Books I Sell

2023/04/15

Sci-Fi: Sensors and Navigation

Sensors are often overlooked in science fiction. They're a tool that allows the plot to happen, and have whatever capabilities and author requires at the moment.

 In fiction:

"Captain, we've detected a ship at 100,000km."
"Onscreen."
[Image appears, crisp and detailed]
"Any life signs?"
"Negative."
"We're detecting high levels of applied phlebotinum in their hull."

In reality:

In the Bright Conference, sensors aren't magic. Let's see what is actually possible.

The Panopticon

Most civilized systems have a cloud of small telescopes and/or few very large telescopes to track asteroids, debris, in-system ships, solar flares, and traffic violations, linked to pattern-matching computers and live decision-makers. There's no stealth in space... if you've got a moderate budget.

But starships have mass restrictions. You can’t pack all the toys.

Large ships can, of course, pack quite a few toys. Distributed array telescopes are a plausible reason to carry a drone or two. But small ships, like the ~20 ton single-astronaut cans of the Bright Conference, can’t do everything. Their sensors need to be:

  • Plausible using current technology. Off the shelf components are ideal.
  • Low maintenance. Minimal cooling tech, and robust enough to survive thumps, bumps, and lurching.
  • Low mass.
  • Small enough to launch inside a faring. No boom arms or long aerials; a Bright Conference pod has to maneuver and dock, not just cruise.

The intention is not to collect space science. It’s easier just to ask the locals for their data. The intention is to navigate, take photos, and locate objects in a system.


Sensors in the Bright Conference

The Mk. 1.0 Human Eyeball

Also known as looking out a window. Pack a camera with a selection of expensive lenses and a pair of binoculars. This sensor is in your living area, which means you can use it as long as you’re alive, and it’s cooled and maintained by the most important systems in the ship at no additional cost. In an emergency you can navigate (badly) with nothing but a grease pencil, a window, a few bright objects, and a notepad.

The Mk. 1.0 eyeball is adaptive, which is annoying in some ways. It can't tell the difference between different stellar classes visually up close. If you're around a yellow star or a blue star, all light "seems" white.

 

Basic External Cameras

In an ideal world, a pod would have full coverage fixed cameras (at least one in every direction).  Cooling and power requirements mean that might not be viable. Cameras need to be toggled on. Each camera's shutter is connected to a simple lux detector so you don’t Bean your camera by pointing it at the sun. 

Instead of, or in addition to, fixed hull cameras, you could mount something like the ISS EHDCA (PDF link). This is not a highly precise automatically tracking camera. This is a basic look-around-the-hull camera. A periscope without a direct optical link. The more I look at the EHDCA, the more delighted I am by its design. 

You'll probably want at least one camera on the end of your robotic arm, for inspection and tool manipulation, and possibly one low-resolution fixed forward-facing camera for docking or maneuvering.

Alex Ries

Main Telescope

You want the widest and longest telescope possible. You can use tricks to fold the telescope’s length into itself, but you can't escape the laws of optics. 

Ideally, you want a rotating mount with independent stabilization. You can set it to track an object and it will, provided your spacecraft isn’t manuvering or you aren't throwing your weight around. Most space telescopes point the entire craft at their target; this is not feasible with a large heavy craft full of air, unpressurized liquid, cargo, and a wiggly human being. Independent stabilization is needed.

You can divert the telescope's image to a diffraction grating and spectrometer or a number of specialized CCDs (IR, visual, UV). You'll probably have a lower resolution conventional off-the-shelf full-colour option and a higher resolution black-and-white option for specific wavelengths that your computer can use to create a false-colour composite.

Ideally, you'll want to point the telescope 90 degrees from your direction of travel so you don’t chip your mirrors.

How big?
Given the mass, size, and complexity restrictions, let's say a Bright Conference pod defaults to an 8” reflector telescope. You could easily go with a 12”, but anything larger and I’d start to worry about stabilization and complexity. 

What can you see?
Conveniently, lots of people on earth have 8” reflector telescopes. We can use their images to calibrate our expectations. Yes, focal length and eyepieces and digital layering and all that will change the results, but we're eyeballing output, not calibrating an actual telescope.

Reddit user McTaSs

You can spot the shape of the ISS, but you can’t tell one satellite from another. A fleet of Star Destroyers could park next to Jupiter and you’d have no idea. Still, an 8" telescope is a lot better than the Mk. 1 Human Eyeball, especially when it's connected to an image-averaging computer. With enough time, you can smooth out the fuzziness of an image, or spot small changes and highlight them. 

For the purposes of Fermi estimates, ISS is 400km up and is 100m wide. So if you want to read a 1m license plate with the same blurry resolution as that ISS pic, you need to be within 4km. 

Spectral Analysis

Bounce incoming light off a diffraction grating, so that only one wavelength reaches the detector, then record the intensity. Slowly scan across all available wavelengths, then repeat several times, then do some math on the output. From this, you can learn:

  • The main elements in the atmosphere of a planet... if you can point your telescope at it for more than an hour and if your software can peel signal from noise.
  • What fuel a ship is burning (H2/O2, CH4/O2, etc.) and if the ship is using a nuclear heat source to boost fuel temperature.
  • The spectral type of the nearest star (if your telescope is set up as a solar telescope; otherwise, do not point it at the sun.)

You cannot detect:

  • Life signs inside a ship.
  • The exact elemental composition of any given distant object.

If you can estimate the mass of a ship, you can work out all possible trajectories it can take. You might not know how full its fuel tanks are, or the type of engine, but the rocket is a tyrant.

The Magellan probe. The dual dish (top) and altimeter (cone-shaped thing, left) are probably worth including on a Bright Conference pod, in an updated form.

Radar

Tuneable, so you can listen to ambient radio waves bouncing off of planets and objects. Shielded from your own ship as much as possible. A dish works. The bigger the better, provided it fits in your faring and can be cooled. A forward-facing dish could also act as a debris shield, as a dish antenna is more resistant to damage than a telescope mirror. The same laws of optics apply, so a sensibly-sized dish won't give you magically more detail than your optical telescope.

Active radar (such as SAR) can provide fairly detailed results without implausibly complicated equipment. Resolution is unlikely to beat your optical telescope's resolution.

X-Ray Telescope

Useful for pulsars, useful for navigation. X-ray optics are getting smaller, better, and sturdier these days, so a dedicated telescope seems viable. You could possibly mount it on the same stabilization platform as your main optical telescope. Long focal lengths probably aren't viable.

A scintillation detector consists of a shielded tube, a crystal, and a photodetector. Gamma rays and/or X-rays hit the crystal and make it sparkle, and the detector converts those flashes of light into a signal. No moving parts. Aside from spotting stars, a scintillation detector is useful for spotting unshielded reactors or other human-killing hazards... hopefully before they become hazardous.

In theory, you could pack a gamma ray or X-ray spectroscope, which could provide another method of elemental analysis. Watch cosmic rays hit an object, then analyze what bounces back. In practice, I'm not sure they're worth it. They need a lot of shielding compared to a simple rate-based "where are we and are we in trouble" scintillation detector and a lot of time to generate useful results. Still, if you can make a small one, you might as well bring it along.

Laser Rangefinder

Bright Conference pods have tiny lasers for docking. The larger the laser, the more complex the cooling system. Calibrated for distances below 500m. Not useful for long-distance signalling.

Gravity Detector

I've included this because gravity detectors are cool, but they're probably not useful. They're extremely fiddly to use on the ground. In theory, they can detect a spoon at 5m. In practice, you don't want them to detect a spoon at 5m, you want them to detect large and distant objects, which means you need a very stable platform, long observation periods, and a lot of noise correction. 

Neutrino Detector

Again, not useful. High mass, very little immediately useful information.


The Scale of the Galaxy

The Galaxy Song is still, 40 years later, accurate enough for RPG purposes. It's worth storing in the back of your mind for quick reference.

Our galaxy itself contains a hundred billion stars.It's a hundred thousand light years side to side.It bulges in the middle, sixteen thousand light years thick,But out by us, it's just three thousand light years wide.
We're thirty thousand light years from galactic central point.We go 'round every two hundred million years.And our galaxy is only one of millions of billions,In this amazing and expanding universe.

Thought experiment: You are Star Tyrant Ludicrous the 2nd. Your vast fleet of space warships can conquer an astonishing 100 star systems per second. Tick. 100 flags over 100 suns. Tick. Another 100 flags over another 100 suns. 

How long does it take your fleet to conquer the Milky Way?

100 billion stars / 100 stars per second = 1 billion seconds (1x10^9). We know that there are 3.2x10^7 seconds in a year (it's a handy number to memorize). So that's 32 years. Tick. Tick. Tick.

