Inspired by Mapping the far side of the galaxy, I wondered how it would be possible to create a real-time map of the entire universe. As that question dealt with the issue of a one-time travel to the other side of the galaxy and figuring out where you are based on existing mapping data, this question focuses more on how to make a total map of the universe with all the resources you could ask for.

Let's make some assumptions:

  • A civilization has the means to travel to any point in the universe in an instant
  • Said civilization has enormous resources (due to aforementioned zero-time travel) to place a tracking beacon by every star, planet, moon, and asteroid in the universe.

So here comes the issue: how can the universe be accurately mapped when the stars in the night sky look different from every solar system?

Perhaps more importantly, how could any given section of space-time be considered "correct" if a planet looks like a primordial puddle from one section of space-time and a thriving trade hub from another?

  • $\begingroup$ Couldn't you simply use the same FTL method for communication between beacons as you do for travel? That would solve the problems you identified. $\endgroup$
    – Frostfyre
    Aug 17, 2016 at 19:34
  • $\begingroup$ Oddly, this question might work as an answer to this question: worldbuilding.stackexchange.com/questions/4918/… $\endgroup$
    – Twelfth
    Aug 17, 2016 at 19:43
  • $\begingroup$ According to our understanding of physics, there cannot be any such thing as a "real time" map of the galaxy, much less the universe (what ever that means) There is no such thing as absolute time. Time is relative. $\endgroup$
    – Seeds
    Aug 17, 2016 at 20:14
  • $\begingroup$ @Seeds there is absolute time - now. Observer perception is another thing here, but not absence of objective time. You might imagine that as ability to generate prediction what observer will see at arrival at any time at any destination point. $\endgroup$
    – MolbOrg
    Aug 17, 2016 at 23:35
  • $\begingroup$ Given that teh univers is infinite (not the observable universe) the answer is that any map must be infinite as well to store the information. So NO. $\endgroup$
    – lijat
    Aug 18, 2016 at 6:58

6 Answers 6


The challenge with the galactic far-side question was in overcoming obstruction from the core and light speed lag measured in tens of thousands of years. This question, given FTL, has a straightforward answer:

Construct a universe-wide grid of sensors no more than 1,000 LY apart. Each sensor regularly pings its nearest neighbours (at FTL comm speeds) and they compare positions for each star within their overlapping monitoring areas. This eliminates light lag because each point (star) is being seen from multiple positions at known distances, so triangulation solves the problem. (They also compare their own relative positions and adjust if needed.) Within orbit of each star or other significant feature construct another handful of sensors to monitor planetary bodies and such, again comparing positions among their local neighbours. At local levels they're only dealing with light lag measured in fractions of a day.

With FTL travel and communication the civilization wouldn't rely on senses or maps based on the natural light from a distant point, but rather from the sensors nearest the target (which update their maps in real time). Your eyes might tell you a star is there, but your FTL maps will know better.

  • $\begingroup$ Permanent sensors are not necessary, it's enough to make snapshot one time to predict locations for long time. $\endgroup$
    – MolbOrg
    Aug 17, 2016 at 23:45
  • $\begingroup$ @MolbOrg Predicting n-body interactions (in which n is > a trillion) based on a single snapshot is asking for trouble. $\endgroup$
    – rek
    Aug 18, 2016 at 15:19
  • $\begingroup$ you missed the point, different points of that snapshot give information about different point in time for particular star - so you do not have to predict nbody interaction, you just measure vector velocity and position. Most stars moving slow compared to distances between them, and just measuring position will be accurate for 1000-2000 years with accuracy 0.1ly. Adding velocity will make it accurate for longer time. Maybe millions of years - interaction between stars is weak. Calculation can be optimized, also it rather first pass, nothing stops from having map unit in each system. $\endgroup$
    – MolbOrg
    Aug 19, 2016 at 7:57

If you have FTL capabilities, it only makes sense that the beacons would also have the capability of communication similarly either by an FTL comm array or with FTL drones going to and fro constantly. If a beacon blinked out you'd send a drone to deliver another - might be a super nova or something you don't want to experience first hand.

