OLED Film Could Provide Cheap Night Vision For Cars

thecarchik writes "Night vision systems are already available in the higher-end luxury sedans from companies like Toyota, Volvo, BMW, and Mercedes-Benz. It's expensive technology that few drivers can afford, and at $4,000 for the system without a display, it's a pricey upgrade. That may all change soon, as DARPA-funded scientists have developed a cheap way to turn any infrared light into visible light with a thin film."

Comments

Night Vision Systems for Seeing in Darkness

[pitchpipe]

Night vision systems are already available in the higher-end luxury sedans from companies like Toyota, Volvo, BMW and Mercedes-Benz, but it's expensive technology that few drivers can afford, but at $4,000 for the system without a display, it's a pricey upgrade.

Sometimes I will proofread my comments, and sometimes I will read them over again just to make sure there are no mistakes.

Re:Night Vision Systems for Seeing in Darkness

[noidentity]

So tell me again what a night vision system for seeing in the darkness does? I'm thinking it has something to do with sight, and maybe night time, but I'm not sure.

Re:Night Vision Systems for Seeing in Darkness

[iamhassi]

"sometimes I will read them over again just to make sure there are no mistakes."

And sometimes I try to get my facts straight, like that Toyota and Volvo does not offer Night Vision at all (only on concepts), and Mercedes and BMW just started offering it in 2006, but Cadillac offered Night Vision 10 years ago on the 2000 DTS :
"on the DTS, and options include a better-than-average navigation system and the "Night Vision" system, which features the first civilian use of infrared thermal-imaging technology to allow the driver to better see pedestrians and animals along the road at night.... A DTS with Night Vision has been my transportation for the past week. It has been an interesting week, with more night driving than usual.... Night Vision really does work."

If the author can't bother to get their facts straight I can't be bothered to read their article. Author completely left out the Cadillac which is plain wrong considering they were one of the first to offer it and it showed up in many popular reviews. Instead, they gave all the credit to German and Asian manufactures for introducing systems either many years later or offering concept cars with night vision

Re:Night Vision Systems for Seeing in Darkness

[iamhassi]

I stand corrected: Toyota began offering in 2008 a Japanese only vehicle called the "Crown Hybrid" with optional "Night View". But good luck finding any vehicles with this system, but you can buy a 2000 DTS with Night Vision if you call Nick

Re:Night Vision Systems for Seeing in Darkness

[DriedClexler]

If I could get underpriced labor by making pension promises I wouldn't have to keep, I could have pretty innovative products too.

This little foray into cool night vision for a few ultra-rich came out of the hide of workers who had nowhere to go when the benefits they were promised got pissed away.

Fuck GM. Fuck GM long, fuck GM hard. (Don't forget to mod this troll.)

Re:Night Vision Systems for Seeing in Darkness

[TBoon]

If you bothered reading your own link to the Cadillac you'd see that it wasn't using OLED film to achieve it. It uses a camera and a projector. Different technology.

Re:Night Vision Systems for Seeing in Darkness

[JWSmythe]

    Well?

    Were there any names, or was he just pissed that people rarely show up to his meetings? :)

Re:Night Vision Systems for Seeing in Darkness

[ChrisMaple]

It's not troll, it's flamebait.

The unions forced their absurd terms on the car companies. The failure of GM and Chrysler is more the fault of vicious unions than cowardly management. The idea that it's malicious action by management is absurd.

More to the point, expensive high-tech systems like night-vision are more cost-driven by development costs and low volume constraints than by the wages and benefits of production line workers.

Re:Toyota

[TrisexualPuppy]

Indeed. I looked into purchasing a FLIR camera for my Corvette since deer are a huge problem around here. The funny thing is that it is hard to even find an aftermarket one now, and the few out there were about $8k for 320x240 resolution last time I checked. But will this tech be any cheaper than FLIR? Just throwing out the OLED buzzword doesn't mean much.

And any links to more substantial coverage on this project? The article is just a blurb.

What night-vision goggles could look like

[Conspiracy_Of_Doves]

What night-vision goggles could end up looking like

http://www.fixandweld.com/images/SSP11.jpg

Re:What night-vision goggles could look like

[The Penguin Man]

What night-vision goggles could end up looking like

http://www.fixandweld.com/images/SSP11.jpg

Why not just coat your spectacles with this thin film of OLEDs?

Re:What night-vision goggles could look like

[Conspiracy_Of_Doves]

Because with that, they cover your entire field of vision.

Besides, not everyone wears glasses.

Re:What night-vision goggles could look like

[The Penguin Man]

Yeah, I get your point. But I think it would be very useful to have glasses coated in this stuff.. That would be super neat!

Re:What night-vision goggles could look like

[Pinky's Brain]

Despite what the article pretends, it fundamentally can't work that way.

In essence this film is a a CCD with a LED display the back ... like a CCD you still need a lens in front to get an image, and the image is only 2D.

Green or BW?

[notjustchalk]

Is there a typo in TFA here?
"The result is a monochromatic green-tinted view of the area in front of it."

But the image caption under the green-tinted rendering is "Conventional night vision image, via David Kitson, Creative Commons 3.0". Plus, it looks pretty grainy/low-res compared to the newer BW rendering which is captioned "New 7-series to feature high-tech night vision system".

Regardless, if the IR image is overlaid onto the screen (HUP) display, mightn't oncoming traffic lights be blinding and actually restrict vision at times?

Re:Green or BW?

[omnichad]

I doubt it's a typo. Neither picture uses the OLED technology. The 7 series has a high-tech night vision system, but it's not OLED. If you read on...

So far they've only managed to make a tiny 1-square-centimeter example, but due to the OLED technology underpinning it, it should be scalable to much larger sizes--and the scientists behind the discovery think they can do it within 18 months.

Re:Green or BW?

[canajin56]

As others have mentioned, the second image isn't of the new tech, which isn't scaled up enough for a prototype yet. The top image is from the 7-series, as you've already noted. The reason it's so crisp is that it uses Active Illumination Infrared. This works by having headlights that are so bright, they would be illegal if they were emitting any visible light. As it happens, they are absurdly bright, but only in the near-infrared band (NIR). Then, an infrared CCD device records the NIR levels, and converts it to a monochrome image. I'd thought they all worked by a CRT projector that projects the image onto the window, which has a special and expensive film that is only reflective to a specific wavelength of green light, allowing it to display the green image overlay without any (noticeable) change to what you see during the day. I didn't think the same trick would work with a white projection, since you'd need to be reflecting at least red, green, and blue light, so you'd need at least 3 layers of film, making it 3 times as expensive, at least! And, I would think that while looking "better" it would be counter-productive. There's a reason green is the most commonly used nightvision color. It's not because that's what works, but because green light doesn't cause your eyes to adjust. So, however bright the green light is, when its turned off, your eyes are still adjusted to the dark. Same thing happens with red light, but red night vision was found to be more unsettling. (This is, by the way, why blue LED clock radios are obnoxious.) So, I find it odd that they'd use white light for nightvision, since it would wreck your night vision. On the other hand, maybe that's good: If your eyes are adjusted for fairly bright light already, then oncoming headlights won't be as blinding as they are without any kind of night vision, so it might be superior all around. Except, when you look in your side mirrors or windows, you won't be able to see shit all, I wouldn't think.

Anyways, Active Illumination is nice and crisp because everything is very well illuminated, so the signal:noise ratio is quite high. The second shot there is conventional night vision, which doesn't use infrared at all, but rather, uses light amplification. In starlight situations, the signal:noise ratio is tiny, and so you are amplifying a lot of noise along with the signal, so you get a very grainy image. It's not film/CCD grain, just background noise. It has nothing to do with the color of the rendering, as you seem to be implying. It has to do with the technique you're using to see in the dark. Active illumination is much superior to light amplification because of its crispness alone. But, it has the added benefit of not being so sensitive to bright lights overloading it. If you have light amplification going, and a headlight shines in the camera, that will be amplified to super bright. Not good! On the other hand, headlights do emit lots of infrared...but not THAT much. With passive infrared, your image would be washed out by headlights, since they'll be by far one of the brightest NIR sources. But with active infrared, everything is being bathed in NIR, so headlights won't brighter than the car they're in (not by a lot, anyway), which will reflect a lot of your lamps back. If the oncoming car also has NIR lamps, they'll be brighter than any reflections, but again, not by a whole lot still. Optimally, your display would "clip" any really bright sources, rather than just darkening the rest of the image. Ugly, but you're driving, not taking pictures. You care about seeing the darker bits, not about maintaining relative light levels. The white light would still shine through your glass like normal...but that's because that's how glass works, and it has nothing at all to do with the nightvision. And, as I said, best case scenario is that it's not as bad because your eyes would be adjusted to the night vision display, which is brighter than what a normal drivers are looking at when oncoming traffic rounds the bend and s

Re:Green or BW?

[Antidamage]

If you have an OLED overlay, you also have a dynamic brightness blocker that can tint itself to dim blinding objects.

What they don't mention is how they'll introduce focus to the system. I suspect we'll see see another filter layer of millions of tiny tubes or tiny lenses.

Re:Green or BW?

[Anonymous Coward]

I suspect we'll see see another filter layer of millions of tiny tubes or tiny lenses.

So you're saying that the technology requires a nano-Internet?

Re:Green or BW?

[GrpA]

Actually, most car headlights throw out phenomenal amounts of near-IR energy - usually far more than is visible ( unless they are LED headlights ) - I built an IR driving system into my Starion over a decade ago as a project.

However, what is important is how sensitive the receiver/camera/tube/film etc is to the IR light and at what frequency.

Also, the green image in the picture is grainy because:

a) It was taken through an ESI Inverter tube, such as was invented around 1970... Sensitivity 270 uA/lm.
b) Most of the image is clipped.
c) It was very dark when the image was taken ( ambient urban starlight and reflected streetlights, with the image mostly in shadow )
d) The S/N ratio of that tube is 3.7:1
e) The resolution of that tube is arund 32 lp/mm.

Compare that to modern night vision which is,
a) Gen 3 wafer tube with GaAs photocathode sensitivity > 2000 uA/lm
b) Provides resolution higher than HDTV in an unclipped circular region.
c) Significantly more sensitive to starlight than the P25 photocathode in the grainy image.
d) The S/N ratio is around 28:1
e) Resolution would be above 64 lp/mm..

It would have been better if they used a more recent image, but even as Gen2, there was still a lot of detail that could be seen in the photograph.

GrpA

Not only Night Vision

[your_neighbor]

If the OLED is already there, you can use it to display all kind of virtual stuff:
GPS
Ads
Traffic Info
Hookers
Create beautyfull landscapes in polluted areas

Re:Not only Night Vision

[FlyingBishop]

It sounds like this is not an actual display, just a film that translates infrared light into visible light when it is supplied with power.

Re:Not only Night Vision

[bigspring]

Don't forget porn!

Re:Not only Night Vision

[sconeu]

Wouldn't you be blinded by the hood?

Re:Not only Night Vision

[nschubach]

Please, just because the buildings may be out of repair and the sidewalks covered in graffiti doesn't mean it's blinding. /badpun

It is a good question though. Is infrared generated by the heat of an engine or are you referring to the reflected light from other sources? If reflected, I would imagine it would be no worse than a sunny day.

Re:Not only Night Vision

[YourExperiment]

Traffic Info
Hookers
Create beautyfull landscapes in polluted areas

Spell-check Slashdot postings composed while driving

Re:Not only Night Vision

[sconeu]

I was referring to engine heat.

And that was a fun pun, by the way.

Re:Not only Night Vision

[fluffy99]

If it's just pixles glowing in the presence of incident IR, it would be worthless. You lose the whole concept of what direction the IR came from and end up with the equivalent of major glare on a smeary windsheild.

Re:Not only Night Vision

[CAIMLAS]

I suspect that the actual infrared sensors would be mounted in front of the vehicle and telemetric computation would be performed to put the thermal image up in a representative fashion.

