Author Topic: Lockheed Martin SPIDER optical system  (Read 13208 times)

Offline gosnold

  • Full Member
  • ****
  • Posts: 572
  • Liked: 243
  • Likes Given: 2116
Lockheed Martin SPIDER optical system
« on: 01/27/2016 07:40 pm »
Lockheed seems to have developed technology to create "flat telescopes": optical systems made of only a flat sensor array, without mirrors or lenses.



http://www.telegraph.co.uk/news/science/space/12117192/The-telescope-gets-its-first-major-upgrade-in-centuries.html?sf19788618=1

They call it SPIDER (Segmented Planar Imaging Detector for Electro-optical Reconnaissance).
It seems to be an 2D interferometer, almost like a lytro camera but with coherent light.
« Last Edit: 01/27/2016 10:07 pm by Chris Bergin »

Offline llanitedave

  • Senior Member
  • *****
  • Posts: 2286
  • Nevada Desert
  • Liked: 1545
  • Likes Given: 2052
Re: Lockheed Martin SPIDER optical system
« Reply #1 on: 01/27/2016 11:10 pm »
They've got lenses, they're just very small, one per pixel it seems.
"I've just abducted an alien -- now what?"

Offline JamesG123

  • Full Member
  • *
  • Posts: 133
  • USA
  • Liked: 49
  • Likes Given: 20
Re: Lockheed Martin SPIDER optical system
« Reply #2 on: 01/27/2016 11:50 pm »
More the model of the insect compound eye than the human eye.  I bet calibrating all those elements is the tricky part.

Offline Asteroza

  • Senior Member
  • *****
  • Posts: 2836
  • Liked: 1084
  • Likes Given: 33
Re: Lockheed Martin SPIDER optical system
« Reply #3 on: 01/28/2016 12:59 am »
Um, this doesn't avoid the light collection issue though, right? Which means your sensor array is limited by flat panel packing constraints, and this probably needs to be monolithic, so roughly limited to payload fairing diameter still? I wonder if this could be combined with the MOIRE project's membrane lens to improve light collection?

Offline JamesG123

  • Full Member
  • *
  • Posts: 133
  • USA
  • Liked: 49
  • Likes Given: 20
Re: Lockheed Martin SPIDER optical system
« Reply #4 on: 01/28/2016 01:31 am »
Not necessarily. At astronomical distances parallex is a non-issue. So as long as you can get individual elements within arrays precisely calibrated (collimated?) with each other and keep them that way,  you could build up an an array of arrays which could fold up the same way the JWST's main mirror does. Each array panel would increase the light gathering ability of the telescope. And probably the angular resolution as well (since its all done via software image processing).

Offline gosnold

  • Full Member
  • ****
  • Posts: 572
  • Liked: 243
  • Likes Given: 2116
Re: Lockheed Martin SPIDER optical system
« Reply #5 on: 01/28/2016 06:32 am »
Um, this doesn't avoid the light collection issue though, right? Which means your sensor array is limited by flat panel packing constraints, and this probably needs to be monolithic, so roughly limited to payload fairing diameter still? I wonder if this could be combined with the MOIRE project's membrane lens to improve light collection?

Apparently the interferometer is built under the lenses, using silicon channels, so if you want an unfolding solution with the same properties as a single panel you have to connect these channels across the different panels, which seems complicated. So for now it's probably monolithic if you want to keep it diffraction-limited.

However if you only care about collection area and not resolving power there's no reason you could not unfold an array like a solar panel.

You could launch a monolithic array of size fairing width x fairing height though, and consequently get extra resolution in on direction.
« Last Edit: 01/30/2016 08:00 am by gosnold »

Offline Nilof

  • Full Member
  • ****
  • Posts: 1177
  • Liked: 597
  • Likes Given: 707
Re: Lockheed Martin SPIDER optical system
« Reply #6 on: 01/29/2016 09:44 pm »
Here's the original source from LockMart, most press articles leave a lot to be desired: http://www.lockheedmartin.com/us/innovations/011916-webt-spider.html

« Last Edit: 08/05/2017 02:31 pm by Nilof »
For a variable Isp spacecraft running at constant power and constant acceleration, the mass ratio is linear in delta-v.   Δv = ve0(MR-1). Or equivalently: Δv = vef PMF. Also, this is energy-optimal for a fixed delta-v and mass ratio.