The map would no longer be based on objects you see with the naked eye but with images produced by the information being collected/transmitted by the beacons. You'd essentially have a navigate by wire system where visual navigation wouldn't come into play until you were in local space.

We do the same thing on earth today with our satellite and beaon systems, producing virtual navigational aids for pilots around the world.


There are two different scenarios in your question and I'm not sure which you're asking about so I'll answer them both.

  • You want an accurate map that shows the "current" location of everything in the galaxy

And you have tracking beacons attached to everything to aid you with this. This is only possible if these tracking beacons can communicate using your ftl-travel method, if they were limited to the speed of causality there is no way they could communicate a change in location instantly to someone lightyears away.

There is one exception to this. If you ignore the potential changes resulting from randomness (in quantum mechanics there is true randomness and it can affect macroscopic environments as well, as proven by the simple fact that we can observe such randomness and act on it), you could, given a very powerful conputer, calculate the current position of everything based on a one-time shot of the universe. In this case the humans who go around attaching tracking beacons to everything could just record its position, speed, direction etc. and start the simulation once they gathered enough information (they might have to initialize it to account for different objects being recorded at different times, but this shouldn't be a problem so long as they make sure they there were no interactions between two objects while one was recorded and the other wasn't).

This simulation could either be run on a central computer which has to communicate through ftl-based methods with every client or run on every client in a manner that completely avoids ftl communication but requires every new client to wait until the nearest existing client can send it the current state of the universe.

  • You want a map that shows not what is, but of what can be seen

This would mean that the map would be different for every client but accurately represent the current state of the universe as seen from that point. On this map you could still see a star 5000 lightyears away even if it actually went supernova 3000 years ago.

This is fairly easy to accomplish and requires no ftl communication but requires you to wait a while if you want data about faraway systems.

You can do this by placing a tracking beacon on everything and just have it broadcast its position. Assuming communication methods which travel at the same speed as light this will cause the relevant data to arrive exactly when the object can be seen as described.

Of course, if an object is 4000 light years away but the tracking beacon was only installed 2000 years ago that mean you're not yet receiving any data. To counteract this you could combine this method with the "simulation" method described above to have a (backwards) simulation bridge the timespan between installment of transmitter and first transmission receival. This would require the simulation to be run on every client or the central computer to generate a new universe state for every map request.

  • $\begingroup$ I'm writing this as I'm pretty tired so if something is unclear or I forgot to mention a detail just leave me a comment. I'll correct it tomorrow (that's in ~9 hours) at the latest (assuming no unforeseen majorly disruptive event, such as something that causes my death, occurs. If an earthquake makes the roof collapse above me while I sleep and kills me I'm afraid you'll be left to fend for yourself, but maybe another member of this community will be gracious enough to edit the answer in my stead). $\endgroup$
    – Annonymus
    Aug 17, 2016 at 20:14

To some extent, there already exists a complete map of the universe: the universe itself. In order to avoid a variant of the Coastline Paradox, you have to decide a few things about your map:

  • What axes you are mapping. For instance, do you only care about solid mass? Solid mass above a certain amount? Solid mass at a single point in time or across time? Regular matter or 'dark' matter, too? Does a thing being mapped need to have its trajectory, mass, spin, and velocity mapped as well? Or just the volume it occupies at a particular point in time?
  • What navigation type are you optimizing for? Understanding the state of the universe at a given point in time will be a different beast than understanding where things will be at x point in time in the future. Somewhat tangential, but the map of Napolean's march into Russia is great for mapping a population across time and space... but the space/geography aspect is rolled up into a single linear axis.
  • What distortions are you willing to accept in order to project the actual universe into a map? It's well known that the Mercador projection, for example, skews the size of continents (in particular in the northern hemisphere) quite a bit.