Of course, it would still be almost useless during warmer months, particularly in more developed areas. The ambient temperature of a road can remain well over 100F throughout the night, and the air a similar temperature, even if it drops into the 70s. I have no doubt it's worth in the South. Any critters on the road would be 'invisible' to this tech, resulting in people either not using it or trusting it implicitly (and resulting in additional accidents/vehicle damage that would not occur with people simply paying attention).

How?

[brian0918]

How could this be used for cars, unless everyone drives around with their headlights off at night? Besides the headlights, the heat from the cars' engines would also produce a very bright result in IR, no?

Re:How?

[ShadowRangerRIT]

It's converting infrared to visible light, it isn't a light amplification system (where headlights would cause blinding glare). If people are running headlights, then the front end of their car will be slightly warmer, that's all. You want the heat from the cars themselves to show up; enough morons forget to turn on their lights that being able to see them regardless of lighting is a good thing. TFA is light on details, but I suspect it would overlay infrared hot spots on the regular light being let through. Applying a logarithmic progression to the conversion would prevent *really* hot things from blinding you. So anything well lit will be visible, anything hot will be visible (in the infrared overlay), and anything both well lit and hot will look funky. I'm sure there's a joke in "anything both well lit and hot will look funky," but I'll leave that up to another /.-er.

Re:How?

[theheadlessrabbit]

How could this be used for cars, unless everyone drives around with their headlights off at night? Besides the headlights, the heat from the cars' engines would also produce a very bright result in IR, no?

OLEDs, like all display technologies, have a limited range from brightest light to darkest dark. It wont simply "amplify all light by 100X" - it can't. at some point, highlights will get clipped. It's just like lens flare on a camera. Looking at a picture of the sun won't blind you, even though the real thing will.
Oncoming headlights might wash out the screen temporarily, but real headlights can do that, too.

Re:How?

[Ephemeriis]

How could this be used for cars, unless everyone drives around with their headlights off at night?

Depending on the time of day, weather, etc. headlights aren't as useful as they could be. And even under the best of circumstances there's a limit to just how much is illuminated by your headlights. Anything that can increase your visibility is a good thing.

Besides the headlights, the heat from the cars' engines would also produce a very bright result in IR, no?

You've got to put the camera somewhere. It would probably be possible to place the camera far enough to the front of the car that you aren't seeing the hat coming off of the hood.

Or you could maybe do some digital processing to block out that heat source

Re:How?

[CorporateSuit]

The spectrum for Infrared is large -- larger than our visible spectrum. Depending on how broad a spectrum these oled's accept as input, you should easily be able to register sources like channel changers and DS communications without dipping into heat signatures.

Re:How?

[canajin56]

It's not light amplification, headlights won't overload it and blind it/you. As TFS says, high end cars already have expensive night vision systems, where a film on the windshield has a projector overlay the screen with a shot from the IR camera. To show inanimate objects better, there is also a set of IR lights to illuminate, in addition to normal headlights. And yeah, engines and pedestrians glow pretty bright on these displays. That's the damn point. It's not blinding, because the total amount of infrared just isn't that big.

Re:How?

[blincoln]

The authors of TFA (and the source article they based it on) confuse the issue by assuming that all infrared is the same thing. Based on the "green glow" look and the "830nm illumination" mentioned in another post, this system is working with near-infrared light (like conventional night vision systems), which is completely different from thermal infrared systems.
This system will not function as a thermal imager. If something is very hot (just below the point of glowing incandescently in visible light), it will register on a near-infrared system, but it's not a cheap FLIR replacement (sadly).

Re:How?

[sconeu]

It would probably be possible to place the camera far enough to the front of the car that you aren't seeing the hat coming off of the hood.

But if I left my hat on the hood, I'd want to see it go flying off, so I would know where to pick it up!

Re:How?

[mr_mischief]

Cars produce IR, including halogen or incandescent headlights.

Liking it, but...

[Uniquitous]

Hopefully they'll also engineer it to compensate for jerks who drive with their brights on. I find on-coming headlights dazzling enough already without amplification, thank you.

