Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!

 


Forgot your password?
Close
typodupeerror
×
Space

Astronomer Makes Navigation System For Interstellar Space Travel (sciencealert.com) 68

Posted by BeauHD from the sky-is-the-limit dept.
rushtobugment shares a report from ScienceAlert: Using the positions and shifting light of stars, both near and far, astronomer Coryn A.L. Bailer-Jones has demonstrated the feasibility of autonomous, on-the-fly navigation for spacecraft traveling far beyond the Solar System. "When traveling to the nearest stars, signals will be far too weak and light travel times will be of order years," Bailer-Jones wrote in his paper, which is currently available on the preprint server arXiv, where it awaits peer review from the astronomy community. "An interstellar spacecraft will therefore have to navigate autonomously, and use this information to decide when to make course corrections or to switch on instruments. Such a spacecraft needs to be able to determine its position and velocity using only onboard measurements."

With a catalog of stars, Bailer-Jones was able to show that it's possible to work out a spacecraft's coordinates in six dimensions -- three in space and three in velocity -- to a high accuracy, based on the way the positions of those stars changes from the spacecraft's point of view. "As a spacecraft moves away from the Sun, the observed positions and velocities of the stars will change relative to those in a Earth-based catalog due to parallax, aberration, and the Doppler effect," he wrote. "By measuring just the angular distances between pairs of stars, and comparing these to the catalog, we can infer the coordinates of the spacecraft via an iterative forward-modeling process."

Bailer-Jones tested his system using a simulated star catalog, and then on nearby stars from the Hipparcos catalog compiled in 1997, at relativistic spacecraft speeds. Although this is not as accurate as Gaia, that's not terribly important - the aim was to test that the navigation system can work. With just 20 stars, the system can determine the position and velocity of a spacecraft to within 3 astronomical units and 2 kilometers per second (1.24 miles per second). This accuracy can be improved inverse to the square root of the number of stars; with 100 stars, the accuracy came down to 1.3 astronomical units and 0.7 kilometers per second. [...] The system hasn't taken stellar binaries into consideration, nor has it considered the instrumentation. The aim was to show that it could be done, as a first step towards actualizing it. It's even possible that it could be used in tandem with pulsar navigation so that the two systems might be able to minimize each other's flaws.
This discussion has been archived. No new comments can be posted.

Astronomer Makes Navigation System For Interstellar Space Travel More

Astronomer Makes Navigation System For Interstellar Space Travel

Comments Filter:

  • It can’t determine roll, pitch, and yaw? Seems pretty simple.

    • First I suppose they they need to figure how to get more air in space.

    • Re:Why only six dimensions? (Score:4, Insightful)

      by sabbede ( 2678435 ) on Wednesday March 24, 2021 @07:48AM (#61192248)
      Orientation is a trivial local issue. it doesn't have an effect on position or velocity, so it's not really relevant to the project. That said, look out a window. Which of the reference stars you see and how they are arrayed gives your orientation.
    • You can those (and the 3 linear accelerations) from standard instruments like gyroscopes and accelerometers. You could integrate those to get position and velocity, but it probably would become usefully inaccurate very quickly.

  • Finally! (Score:5, Informative)

    by Ecuador ( 740021 ) on Wednesday March 24, 2021 @05:18AM (#61191978) Homepage

    Finally! The last piece of technology needed for interstellar travel! We are launching tomorrow, boys!

  • Where is the uncertainty? (Score:4, Interesting)

    by JoeRobe ( 207552 ) on Wednesday March 24, 2021 @06:01AM (#61192038) Homepage

    The sciencealert article don't mention what the source of error is, but this seems quite important. He assumes an angular measurement precision of 1" ("modest"), and that the accuracy would increase 1000-fold if the spacecraft has a 1 mas resolution.

    So using more stars would improve precision by sqrt(n), and using more precise angular measurements would improve precision linearly.

    I assume somewhere in there it's eventually limited by the accuracy of the database as well.

    • Author assumes stars are moving on unaccelerated or at least predictable orbits; this is not true, it may be a good approximation for a limited time (thus maybe for relativistic travel speeds), but certainly not for any realistic travel speeds..
      • But the predictability of their movement is how we learned that we are on a large body that orbits another, how we learned when to plant crops, how we learned math, and how we learned to navigate on the ocean. It's how we can say for sure what the sky looked like on a specific day from a specific position five thousand years ago, and have a pretty good idea about what it will look like five thousand years from now.

        So long as you have enough reference points, one star unexpectedly exploding while another

      • Re: (Score:3)

        by XXongo ( 3986865 )

        Author assumes stars are moving on unaccelerated or at least predictable orbits; this is not true, it may be a good approximation for a limited time .

        Yes, it is true for a "limited" time, and that limited time is on the order of 10,000,000 years.

        The orbits of the stars around the center of the galaxy are well known.