Similarly, if you imagine a line sweeping across the galaxy at 100 stars per frame and 60 frames per second, it'd still take 192 days for the line to reach the other side. 100 billion is a ludicrously large number. It boggles the mind. And that's a low estimate; some papers suggest 400 billion stars is more accurate.

The point is, if you're telling a human-scale story and you feel the need to include other galaxies, consider just how much sand is currently in your sandbox.

Lost In Space: Galactic Orienteering

Thought experiment: You are teleported somewhere in the galaxy. How can you determine your position?

In a Bright Conference scenario, you can ask the gate that you exited through where you are, and compare it to any number of moderately accurate maps of the network. The gate's automated system will give you all the information you need about your location in the galaxy and in the local star system. It's always acceptable to ask for directions.

But let's imagine that you can't ask. This is an interesting experiment, and one that doesn't seem to have any well-documented solutions. If you have a cunning answer, post it in the comments.

1. Broad Position
All the constellation are different. You can't expect a computer to accurately store the relative position of 100 billion stars and then, just by looking at a portion of the starfield, calculate your location. It's possible, but you'd need a very, very accurate galactic map and a very fast computer.

Instead of looking at all the stars, why not use a special type of star? Pulars seem like they're designed for celestial navigation. They're unique lighthouses. If you have a database of pulsars, you can slowly scan the sky with your X-ray telescope or radar dish, locate a few pulsars, and triangulate your position. The more pulsars you identify, the more accurate your position. Wikipedia claims +/- 5km but I'm skeptical.

2. Local Position
But before you determine your broad position, it's best to determine your local position.

First, turn on your navigation computer. Tell it to lock your pod's position relative to the starfield ahead of it. Basically, your computer can take a picture of the star and use your pod's reaction wheels to keep the stored image in line with what the camera sees. It's tricky to take celestial measurements if your ship is rotating or tumbling.

Once your INS is set, you can adjust your pod's attitude without fear of losing track of external objects.

Second, check your habitat dosimeter. If you're in a high radiation area you might not live long enough to do anything about it, but it's nice to know. Radiation can give your next steps a sense of urgency.

Third, try to spot the nearest star. The easiest method is to unlock the roll axis and gently spin your pod while looking through the windows. Spinning along a pod's long axis requires less energy than tumbling end over end. Within a system's heliosphere, the local sun is still probably bright enough to identify with the Mk. 1 eyeball.

If there's no obvious candidate for a nearest star, you might be in interstellar space. Proceed to the next steps, but if you don't spot any planets or bright objects, then your broad pulsar-determined position from step 1 is probably acceptable. Space is big and mostly empty. 

If you do spot a star, mark its approximate position in the computer. During the next phase, your computer will try to avoid pointing your ship's delicate instruments anywhere near the star.

Activate survey mode. Your navigation computer will use your ship's cameras to build a complete map of the starfield around your ship. This process takes approximately one hour. Your pod will alter its orientation using reaction wheels. You should avoid moving around the pod during this process. Bunk down or strap into your seat.

The survey may identify bright spots. These could be planets, ships, or stations. You can export the coordinates to your telescope system and take a closer look.

3. Velocity
Calculating your velocity is difficult. Your spaceship doesn't come with a magic speedometer.

The good news is that you can use your radar system to determine your velocity relative to any dangerously close planets or vessels. Eyeballing some existing projects, I'd say a Bright Conference pod could determine the relative velocity of any planet, moon, or asteroid within 500,000 km via active Doppler radar. Ships and stations might need to be within 5,000km.

But you can't calculate the orbits of distant planets if you don't know your relative velocity. You could be stationary, relative to the star, and falling like a stone, or you could be speeding through the system like a bullet. You can't tell if the apparent shifts in a planet's position are due to its velocity or your velocity.

You could potentially use the red/blueshifts of various pulsars to calculate your velocity relative to Sol, but that's not useful. You might assume that, if you can triangulate your position within the galaxy within +/-5km, you could just take multiple pulsar fixes and calculate your velocity that way. This is true, but it'd be your velocity relative to various pulsars, not relative to other objects in the system.

You could use red/blueshifts of local spectra to determine your velocity, but only if your relative velocity is alarmingly large. You only have an 8" telescope.

Also, you can't assume you're near the invariable plane of a system, so you'll need to take many, many measurements. Since a planet's moons typically fall along the same plane as the other planets, you can use them to quickly determine the approximate location of the invariable plane. Moons might also be more useful for the estimates above. I'm sure astronomers can do fiendishly clever things with transits and shadows. I've tried to work out a few basic calculations with the moons of Jupiter. The results suggest it's possible, but very difficult, and you'd need to know or estimate the size or mass of the moon to get any useful information.