Offline Hanelyp

  • Full Member
  • ***
  • Posts: 368
  • Liked: 65
  • Likes Given: 252
Re: Lockheed Martin SPIDER optical system
« Reply #7 on: 01/30/2016 02:19 pm »
They intend to combine many low resolution images in post processing to a single high resolution image?  I have my doubts that will work.

Offline gosnold

  • Full Member
  • ****
  • Posts: 572
  • Liked: 243
  • Likes Given: 2116
Re: Lockheed Martin SPIDER optical system
« Reply #8 on: 01/30/2016 03:43 pm »
They intend to combine many low resolution images in post processing to a single high resolution image?  I have my doubts that will work.

It's not in post processing, it's with interferometry. They combine light waves (phase and amplitude), not just images (amplitude only). That's why it can work.

Offline Nilof

  • Full Member
  • ****
  • Posts: 1177
  • Liked: 597
  • Likes Given: 707
Re: Lockheed Martin SPIDER optical system
« Reply #9 on: 01/31/2016 01:13 am »
My guess is that the aperture is similar to diffractive lenses, but using interferometry to fold it up so that the depth isn't to big. I don't think it is designed to work at more than one wavelength, and it likely has significant drawbacks compared to regular optics. The press release looks intentionally vague.

The key issue is that to beat the uncertainty principle, when measuring a photon, in order to measure it's momentum you have to delete the information about where on the detector it hit, and the measurements have to be made non-local somehow.

You can do this by interfering the light that entered at two spots like in a classical mirror system, or you could interfere the light coming in with light coming from a laser source to measure in the momentum basis. However, the latter essentially means you have to know the frequency and amplitude of the incoming light, and there would be strong coherence requirements.
For a variable Isp spacecraft running at constant power and constant acceleration, the mass ratio is linear in delta-v.   Δv = ve0(MR-1). Or equivalently: Δv = vef PMF. Also, this is energy-optimal for a fixed delta-v and mass ratio.

Offline MikeFidler

  • Member
  • Posts: 6
  • Jagna, Bohol, Philippines
  • Liked: 2
  • Likes Given: 3
Re: Lockheed Martin SPIDER optical system
« Reply #10 on: 02/01/2016 09:57 am »
Very good infographic pdf;
https://www.nasa.gov/sites/default/files/files/Ben_Yoo_LowMassPlanarPhotonicImagingSensor.pdf
A more in depth write up;
http://www.amostech.com/TechnicalPapers/2013/Space-Based_Assets/KENDRICK.pdf

Yes, it will work except some light loss do to design but probable will be correct once the design is improved.
Good possibility that correction for atmospheric turbulence is easily done because of the small size of sensors.
The big deal is the system should be scalable and the PIC can be mass produced, in fact Vice President Biden was at the launching of a new lab and manufacturing facility for producing these photonic devices - Military Money big time! ;)

Offline as58

  • Full Member
  • ****
  • Posts: 835
  • Liked: 300
  • Likes Given: 186
Re: Lockheed Martin SPIDER optical system
« Reply #11 on: 02/04/2016 07:29 pm »
Very good infographic pdf;
https://www.nasa.gov/sites/default/files/files/Ben_Yoo_LowMassPlanarPhotonicImagingSensor.pdf
A more in depth write up;
http://www.amostech.com/TechnicalPapers/2013/Space-Based_Assets/KENDRICK.pdf

Yes, it will work except some light loss do to design but probable will be correct once the design is improved.
Good possibility that correction for atmospheric turbulence is easily done because of the small size of sensors.
The big deal is the system should be scalable and the PIC can be mass produced, in fact Vice President Biden was at the launching of a new lab and manufacturing facility for producing these photonic devices - Military Money big time! ;)

Thanks especially for the second link, which finally helped me understand what this is about. I also found a newer paper on the same website:

http://www.amostech.com/TechnicalPapers/2015/Optical_Systems/Duncan.pdf

Basically (if I understood correctly), they're building long 1-D strips of tiny Michelson interferometers with PIC; to sample the whole uv-plane they're planning to use multiple independent strips in different directions. That allows high resolution, but it comes with a rather massive sensitivity penalty. I guess that's why they're pitching it for Earth observation and planetary missions instead of other fields of astronomy where you really need the sensitivity.

Offline Asteroza

  • Senior Member
  • *****
  • Posts: 2836
  • Liked: 1084
  • Likes Given: 33
Re: Lockheed Martin SPIDER optical system
« Reply #12 on: 02/04/2016 11:19 pm »
The papers seem to imply that linking sensors together via fiber optic cable is doable, in the sense of traditional interoferometer linkage, so an accordion stack seems doable.