        • not really, the uncertainty in e.g. the local circular velocity is in the second digit.. furthermore the navigation errors quoted in the paper depend on the relative positions of 20 or so nearby star and their relative motion depends on the local mass distribution, which is also poorly constrained (error in the first digit)

          • Re: (Score:2)

            by XXongo ( 3986865 )

            not really, the uncertainty in e.g. the local circular velocity is in the second digit.

            Yes, but the statement I was addressing was saying that the acceleration was unpredictable.

            furthermore the navigation errors quoted in the paper depend on the relative positions of 20 or so nearby star and their relative motion depends on the local mass distribution, which is also poorly constrained (error in the first digit)

            This deviation of the acceleration from simply orbiting around the center of the galaxy is primarily due to perturbation by the gravitational pull of giant molecular clouds. Over a time scale of tens of millennia, these may make a difference in the position of nearby stars large enough to measure, but over a lower time scale, the deviation from predicted motion is going to be undetectable.

    • This is just a stop gap until we start using the interstellar beacon system ;-)

      • This is just a stop gap until we start using the interstellar beacon system ;-)

        Need to hurry up and more psykers to the Golden Throne!

    • He also assumes that his system is locked on to 20 stars and somehow remains locked. If the craft wakes up from sleep every year or ten, then it needs to figure out its orientation and find those 20 stars. That would not be easy.

      • If the craft wakes up from sleep every year or ten, then it needs to figure out its orientation and find those 20 stars. That would not be easy.

        It is not only easy, it is done all the time. That's how most spacecraft measure their orientation. Here's an article on how it was done back in 1964: https://pubmed.ncbi.nlm.nih.go... [nih.gov]

      • That would not be easy if you were magically transported (or warped) to a different sector of the galaxy.
        As that is not a probability to account for in a scientific writing, you just need to cope with "normal" movements - i.e. on the order of light-seconds per minute or less, and basically you need to only refine your position.
        To make a car analogy, you're in your car with GPS and go into a low coverage area. While your GPS localization precision jumps from 10 meters to 500 meters, the included GPS service

        • Re: (Score:3)

          by jabuzz ( 182671 )

          They tend to use just the mapping data and speed because you are most likely on a road. Try it driving through a tunnel where GNSS coverage is zero. You will also sometimes when passing or taking grade separated junctions see it snap your position from the main road to the slip road or vice versa, especially if you fail to take the junction or do take the junction by mistake.

          • If "they" are GPS units built into phones (as opposed maybe to built into cars), not true; I have often used my phone's GPS on hikes.

    • Re: (Score:2)

      by Zocalo ( 252965 )
      I'd guess that in practice that wouldn't really matter. A margin of error measured in a few AU is nothing in terms of the kinds of stellar distances where this approach would be used, assuming that a more conventional approach based on pre-calculating the trajectory and any major course corrections would likely be used for destinations within the local stellar group. In both cases though, my assumption would be that you'd need to have a second system on-board to handle in-system navigation that could star
    • I'd think it was just as it is for a sailor navigating at night - the accuracy of your charts and the precision of your tools.

      • For sailors, they're trying to find their location on a more or less spherical Earth, where every point on Earth has a different zenith at any given time, and the position of the horizon is mathematically determinable from your height above the water. Whereas for interstellar spacecraft, you're trying to fly between stars, and there is no up, nor any horizon.

        But maybe I misunderstand what you're saying.

    • Measuring versus more stars does not make it more precise.

      Having more stars to pick from, lets you pick pairs which give a precise measurement.

      Big difference.

      Think about land based navigation. You have a precise map wich shows 2 towers in 10 miles distance. The two towers have 400yards distance to each other. To pin point your location you measure the angle to those towers versus true north. Unfortunately your measuring equipment, or your way to hold the compass gives you an error of +/- 5Â for each to

  • How is this different from 'star trackers' that you today can buy and put on satellites? I know the satellite EDRS-C [1] use a star tracker, primarily for attitude estimation, but if I remember correctly the star tracker also outputs a position estimate.

    Note: Each of the two star trackers (STR) in EDRS-C is basically a camera, a computer and a database of star locations so it can compute the attitude/position. The satellite have two STR:s for redundancy, each in different locations so they can't both acci

    • Re: (Score:2)

      by wfj2fd ( 4643467 )
      Most earth based star trackers aren't moving at relativistic speeds, which changes the game a bit, and the relative positions of the stars can be very different. We're not significantly moving between stars, the relative positions / angles between them doesn't change too much, so it's easier to identify a star or cluster of stars. So we can think of the stars as just points on a sphere and that greatly simplifies the calculations. When moving between things, it makes it more difficult. Think about what th

      • Most space based trackers aren't moving at relativistic speeds either. The Voyager 1 probe has a velocity of some 15 km/s - with the lightspeed at 300,000 km/s.

      • At different relativistic speeds, the positions would look different,
        No they would not. Why would they, lol?
        and because you don't know your speed, you have to solve for that as well as your position.
        You know your speed precisely, because you burned your engine a certain time to get exactly that speed.

