Of DAS, EOTS etc..

[spazsinbad]

'This article' link does not work for me with two browsers: http://www.wpafb.af.mil/news/story_prin ... =123066261

[bring_it_on]

This article and this data sheet confirm it; the EODAS system uses L3's 4MP sensor.

Specifically; you can see the edges of the FPAs on the new 4" diameter wafer; if you use an image editing or CAD program you can then see for yourself that a 1.2" (30.7mm) sided FPA (as per the datasheet) creates the exact same pattern scale as in the AFRL image.

It's possible that it's just a 1MP sensor; L3 doesn't have a datasheet for their 1MP sensor, and it's not uncommon to use (eg) 4MP arrays downsampled to achieve a more reliable or better quality 1MP image. However, the fact that they don't provide a datasheet for the 1MP sensor, plus the fact that the EODAS sensor is physically larger than the 2MP sensor in the HMDS, and needs to achieve a decent level of resolution over it's ~120 degree FOV means that it's almost certain that it has a resolution of 4MP.

The 16MP sensor is twice the width / height of the 4MP and wafer size is measured in diameter by convention, so it's (unfortunately, and only just for now) not the 16MP sensor.

Here you go

https://www.scribd.com/doc/254279654/Fi ... st-Century

[Dragon029]

'This article' link does not work for me with two browsers: http://www.wpafb.af.mil/news/story_prin ... =123066261

Here's a screencap:

Here you go

https://www.scribd.com/doc/254279654/First-Electro-Optical-System-of-the-21st-Century

Thanks! So at least pre-2006 (do you know when exactly the paper was published? I can only find a reference to 2006 as being the future and 1996 as being the most recent reference used) the DAS system was meant to use 1MP sensors, using microscanning to increase that to a ~4MP final image, dependent on the available processing power. While that'd be okay, I'm hoping that the FPA they use today is a 4MP variant, as that'll reduce processing requirements and give better future software-based growth (to use microscanning again if they have the power present and no desire to drive for arrays with a larger resolution).

[spazsinbad]

Thanks for the info (not that I understand it much ) and here is an old screed about DAS/EOTS from Wiley (Coyote?) Military Avionics Systems. See attached PDF (replicated in .GIF graphic)

Military Avionics Systems
2006 Ian Moir and Allan G. Seabridge; John Wiley & Sons, Ltd. ISBN: 0-470-01632-9

"...The distributed aperture system (DAS) being developed by Northrop Grumman together with BAE SYSTEMS comprises six EO sensors located around the aircraft to provide the pilot with 360° situational awareness information that is detected by passive means. The concept of horizontal coverage of the DAS is depicted in Figure 5.37. The six DAS sensors provide a complete lateral coverage and are based upon technology developed for the BAE SYSTEMS Sigma package (shown in the inset). Key attributes are dual-band MWIR (3–5 mm) and LWIR (8–10 mm) using a 640 x 512 FPA. Each sensor measures ~7 x 5 x 4 in, weighs ~9 lb and consumes less than 20W. Sensor devices with megapixel capability (1000 x 1000) are under development and will be incorporated...."

Source: http://www.helitavia.com/books/Mil%20Av ... ystems.pdf (10.4Mb)

[Dragon029]

Odd; the author of that article believed in 2006 that there would be 6 cameras in a horizontal ring (rather than the current upper rear / forward, lower rear / forward and left / right). Does anyone know when the first F-35 flew with EODAS installed? It might give us another clue as to whether or not they used 1MP sensors or if 4MP sensors became available in time.

[spazsinbad]

'Dragon' do you refer to the WILEY article? The illustration is just that - not a specific 'device located here' F-35 exact one.

[SpudmanWP]

FF of EODAS on an F-35 was in 2010

http://www.aviationtoday.com/av/militar ... 24687.html

The company had delivered three sets of EO DAS sensors to JSF prime contractor Lockheed Martin for its mission systems integration lab in Fort Worth, for the Cooperative Avionics Test Bed (CATBird) aircraft, and for BF-4, the first F-35 slated to fly with the system. Bouchard said CATBird, a modified Boeing 737, will start flying with mission systems hardware this fall, including the AESA radar. The EO DAS "is about nine months to the right of radar," he said, putting its first flight on CATBird in latter 2009 and on F-35 in 2010.

[mrigdon]

It's possible that it's just a 1MP sensor; L3 doesn't have a datasheet for their 1MP sensor, and it's not uncommon to use (eg) 4MP arrays downsampled to achieve a more reliable or better quality 1MP image.

There are pixels and there are pixels, you can get different resolutions for the same numbers.

Most commercial digital sensors (the ones most people have experience with) in digital cameras use Bayer filters to combine four pixels on the sensor into a single color pixel in the final data file. So, you end up "losing" a fourth of your resolution. Foveon sensors use layers of light sensitive pixels, each layer receptive to one of the primary colors (Canon is supposed to be working on a similar sensor for future cameras). Those sensors have lower pixel counts, yet provide the same resolution.

The data sheet makes it sound like L3's sensor is the former, photosites with a filter applied over the sensor. If that filter doesn't require combining the same number of pixels for the final output pixel, like a Bayer filter does, then even a 1MP sensor can have much greater resolution than what you'd expect if you're used to hearing the numbers in your camera phone, for instance.

[hornetfinn]

If we assume that DAS uses 4M array and it has a field of view of 120 degrees, we can calculate the following:

1000 meter (1 km) circle has a diameter of about 6280 meters. This means each DAS sensor will see an area of slightly over 2090 meters in both width and height. With 4M array, each pixel would equate almost exactly 1 meter x 1 meter about 1000 meters away. At 10 km away that would mean each pixel would equate about 10 meter x 10 meter area.