Offline Stardhingy

  • Member
  • Posts: 24
  • Liked: 1
  • Likes Given: 1
Re: Lockheed Martin SPIDER optical system
« Reply #13 on: 02/05/2016 03:45 am »
They intend to combine many low resolution images in post processing to a single high resolution image?  I have my doubts that will work.

It's not in post processing, it's with interferometry. They combine light waves (phase and amplitude), not just images (amplitude only). That's why it can work.

It can definitely could work in post, that's exactly how lightfield cameras work. (Interferometry is obviously superior for scientific observations though)

http://imgur.com/atZr9IC

Offline MikeFidler

  • Member
  • Posts: 6
  • Jagna, Bohol, Philippines
  • Liked: 2
  • Likes Given: 3
Re: Lockheed Martin SPIDER optical system
« Reply #14 on: 02/05/2016 06:59 am »
These are not fiber optics but solid state waveguides all on one integrated photonic circuit and that is what is so new about this! If you read the infographic on page 12-20, all of this will be shrunk into a Photonic Integrated Circuit (PIC)
board. Take a look at the latest images from Lockheed Martin Coherent Technologies, they have changed from a 1mm lens to around 8mm and made it into a square array and they plan on using a square lens so that there will be limited light loss.
SPIDER’s photonic integrated circuits do not require complex, precision alignment of large lenses and mirrors. That means less risk on orbit. And its many eyes can be rearranged into various configurations, which could offer flexible placement options on its host. Telescopes have always been cylindrical, but SPIDER could begin a new era of different thin-disk shapes staring in the sky, from squares to hexagons and even conformal concepts.

Offline Nilof

  • Full Member
  • ****
  • Posts: 1177
  • Liked: 597
  • Likes Given: 707
Re: Lockheed Martin SPIDER optical system
« Reply #15 on: 02/05/2016 02:26 pm »
They intend to combine many low resolution images in post processing to a single high resolution image?  I have my doubts that will work.

It's not in post processing, it's with interferometry. They combine light waves (phase and amplitude), not just images (amplitude only). That's why it can work.

It can definitely could work in post, that's exactly how lightfield cameras work. (Interferometry is obviously superior for scientific observations though)

http://imgur.com/atZr9IC

That won't get you far-field resolution. Focus length =/= diffraction, focus length can be described with plain geometric optics while diffraction is related to the behavior of light as a wave. The interferometry aspect is required to obtain a higher resolution, because otherwise you simply aren't collecting the information you need from the light.
« Last Edit: 08/06/2017 01:11 am by Nilof »
For a variable Isp spacecraft running at constant power and constant acceleration, the mass ratio is linear in delta-v.   Δv = ve0(MR-1). Or equivalently: Δv = vef PMF. Also, this is energy-optimal for a fixed delta-v and mass ratio.

Offline MikeFidler

  • Member
  • Posts: 6
  • Jagna, Bohol, Philippines
  • Liked: 2
  • Likes Given: 3
Re: Lockheed Martin SPIDER optical system
« Reply #16 on: 02/06/2016 02:56 am »
Good point Nilof, wondering if the idea for a Quantum telescope could be integrated into the SPIDER concept:

http://physicsworld.com/cws/article/news/2014/apr/29/quantum-telescope-could-make-giant-mirrors-obsolete

Found a good infographic on Plasmonics, Metamaterials and Their Applications in Light Manipulations:

http://s3.amazonaws.com/sdieee/1722-IEEE_PhotonicsSociety-2013_Nov.pdf

Have not read anything on using this in telescopes just microscopes, do you have anything relating to it and telescopes? :)

Offline MikeFidler

  • Member
  • Posts: 6
  • Jagna, Bohol, Philippines
  • Liked: 2
  • Likes Given: 3
Re: Lockheed Martin SPIDER optical system
« Reply #17 on: 02/07/2016 01:56 am »
Good 64 page infographic on interferometry, see page 26 on nulling:
https://lagrange.oca.eu/IMG/pdf/patru.pdf

14 page infographic on Photonics interferometry: Dragonfly:
http://s14.sydney.edu.au/science/physics/staffinfo/nanoscience/nanotechnology_meeting_leon-saval.pdf

Offline Dao Angkan

  • Full Member
  • **
  • Posts: 233
  • Liked: 77
  • Likes Given: 44

Offline KelvinZero

  • Senior Member
  • *****
  • Posts: 4286
  • Liked: 887
  • Likes Given: 201
Re: Lockheed Martin SPIDER optical system
« Reply #19 on: 08/05/2017 03:08 am »
Phone cameras are about to get a lot better  :)

