Stories
Slash Boxes
Comments

SoylentNews is people

SoylentNews is powered by your submissions, so send in your scoop. Only 12 submissions in the queue.

5,000 Robots to Build 3D Map of the Universe

posted by n1 on Wednesday August 10 2016, @07:16PM   Printer-friendly
from the robot-hell dept.

Arthur T Knackerbracket has found the following story:

Five thousand robots will get busy creating a 3D map of millions of galaxies in 2019.

The Dark Energy Spectroscopic Instrument (DESI) has received US Department of Energy (DoE) approval to move from the design phase to construction, which will start next year.

That includes building the 5,000 10 cm-long, finger-width robots which will have the job of aiming fibre-optic cables at galaxies, stars, and quasars.

DESI's builders have just begun a two-month prototype run of the light collection system in Arizona.

Original Submission

 
This discussion has been archived. No new comments can be posted.
5,000 Robots to Build 3D Map of the Universe | Log In/Create an Account | Top | 8 comments | Search Discussion
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.

  • (Score: 1, Informative) by Anonymous Coward on Wednesday August 10 2016, @07:44PM

    by Anonymous Coward on Wednesday August 10 2016, @07:44PM (#386359)

    I'm trying to get my head around what's actually happening here. At first reading I thought an army of robots were going to string fiber optic cables to make a giant physical 3D map "room" of a portion of the universe, kind of like a fiber-optic planet mobile.

    But it turns out that these fiber-optic cables are probably each independent mini-telescopes, all being aimed different places at the same time (within the constraints of their "section").

    However, I thought small apertures cannot get enough resolution to isolate the light of a given object. You need a big lens to get sufficient resolution, due to the quantum oddities of light. How is a small aperture going to isolate the light of two close-together objects?

    By the way, another article, with a diagram:

    http://www.digitaltrends.com/cool-tech/desi-prototype-test-berkely/ [digitaltrends.com]

    • (Score: 2) by Scruffy Beard 2 on Wednesday August 10 2016, @08:17PM

      by Scruffy Beard 2 (6030) on Wednesday August 10 2016, @08:17PM (#386369)

      Sounds like they are simulating between 1 and 6 larger telescopes with a whole bunch of small ones.

      They probably need to make sure the fibre is all the same length and that the robots all point in the same direction. They probably have a very narrow field of view.

      • (Score: 2) by Scruffy Beard 2 on Wednesday August 10 2016, @08:24PM

        by Scruffy Beard 2 (6030) on Wednesday August 10 2016, @08:24PM (#386370)

        Michael H.F. Wilkinson writes:

        The Sloan Digital Sky Survey release 12 has far more objects: 208,478,448 galaxies to be precise. It only contains spectra for 2,401,952 galaxies and 477,161 quasars, so DESI becomes particularly interesting because it captures spectra of so many objects.

        Pointing an optical fibre directly at the sky is of course pointless, as the aperture is pathetically small. The trick is to get a big mirror (4 m in this case), which gathers a lot of photons, and use fibre optics to guide the light to (multiple) spectroscopes. The more usual trick is to use a slit, which captures spectra from a little stripe across the image plane. By using carefully placed fibre optics, the project (as I read it) wants to tap into more of the optical plane, to record spectra of far more objects simultaneously. The robots pick out the right bits of the image from which to obtain spectra. Sounds very interesting

    • (Score: 0) by Anonymous Coward on Wednesday August 10 2016, @08:24PM

      by Anonymous Coward on Wednesday August 10 2016, @08:24PM (#386371)

      small apertures cannot get enough resolution to isolate the light of a given object.

      After thinking about this, it occurred to me it may not matter. If the light of a few objects get mixed together, than statistical methods can be used to isolate the spectrum of each as long as you know or can determine generally what kind of objects are included. The project's main job is to measure the red-shift, not study specific chemicals etc. in the spectrum such that overlapping spectrum will generally not matter as long as the shifts of multiple objects are not close. The handful that have close shifts can probably be followed up using traditional methods.

    • (Score: 0) by Anonymous Coward on Wednesday August 10 2016, @08:25PM

      by Anonymous Coward on Wednesday August 10 2016, @08:25PM (#386373)

      I was surprised to learn from an astrophysicist friend that the way of obtaining spectroscopy of stars in telescope images was literally to place fibre optics on an image screen of the telescope (where you see the stars as an image) that take the light from the image of that star to the stectrometer. Several fibres can be placed at the same time and multiple spectra obtained simultaneously.

      I'd imagined some fancy spectroscopic image sensor would take the place of the camera in a regular telescope, but that doesn't exist apparently.

      This sounds like a big automated version of that.

      • (Score: 0) by Anonymous Coward on Wednesday August 10 2016, @08:48PM

        by Anonymous Coward on Wednesday August 10 2016, @08:48PM (#386381)

        So there is to be a big lens(es) or mirror(s) above or feeding focused light to this bucket-looking thing?

  • (Score: 0) by Anonymous Coward on Wednesday August 10 2016, @10:55PM

    by Anonymous Coward on Wednesday August 10 2016, @10:55PM (#386422)

    Hell of a job you doing, "editors." SMH.

  • (Score: 0) by Anonymous Coward on Wednesday August 10 2016, @11:07PM

    by Anonymous Coward on Wednesday August 10 2016, @11:07PM (#386424)

    those had better be self replicating robots