    • How is this different from 'star trackers' that you today can buy and put on satellites?

      Star trackers on satellites assume that the stars are infinitely distant. They give you attitude but not position.

      I know the satellite EDRS-C [1] use a star tracker, primarily for attitude estimation, but if I remember correctly the star tracker also outputs a position estimate.

      Nope, just attitude.

      Inside the solar system, you can get position by looking at the sun and at planets (and asteroids), and comparing their position to an ephemeris. This capability isn't in the usual star tracker algorithms because it's nowhere near precise enough to be of any use in Earth orbit, but some planetary probes do this (although the primary position measurement on planetary probes i

    • Re: (Score:2)

      by ledow ( 319597 )

      You can do star tracking from the ground or in the air. This is just a more-accurate version when at speed, one presumes.

      Old BBC micros used to be used to tell you what you were pointed at with your telescope just using a bog-standard CCD image of a starfield.

      Some guided missiles literally fall back to star-tracking to home in on their target.

      Now, they're based on the assumption that the stars are going to be "as seen from Earth", but I can't help think that if there was a need for it we would have this te

  • It's how the Enterprise finds its way, how Heinlein's astrogators did it, how actual sailors have done it for hundreds of years.

    I'm glad someone is working on the math to actually do it, and it's impressive work, but it's not like this is an astonishing new concept as implied by the tone of the article.

    • It is not. But doing it on earth is more simple than in earth orbit. And then again it gets simple when you are far away - or moving away.
      However, it is pure math, having to write a simulator for it to prove a point makes no real sense.

    • It is nothing like what sailors have done. They have an "up" and a horizon; interstellar spacecraft have neither.

  • Celestial navigation is already being used by SLBMs.

  • 1950s technology, neat! (Score:5, Informative)

    by Xenolith0 ( 808358 ) on Wednesday March 24, 2021 @08:59AM (#61192464)
    The SR-71 had an astroinertial-navigation-system that worked this same way, by tracking stars and comparing them to a known database.

    https://alopez128.tumblr.com/post/15406861921/sr-71-blackbird-astroinertial-navigation-system [tumblr.com]

    A time-consuming primary alignment was done on the ground to bring the inertial components to a high degree of accuracy for the start of a mission. Once in flight, a âoeblue lightâ source star tracker, which could detect and find stars during day or night, would continuously track a variety of stars as the changing aircraft position brought them into view. The systemâ(TM)s digital computer ephemeris contained data on 56 (later 61) stars. The ANS was later modified to speed up the ground alignment and even start it in the air.

    • A number of people have claimed this in this thread, but they--and you--are wrong. An SR-71 is trying to navigate around a spherical object (Earth), and at every point around that Earth there is a different up (although since the SR-71 may be accelerating, it's probably better to determine "up" wrt the horizon), and a known horizon. Furthermore, the SR-71 is essentially unmoving wrt the stars. A spacecraft navigating between stars has neither an "up" nor a horizon, and it's moving (by hypothesis) signifi

    • One of the SR71 books I read said that they started up the celestial navigation inside a hanger and it locked onto a tiny hole in the ceiling.
      Everything else was then based on treating that single hole as one of the stars in its database.

  • Using the positions of 6 stars as coordinates to determine a fixed point. They managed to steal science from the plot of Stargate only 25 years after the film came out.

    • It isn't theft when science fiction is transformed into science reality. It is cool and unexpected while at the same time something we all assume will happen.

  • Also, why do navigation units sound so annoyed when they say "recalculating"?

  • What? (Score:2)

    by ledow ( 319597 )

    1.3 astronomical units = 195m kilometers

    Doesn't appear to be too accurate to me.

    • Doesn't need to be that accurate when you're dealing with interstellar distances.

      1 light year = 63421 AU.

    • When you get near your target star, you switch over to a different navigation system. Like sailors switched from celestial navigation to coastal navigation when they got near the shore.

  • If you're moving at relativistic speeds, don't you need some sort of shields ?A stray piece of dust wold ruin your whole day/

    • No need for shields. You just need the ships of your enemies fleeing ahead of you. This is the real reason for interstellar conquest.

  • Because we will need this within just a few millenia.

    • We already have ships that have left the solar system. There are some others planned for the next couple decades. Ships with humans onboard is still waiting for some breakthroughs in cryonics, but if that tech is possible at all, I would expect it to be viable within a century... our medical tech has been advancing exponentially since 1800s. Barring civilization collapse/extinction from climate change, of course.

  • "We're headed for earth, right?"
    "Well, earth-ish."
    "ish?"
    "Either earth, or the asteroid belt, or... checks notes... the heart of the sun."

  • Does it work in the delta quadrant as well? Just in case...

Slashdot Top Deals

Marvelous! The super-user's going to boot me! What a finely tuned response to the situation!

Close