Offline Nilof

  • Full Member
  • ****
  • Posts: 1177
  • Liked: 597
  • Likes Given: 707
Re: Lockheed Martin SPIDER optical system
« Reply #20 on: 08/05/2017 02:07 pm »
Lockheed martin's release on their site:
http://news.lockheedmartin.com/2017-08-03-Lockheed-Martin-Closes-In-On-Shrinking-the-Telescope-Reveals-First-Images-from-Instrument-as-Thick-as-a-Pen-Cap

They also linked to some pictures taken with their array prototype:
« Last Edit: 08/05/2017 02:37 pm by Nilof »
For a variable Isp spacecraft running at constant power and constant acceleration, the mass ratio is linear in delta-v.   Δv = ve0(MR-1). Or equivalently: Δv = vef PMF. Also, this is energy-optimal for a fixed delta-v and mass ratio.

Offline Propylox

  • Member
  • Posts: 99
  • Colorado
  • Liked: 15
  • Likes Given: 7
Re: Lockheed Martin SPIDER optical system
« Reply #21 on: 08/05/2017 04:41 pm »
... The big deal is the system should be scalable and the PIC can be mass produced, in fact Vice President Biden was at the launching of a new lab and manufacturing facility for producing these photonic devices - Military Money big time! ;)
Yep, this is a DARPA program for UAV surveillance which is hoping to expand into orbital use.
My guess is that the aperture is similar to diffractive lenses, but using interferometry to fold it up so that the depth isn't to big. I don't think it is designed to work at more than one wavelength, and it likely has significant drawbacks compared to regular optics. ...
They intend to combine many low resolution images in post processing to a single high resolution image?  I have my doubts that will work.
It's not in post processing, it's with interferometry. They combine light waves (phase and amplitude), not just images (amplitude only). That's why it can work.
Definitely going to need post-processing based on those diffracted results (note: as a DARPA program we can only be sure those are not their current results). These images appear to confirm Nilof's concern of interferometry's single wavelength sensitivity (likely IR for military applications) and others' resolution concerns.
« Last Edit: 08/05/2017 04:46 pm by Propylox »

Offline edzieba

  • Virtual Realist
  • Senior Member
  • *****
  • Posts: 6104
  • United Kingdom
  • Liked: 9330
  • Likes Given: 39
Re: Lockheed Martin SPIDER optical system
« Reply #22 on: 08/08/2017 05:17 pm »
It's an interesting device, kind of a hybrid between a Lightfield imager and a Phased Array imager (the current line array prototype is a Synthetic Aperture imager in the traditional sense due to being physically scanned). Each lens element is not independent so it's not a Lightfield imager, instead combining multiple lens elements into a single sensor with a known phase relationship using the PIC. But those monolithic arrays then need to be combined to create any appreciable aperture, and fabbing a really big PIC is going to be technically infeasible, so it makes sense to instead combine multiple of these arrays into a larger imager. Combining those blocks is likely going to need the same techniques as combining imaging elements as a Lightfield.
I wonder if they're using Telecentric lenses for the individual elements? If so, the whole array could be thought of as a big Telecentric lens making combination trivial, but losing most/all the computational imaging benefits and limiting you to only contiguous planar arrays.

Amusingly, the current rotating-line-array imager is a bit like the old TOSCA imaging idea for turning a spinning-mirror conical scan seeker head into an imaging seeker head without that pesky expensive focal plane array. Except now the array is cheaper than all those pesky optical gubbins you need to slap on the front!

Offline Propylox

  • Member
  • Posts: 99
  • Colorado
  • Liked: 15
  • Likes Given: 7
Re: Lockheed Martin SPIDER optical system
« Reply #23 on: 08/10/2017 02:27 am »
Phone cameras are about to get a lot better  :)
Low resolution black-n-white isn't an improvement and SPIDER doesn't gather more photons per area than a traditional lens to shrink a camera. As described, it'd make every camera or telescope ever used worse - including ones from a hundred years ago. What it has going for it is form factor. That's it.

 

Advertisement NovaTech
Advertisement Northrop Grumman
Advertisement
Advertisement Margaritaville Beach Resort South Padre Island
Advertisement Brady Kenniston
Advertisement NextSpaceflight
Advertisement Nathan Barker Photography
1