Re:Liking it, but...

[bugs2squash]

No - instead some idiot will think that it's fine to put ultrabright IR lights spraying light all over the road, because no-one can see it, right ! This system will be ruined by its own widespread adoption.

Re:Liking it, but...

[djdanlib]

Don't the regular bulbs spray a lot of light in the near IR field anyway? Are these sensors operating on a band of IR that regular headlights aren't pouring out? Yeah, I know, some idiot is going to get lights that dump on the same band the sensors use and it will nullify the usefulness of said sensor system in some situations, I get it.

Better get a mop, all that spraying and pouring and dumping sounds messy...

Re:Liking it, but...

[canajin56]

Nothing is being amplified. This isn't like nightvision goggles, it's Active Illumination Infrared: You can only non-hot things because there are powerful infrared lamps bathing the scene in (invisible to everything but the camera) light.

Re:Liking it, but...

[canajin56]

Normal headlights are fairly focused. The infrared lamps used for Active Illumination nightvision are very widebeam. So, I don't think their lamps will result in localized bright spots, but rather, just everything being brighter. I'd hope that they thought of other people having the system, and as such, designed it to adjust the total brightness of the display to match the overall brightness of the image.

Re:Liking it, but...

[svtdragon]

I think you a word.

Re:Liking it, but...

[steelfood]

I think GP was saying it could replace the high beam. The infrared imaging system can be switched off when the infrared high beam is off as well, so that when you're driving towards me with your infrared brights, I'll still only see your normal headlights so long as my infrared brights are off.

And because in suburban and rural areas, you can turn these up without bothering people inside a building, you can have them project farther, to the point where you can spot that deer headed straight for the road from much farther away.

Affordable for civilians to spot undead.

[qwerty8ytrewq]

This will make spotting vamps and zombies in traffic easy, and people with SARS, fever etc... just like the body heat scanners in some Asian airports.

Glasses

[Charan]

Forget cars, how long will it be before I can a pair of eyeglasses with this coating? The night vision coating could go on one side of the lenses, and the anti-glare coating goes on the other side. What more could a geek need?

Re:Glasses

[fifedrum]

x-ray specs! that's what more.

Journal Article

[JustinOpinion]

For those with access, the media report is based on this paper:
Organic Infrared Upconversion Device, Do Young Kim, Dong Woo Song, Neetu Chopra, Pieter De Somer, Franky So. Advanced Materials 2010, 22, p.1-4. DOI 10.1002/adma.200903312

The abstract is:

Novel infrared-to-visible light upconversion devices are demonstrated by fabricating an organic light-emitting diode with an infrared-sensitizing layer. With a SnPc:C60 mixed layer as an infrared absorber and fac-tris(2-phenylpyridinato) iridium (III) (Irppy3) as an emitter, an infrared-to-green up-conversion device is demonstrated under 830-nm irradiation (see figure, ITO=indium tin oxide). The maximum photon-to-photon conversion efficiency is 2.7% at 15V.

This is good development, to be sure... but I think TFA exaggerates by saying that the device can be so thin that it can be placed on a windshield. In order to be used for something like night-vision, you'll need some kind of lens/optics as well. This material will not maintain the directionality of light as it is converted (from IR to visible), so you can't just "look through" it and see a night-vision version of the world. But you could use a lens to focus an infrared image onto the film, and look at the visible-light emission from the film. Still, this technology should be able to help make night-vision systems smaller and cheaper.

It's also disappointing how media reports of new sci/tech developments insist on focusing on one possible application. It obscures the real potential. For instance, lighter/cheaper IR-to-vis conversion would not just be cool for night driving, but also for emergency workers, home security systems, scientific instruments (the journal article also lists "semiconductor wafer inspection"), optical computing, and so on...

Re:Journal Article

[Scrameustache]

This is good development, to be sure... but I think TFA exaggerates by saying that the device can be so thin that it can be placed on a windshield.

Yeah, I don't know what's up with the car angle. I see what they did and I think "night vision goggles that are just goggles! Finally!" (as opposed to the bulky getup currently available).

Re:Journal Article

[CrazyColin]

Yeah, the fact that they didn't address the directionality of light makes me assume they didn't solve that problem. So you can't just slap the film on a windshield and have night vision, you need to project an IR image on the windshield (in which case the image would only be relevant for the driver.) A lot less exciting than they make it sound.

Re:Journal Article

[Pinky's Brain]

I don't see what's the point really ... the optics are the bulkiest parts. Instead of focusing an image on a tiny CCD it would have to cover the entire display. A normal camera plus LCD display or HUD projection on the windscreen make a lot more sense.

Re:Journal Article

[KeithJM]

But it would work just like regular light, right? When regular light hits the windshield, with our binocular vision we can judge distance, etc. If IR was emanating from an object and a film on the windshield turned it into visible light, the whole binocular vision thing should still work, right? This isn't my field, so I may be missing something.

Re:Journal Article

[JustinOpinion]

The problem is that the IR-to-visible conversion doesn't maintain directionality of the light. Normally a visible light photon hits the windshield, travels through it, and comes out the other end going in the same direction (well, slightly shifted due to refraction). But with this upconversion film, an IR photon would hit the film, and be converted into a visible-light photon... but that visible light photon will now be traveling in a totally different (effectively random) direction.

Since each "pixel" on your windshield would be hit by IR light coming from a variety of directions, you would in fact just see a total blur on the windshield.

Obviously it would be awesome to come up with a material that could convert the frequency of light without altering direction. In practice this is not easy to do. (Any kind of conversion is going to require a photon to be absorbed by the material, promoting to some high-energy state, and then the material will re-emit a photon of a different frequency. The problem is that the re-emission is basically isotropic.)

Re:Journal Article

[CrazyColin]

See the reply "No, this won't work." It's akin to a frosted piece of glass or translucent plastic. Light enters the film but directionality is not preserved because OLEDs cannot choose where the light goes, they just emit in all directions at once. Optics is not an easy thing to explain or understand fully, which is why so few people picked up on why this doesn't work. I'll try though. Transparency in optics means that light can get through something without being scattered. That is, emitted photons have non-random directionality based on the incident photons. It's frosting of glass that introduces this random directionality, since all incoming photons are sent out of the glass in random directions. That's the problem with this film. Incoming photons are turned into electrical signals by the IR sensors, then those electrical signals tell the OLEDs to emit photons. The emitted photons are emitted randomly, so the effect is scattering. And scattering in a windshield is a very bad thing if you want to be able to make out objects on the other side.

Re:Journal Article

[Pinky's Brain]

There is actually a very simple way to capture/replay wavefronts for which this technology is ideally suited, integral imaging.

You simply put an exactly aligned integral lens array in front and at the back and you're done. You need very high resolution of course, but because the pixels are so simple and completely independent that's not a problem. You would probably want to use diffractive instead of refractive lenses so you can get a decent viewing angle, but again that is no problem here because you are only using a single wavelength on each side.

If you wanted to do this for a windscreen though you would have to interleave the IR integral lenses with clear window. So say 1/2 mm square IR conversion pels (consisting of an IR filter, an integral lenslet, the conversion film with something like 1 um pixels and then another integral lenslet) interleaved with a normal clear window. At an effective 25 DPI the resolution wouldn't be great, and normal viewing would have a bit of screendoor effect though. Smaller and or less IR pels per clear window would help, but hard to say what's appropriate without experimenting.

Should work ... almost certainly a patented concept though. If not ... it probably should be patented :)

Re:Journal Article

[Pinky's Brain]

On second thought ... the image could look great ... from outside. From the inside the direction of travel of the light is mirrored around the center of the integral lens. I guess you need one more lens to flip it back.

Re:Journal Article

[svtdragon]

This is slashdot. Our effort is minimized when the car analogy is already in the summary.

Re:Journal Article

[blincoln]

For instance, lighter/cheaper IR-to-vis conversion would not just be cool for night driving, but also for emergency workers, home security systems, scientific instruments (the journal article also lists "semiconductor wafer inspection"), optical computing, and so on...