Given a willingness to make tradeoffs (less precision in your map, for instance, in order to avoid having to actually keep track of every particle in the universe), then it just becomes a matter of making those choices and spending the energy collecting the data. The only real limitation here is causality: instant FTL travel has some major problems. In this case the problem is a basic quantum one: that by measuring a thing you affect that thing. In this case, in a big way. Say you 'instantly' teleport to a star to mark where it is. By being there you are subtly affecting its trajectory and course. On a galactic scale this 'mapping' will have significant effects - and so you're compromising the goal of knowing where everything is by having to, in effect, move everything.

There is another quantum problem, the Uncertainty Principle, wherein you can know the speed or the location of an object, but not both. Combined with the limits of the speed of light, your map will quickly become out-of-date, no matter which slice of time you select and how much processing power you throw at it.

Finally, there are some more mundane matters to consider:

  • Space is expanding, meaning your map has to 'expand' as well.
  • Gravity bends space, most easily seen in a 'lensing' effect. Because of the relative nature of bodies in space, this means your map has to be particularly fluid - not only because objects are flying around in it, but their mass changes the shape of the volume itself.
  • Singularities are a problem, specifically because getting information out of singularity is quite hard. (If your teleportation device defeats this property, there are larger physical problems.)
  • Your dataset is, quite literally, astronomically large. There are 10 billion galaxies in the universe, with an average of 100 billion stars each. That is one billion trillion stars alone, never mind 10x numbers of planets and probably 1000x (at a minimum) moons, comets, asteroids, etc. Assuming you could store all the needed information about a star (location, size, etc.) in one byte of information (that is one eight-bit number, or between 0 and 255), the storage capacity on Earth right now would just about cover it. However, we know this is impossible based on the Pigeonhole Principle: such a storage scheme would necessitate there be no more than 256 stars. Following this logic back suggests that you need at least a ten bytes per mapped object, for unique identification alone.

In short, barring your 'map' being the actual universe, the answer to your question is 'no'. You're a lot better off choosing a single axis of thing you want to map, and toss the vast majority of information (just like the UK does not map where every single rock on its coast is). You can create predictive models to 'mostly' know where everything is, and chances are that will turn out to be 'close enough'.

  • $\begingroup$ given a billion trillion stars cant you number them 1 to a billion trillion? 9 bytes required. $\endgroup$ Aug 17, 2016 at 22:42
  • $\begingroup$ Oh! Yeah, my bad. So only 10 zettabytes to get a unique identifier, nevermind location, for all the stars. (only.) $\endgroup$ Aug 17, 2016 at 22:44
  • $\begingroup$ Given how much capacities have grown recently and assuming they have 1 atom storage tech, the map could weigh a few grams. $\endgroup$ Aug 17, 2016 at 22:45

The universe is constantly moving all the time but it is moving inertially. This means that we can predict where objects will be by where they where. A real time star chart would be a database of the equations of stars and objects in the universe that we have observed before, we already have something like this. The way we currently measure stuff is the distance from the Solar System, our origin so to speak. When we move to a different origin all the relative distances and positions change, thus a different sky. For a civilization that has obtained interstellar travel the math for this is (hopefully) trivial.

So lets say that your civilization pops into a new star system and they want add the local stars to their map, they know how far they traveled from their home and they know how far the stars are from them. They can figure how far the stars are from their home with some high school trig.

When you observe stars you have to deal with time dilation due to the fact that light travels really far distances as you said. When you find a star that is inconsistent with your map says where it should be, go with the closer observation. The closer you are to a object the less time shift you deal with and the less prone you are to errors.

This is all based off of what we do now but if you already have a beacon on every single object then you don't have to deal with observations you just check where the beacons are and your good, everything files into a 3d plane.


Same as @rek answer in general

With FTL and limited speed of light you can fast determine vector of velocity of each star. Just by taking visual snapshots each 10-100 ly - a 3d grid-mesh.

Those view in each point allows to determine distance and vector velocity of each star. As example star X will be seen with 10ly mesh up to distance 100ly (as example) from few thousands points, up 100 years in the past of this star. 10y, 20y, 30y etc in the past of that star.

Diameter of Universe is at least 91 billion light-years, so with 10ly mesh it needs 3.8151e+29 points.

Send Von Neumann probes which will make map for you.

There is no problem with this task, except instant travel.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .