't be able to with this certain amount of light collecting hardware? Or are there diminishing returns? Is there a formula for effective range based on a combination of light collecting area and diameter?

[aimless]

I guess the quantification I'm looking for would be in the difference in effective range. So let's say with a Really Big Scope I can barely see a target at 1 pixel resolution. I break up that area amongst 10 ships and spread them out to boost my resolution. Now with this diameter I can theoretically get a 1 pixel resolution much farther out, but are you saying that I won't be able to with this certain amount of light collecting hardware? Or are there diminishing returns? Is there a formula for effective range based on a combination of light collecting area and diameter?

Thinking about this my question is a bit silly since effective range for space military application is a such a nebulous concept and I'm realizing that the issues with effective range and collecting area are going to come from how long you have to look at something to get a meaningful exposure. If you magically know where the enemy ship is and can point your scope at it for a long time you'll be able to get an ok image after a long time even if you couldn't get one after a few minutes. So I guess the question of effective range is when performing a scan of an area how long do you have to linger on every section to make absolutely sure you don't miss an enemy ship, given the desired max range you want to detect at and the total collection area.

[Darth Wong]

Reality check: optical interferometry depends on optical wave interference (obviously). To make it work, the elements of an interferometer must be aligned so precisely that the two images will actually be in phase, so we're talking about lining it up (or adding a delay line) to within an accuracy of small fractions of a wavelength. Given that visible light wavelengths are in the sub-micron range, these hypothetical starships must be able to do station-keeping with nanometre precision.

That's not a small difficulty.

In addition, you don't get interference without combining the two light streams, so you need to bounce the light beams from the telescope array off mirrors or something so that they will combine somewhere. Moreover, with two telescopes you can measure the diameter of a distant star, but you don't get enough information for an actual image (astronomers are really helped out by knowing in advance that the target object is round). You need more telescopes to get an actual image.

Just don't get the idea that interferometry allows you to make arbitrarily large virtual telescopes by just telling starship A to go over here and starship B to go over there, even before we look at the light intensity problem (which obviously reduces sensitivity to orders of magnitude below a conventional telescope).

PS. Plain-English description of optical interferometry: we know that light diffracts when it passes through an aperture, eg- a telescope lens. If we pass light through several apertures and then allow the resulting "fringe patterns" to interfere with each other, we can use computer software to reconstruct a virtual image from the interference pattern. The problem is that optical interference requires phase alignment: a real problem when you're talking about wavelengths measured in nanometres. It also requires some kind of "combining" facility, so a three-telescope interferometer actually requires four elements: three telescopes on spokes around a central combiner facility.

[Starglider]

Reality check: optical interferometry depends on optical wave interference (obviously). To make it work, the elements of an interferometer must be aligned so precisely that the two images will actually be in phase, so we're talking about lining it up (or adding a delay line) to within an accuracy of small fractions of a wavelength. Given that visible light wavelengths are in the sub-micron range, these hypothetical starships must be able to do station-keeping with nanometre precision.

This isn't a physical limitation, it's a technological one. We can do aperture synthesis at radio wavelengths without physically combining the signals, simply with digital processing and very accurate clocks. We can't (AFAIK) currently do this at optical wavelengths, due to the lack of an optical sensor that can capture the required wavefront data (and possibly due to the extreme timing accuracy requirements). However progress in solid state physics and computing is constantly raising the frequency at which virtual interferometry is possible. Combined with high-precision measurement of relative position, this potentially allows on-the-fly combination of a fleet of ship's sensor data for resolution enhancement without actual beam combining. Luminosity will likely still be a problem for enemies making a 'cold approach'.

In addition, you don't get interference without combining the two light streams, so you need to bounce the light beams from the telescope array off mirrors or something so that they will combine somewhere. Moreover, with two telescopes you can measure the diameter of a distant star, but you don't get enough information for an actual image (astronomers are really helped out by knowing in advance that the target object is round). You need more telescopes to get an actual image.

The more baselines you have, the less blurry the image gets. Again, we've been making rapid progress in processing algorithms recently, so you can now plug in pretty much any (non-parallel) assortment of baselines, which progressively sharpens the image.

Just don't get the idea that interferometry allows you to make arbitrarily large virtual telescopes by just telling starship A to go over here and starship B to go over there, even before we look at the light intensity problem (which obviously reduces sensitivity to orders of magnitude below a conventional telescope).

[Darth Wong]

Your quoting is fucked up. Anyway, even if you can do away with the central combiner facility, you still need the ships to have nanometre-scale location accuracy. Even if you can simulate all of this with processing, you still need to know the locations of the ships to that same accuracy in order to correct for the phase angles. One way or another, you need a map of the starships' locations and orientations which is absurdly accurate.