A gravity gradiometer might be useful here, but, as stated above, they seem very fiddly and slow. A gradiometer can, as the name implies, only measure a gradient, so it's a bit like trying to navigate a city by only looking at the pavement under your feet.


Best Guess

Also known as doing a lot of estimates and averaging the results.

  • You can determine the type of the star by its spectrum, and can estimate its mass.
  • You can determine temperature of gas giants or atmospheric planets by their spectra. You can use that, plus the stellar type, to estimate their distance from the star.
  • You can estimate the mass of a planet by its type. That, plus its distance from the star, can be used to calculate its orbital period via Kepler.
  • You can then compare the estimated velocity to the observed velocity to get your own orbital velocity.

If your trigonometry is rusty, have an analogy. Imagine you are standing and watching a street through a camera. A car drives in front of you, perpendicular to your camera. You take two photographs of the car. 

By comparing the two photographs and the time between them, you can calculate the car's velocity. If you know the length of the car you don't even need to know how far away the road is. With some slightly trickier math, you can even tell if the car is moving away from you ("changing lanes") as well as moving past you.

In a second experiment, you walk by a stationary car and, while in motion, take two photographs. If you know the time between the photographs, you can calculate your velocity.

In a third experiment, you walk past a moving car. If your velocity is known or the car's velocity is known, you can calculate any missing information. But if you and the car both have unknown velocities, calculations are no longer possible. You can't tell if the apparent shift in the car's position between the photographs is because it's moving, you're moving, or both.

Nevertheless, you know the the bounds of the velocities, and can use them to estimate possible unknown values. You know the speed limit on the street. You know the car isn't stationary. You know that you probably can't run faster than 12 km/h. You know that cars come in a relatively narrow range of sizes.

Why Is Velocity Important?

You want to avoid crashing into things. Space isn't about up. It's about sideways. You can get near space with a hot air balloon or a cannon. You can't stay in space without a whole lot of sideways.

The further you are from a gravity well, the cheaper it is to adjust your path relative to that gravity well. This is why probes that want to look at the sun's poles travel deep into the solar system before swinging back towards the sun. The smaller your relative velocity, the easier it is to adjust your vector. 

So figuring out where you are, how fast you are going, and if you need to adjust your velocity is a very important part of stellar navigation. And since Bright Conference pods have very limited delta-v, it's important to know when to burn.

In the Bright Conference, gates are usually parked in stable orbits. Earth's gate is parked near Earth-Moon L1. You have time to figure out where you are.

In the real world, space navigation relies heavily on dead reckoning. You know where you were, you know what changes you made to your velocity, and you can check your current position against a number of known observations. Apollo 13 wasn't able to take see the stars through their debris cloud, but they were able to use the position of the sun, the earth, and the moon (and some frantic ground-based calculations) to check their position and velocity. 

In the Bright Conference, you need to reset your dead reckoning system every time you travel through a gate. Most systems provide standardized (or at least comprehensible) information. 

But the map is not the territory. Knowing where you are and how fast you're going won't help if you're six months from any interesting destination. Sir Isaac Newton is the deadliest son of a bitch in space, and he's a persistence predator.

4 comments:

  1. Now, you might think it's a long way down to the chemist but that's peanuts compared to space!

    ReplyDelete
  2. Another option for broad position would be a triple sighting, using the galactic core as one point of your triangle and then finding two galaxies that are not in the galactic plane (purely for easier observation since the Milky Way itself is more likely to obstruct observations within the galactic plane). The relative locations of those galaxies to each other and to the core should provide a unique location. As with the pulsar method, more observations will be better since resolution limits on the observation will introduce some uncertainty, but it's another way to quickly narrow down the possibilities with a relatively limited data set.

    ReplyDelete
    Replies
    1. I thought of that, but there are a few problems. What if you're somewhere near the core? Can't sight the forest for the trees. With an 8' optical telescope, getting an accurate galaxy ID from a single snapshot could be a problem, while with a pulsar, the pulse contains several levels of identifying information and can be picked up by a fairly compact dish. It's a good backup method though.

      Delete
  3. GURPS Spaceships 5 has an entire section about scanning new planets and stuff. It really allows you to go in-depth on the mechanics of exploring space.

    ReplyDelete