Of course there are other variables as already mentioned regarding resolution and of course IR sensors operate differently in that a bright heat source would seem much larger to it than to naked eye. For example a pilot could not see a BVR missile launched 30 km away but DAS most likely could due to missile exhaust plume would make it rather large (and bright) in IR spectrum. Another thing besides resolution is how sensitive the sensor is and how small heat differences it can detect (thermal resolution). The 4M sensor has NETD of <30 mK at 25 degrees Celsius, which means it's highly sensitive sensor (current state of art) and can detect extremely small heat differences which means a very long detection range against all target types. This is key why DAS could see that Falcon 9 space rocket launch from 1300 km away. Most IR sensors are not sensitive enough to achieve this feat.

With this high thermal sensitivity, DAS should detect even a cool target like UAV when it basically fills one pixel. So a medium altitude UAV (like MQ-9 Reaper or equivalent) should be detectable at about 5 to 15 km away, depending on viewing angle. A fighter aircraft is more complex as it emits more heat and thus is brighter to IR sensor. It also heats the air around it (especially behind due to exhaust) and thus seems larger to IR sensor than what the physical dimensions of fighter aircraft are. DAS would see an afterburning MiG-31 flying perpendicularly very far away (several tens of kilometers for sure). F-16 or JAS Gripen coming directly towards at low power setting would be seen at much shorter ranges but still very likely more than 10 km away. I'd say that fighter aircraft are detectable about 10 to 50km away, depending on target size, power setting and viewing angle.

Missile launch detection ranges are also heavily dependent on missile size, propellant and rocket motor design and of course viewing angle. SA-10 long range missile launch will be seen very far away as the missile is big and exerts a lot of heat during boost phase. MANPAD missile will naturally be seen at much shorter ranges due to small size and low amount of heat generate. I'd say that most missile launches will be seen at considerably longer ranges than the missile could hit a target (like the F-35). Missile in flight after motor burn out will be seen due to missile heating during high speed flight. Thus the missile will be seen at much longer ranges than it will be seen with naked eye. The detection range will depend on viewing angle and missile skin temperature but I'd say we are talking about maybe 2 to 20 km range (small MANPAD to very large SAM). That'd usually give enough time (about 5 to 30 seconds) to maneuver and employ countermeasures.

These were detection range figure guesses and recognition and especially identification ranges are considerably shorter. Recognition ranges are usually about 3-5 times shorter than detection ranges and identification range is usually 5-10 times shorter than detection range. Of course all my range figures here are just guesses for performance in good conditions and exact figures depend heavily on exact design of the sensor array, signal processing and conditions (background, weather, moisture, pollution).

[uclass]

If we assume that DAS uses 4M array and it has a field of view of 120 degrees, we can calculate the following:

FoV is stated as 90deg in those links and 95deg in AirForces Monthly F-35 special.

[Dragon029]

If we assume that DAS uses 4M array and it has a field of view of 120 degrees, we can calculate the following:

FoV is stated as 90deg in those links and 95deg in AirForces Monthly F-35 special.

Those would refer to non-overlapping fields; Northrop / other sources also say that the system has a total FOV of 4pi, which equates to 120 degrees. If they were only 90 / 95 degrees, the system would have massive gaps in the sphere of coverage; here's a quick little CAD drawing to demonstrate; on the left is 6 cones, 120 deg wide, pointing in each direction, on the right are 6, 90 deg cones. At the corners, you can see right between them.

[uclass]

Hmmm, maybe, here's where I got my quote from anyway.

-AFM 2014 F-35 Special, Page 28

Also here:

The advanced features of the DAS include missile and aircraft detection, track, and
warning for the F-35. DAS also gives a pilot 360° spherical
day/night vision, with the capability of seeing through the
fl oor of the aircraft. And because the DAS is a passive
system, the pilot does not have to point a sensor in
the direction of a target to gain a track. Comprising six
infrared (IR) sensors (each housed in an aperture)
located around the aircraft, Northrop Grumman
classes the DAS as an integrated system and not a
sensor or a series of sensors.
The six apertures each provide 95° fi eld of regard and
a total of 570° to ensure suffi cient overlap in coverage
around the aircraft.
One aperture is positioned on either side of the radome
below the chine line (the right and left side apertures), one in front of canopy (upper
forward), one in front of the refuelling receptacle (upper aft) and two on the under fuselage
(the lower forward and lower aft) one pointing forward and one aft, but not straight down.
The six apertures are positioned so that no part of the aircraft blanks out its view. The
system receives threat information from all directions and stitches it together to give a
simultaneous three-dimensional spherical view, using that information to protect the aircraft.

[linkomart]

hmm, something's fishy about the 95 degree coverage, when I draw it up I get pretty big chunks of uncovered area around the Aircraft. If the aircraft shall have full 360 degree coverage the angle needs to be at least 108.752 degrees, IF the cameras are at the same point... more if not.
This is given that the sensor picture is circular, which no one states. if the sensor picture is square that can give the coverage of the corners.

Anyway, I'll leave the theories now, back to the real world...

regards.

[quicksilver]

In order to determine the notional sensor coverage, one also has to know the mounting angle of the sensor relative to the outer mold line of the aircraft at the aperture. Do we know that?

[SpudmanWP]

Since the sensor is square wouldn't the FOV be square also?

Given this, if the the angle was 95 at the edges this would mean greater than 95 at the corners which would allow for full coverage.