Essentially any digital camera sensor (including webcams) can pick up the type of infrared light that this device does (~1000nm). I'm a little confused about how it could get any cheaper. If this device could also pick up mid- and far-infrared, now that would be a breakthrough, because sensors that can handle those wavelengths are still thousands of (US) dollars.

Re:Journal Article

[bar-agent]

It's akin to a frosted piece of glass or translucent plastic. Light enters the film but directionality is not preserved because OLEDs cannot choose where the light goes, they just emit in all directions at once.

It doesn't seem like this would be much of an issue. I can think of two solutions. One solution is to separate the OLED elements so they form a halftone screen. Each OLED element would be a point source of light so the directionality of its output wouldn't matter so much. Another solution is to put an isolating barrier around each OLED element so that it only shoots light out one direction because light output in other directions gets absorbed. If you use a material that reflects or absorbs differently based on incident angle, it shouldn't affect the light from the outside that much; it would only absorb light from the OLED that is more-or-less parallel to the film itself.

Re:Journal Article

[bar-agent]

...a third solution is kind of the inverse of the second. If you are worried about lateral light from the OLED making everything a blur, just coat the film in material that doesn't let lateral light out, but instead directs it out of the film's edge using internal reflection.

Re:Journal Article

[Pinky's Brain]

The light doesn't just have to be directional ... you have to do something similar to holography. The angular spectrum of the outgoing light has to be the same as the incoming light at a given point on the window (although it can be quantized a bit, like with integral imaging).

Re:Journal Article

[CrazyColin]

As I said, this stuff is hard to explain (and I seem to have failed).

I overemphasized the importance of the OLED's ability to emit light in a specific direction. As you know, OLEDs emitting light in all directions are perfectly fine for TVs and such, so they are absolutely capable of creating an image. And yes, there exists technology to make that image only viewable under a small range of viewing angles. This technology is probably not advanced enough to restrict the viewing to only the driver of the car, but even if it was, the layer of IR sensors is the real issue.

Solid-state IR sensors such as these do not know where the incident light is coming from. They only measure the amount of incident light. If, say, a light is moved around the sensor but kept at the same distance, the amount of incident light will remain the same and the sensor will not know the light is moving. It's this loss of information that makes the whole IR-film-overlaying-an-OLED-film-for-night-vision impossible. If you're sitting in the car, whether or not the OLEDs can emit light directionally, and There Are 4 Lights on the other side of the window, you will not be able to tell whether there are 4 lights or there is one light that's simply 4x brighter. This is because the IR sensors do not even know that, so they cannot convey that information to the OLEDs, which convey the information to you.

The bottom line is that when you want to capture an image (say, of the area in front of your car), you do not use a big array of IR sensors with no lenses. You put an array of sensors behind the focal point of a lens, where the light is focused. i.e. you use a camera.

I hope that helps. In order to display an image, you have to fist capture an image to display. And you can't capture an image with an unadorned array of sensors unless the thing you're imaging is sidled right up next to the array.

Re:Journal Article

[bar-agent]

I hope that helps. In order to display an image, you have to fist capture an image to display. And you can't capture an image with an unadorned array of sensors unless the thing you're imaging is sidled right up next to the array.

Ah. I was thinking in terms of a HUD or an augmented reality display, not redisplaying an infrared image in visible wavelengths. I assumed there would be a scanning IR beam to light up the OLED elements.

From the article, it sounds like each OLED element includes its own IR sensor, or else converts the IR directly to visible light chemically. If they do it that way, the 4 lights situation (nice ref, by the way) shouldn't be a problem. The OLED elements closest to each of the 4 lights would be emit more light than elements more tangential to each light, because of higher IR intensity at that element as compared to other elements where the IR hits more incidentally and thus with a weaker flux density (which I think that is the correct term).

Re:Journal Article

[bar-agent]

The light doesn't just have to be directional ... you have to do something similar to holography. The angular spectrum of the outgoing light has to be the same as the incoming light at a given point on the window

What's an "angular spectrum?"

Re:Journal Article

[Pinky's Brain]

http://en.wikipedia.org/wiki/Angular_spectrum_method

It's a way to describe a propagating wavefront, probably not a good choice of words.

What I meant to say was that for any given point on the window the intensity of light coming from any given direction has to be output at the same intensity and in the same direction on the other side. For all positions and directions at the same time ... you can quantize the spatial and angular resolutions a bit, but simply using a directional LED doesn't cut it.

As I said further down it could in theory work with a variation on integral imaging though.

Re:Journal Article

[Pinky's Brain]

The incoming light from a light source at 50 meter at different points on the window is going to vary in intensity by less than 1%.

Re:Journal Article

[CrazyColin]

Oh, yeah, with the scanning beam it'd work fine. The image would still have to be captured from the perspective of the driver somehow and it wouldn't be relevant to anyone but the driver, but it would work. Since it didn't mention that I assumed the film itself was supposed to be a standalone night vision apparatus.

As far as the second paragraph, what you're talking about definitely works, but only when the lights are up close. In this case, the light right on top of an element will illuminate that element a whole lot more than the other lights because it's many times closer and the incident light is less tangential. When the lights are far away, say a couple hundred yards, the difference in angle of incident light from each light on an OLED element will be a tiny, tiny fraction of a degree; essentially zero. The difference in distance is also essentially zero. So all four lights are illuminating the element approximately equally. (I get what you mean with weaker flux density, but just saying less flux would be more accurate :) Light intensity can be described as a 2D density (e.g. W/m^2), and flux is multiplying that by an area to get a power.)

This goes back again to the frosted glass analogue. With a scanning visible light beam (akin to the scanning IR beam you suggested), you can project an image on it. Normally though, you can't see through it because of scattering. When lights are up close, you can differentiate them, but as they get farther away, they're harder to make out.

sunglasses

[Anonymous Coward]

Screw night vision windscreens. I want some night vision sunglasses!

No, this won't work

[Thagg]

If they are OLEDs, they emit light relatively isotropically. So infrared light hitting the film would be turned into visible light, but it would be completely diffused -- like if you had a piece of translucent plastic. The only way it could work if the light was transformed in frequency but kept the same direction.

Re:No, this won't work

[Kratisto]

Oh great. Next you're going to tell me that there won't be any jetpacks, flying cars, or invisibility cloaks.

A GPS upgrade would be really nice

[NotSoHeavyD3]

I was talking to a co-worker about this a couple years ago. What I was hoping would come out is a GPS device that would display on your windshield. When you are driving down the road that path would actually "light up" while you're driving on it. When you come to an intersection and need to know which particular street you need to go down that one would light up. Of course it'd have to be smart enough to let cars and obstacles show through.

Re:A GPS upgrade would be really nice

[Rallion]

In order to really work well, something like that would also have to have head tracking in order to draw in the right place, and ideally the positioning system and the map itself would be extremely accurate (more accurate than they currently are). Not at all impossible, but I'm not sure it would work well enough to be worth doing quite yet.

Re:A GPS upgrade would be really nice

[robot256]

Sure, you could have it show "right turn" and "left turn" arrows and distance remaining in a heads-up display, but augmented-reality overlays are way too dangerous with today's inaccurate map and position data--what if your "yellow brick road" goes through a ditch or a stopped truck?

Re:A GPS upgrade would be really nice

[nschubach]

With the current upcoming game console wars is the technology that does just that. It's actually been around for some time and only needs a camera facing the driver that can extrapolate facial features in a very easy "low tech" way of video filtering.

Re:A GPS upgrade would be really nice

[nschubach]

I would hope it was a transparent arrow and didn't cover the entire screen...

FAIL!

[srussia]

The money quote is: That may all change soon, as Defense Advanced Research Projects Agency (DARPA) funded scientists at the University of Florida have developed a cheap way to turn any infrared light into visible light with a thin film.

Re:FAIL!

[robot256]

Those scientists are such cheap bastards! Let's see them try that on an infrared photon in a FAIR fight!

Re:FAIL!

[CAIMLAS]

I don't care how cheap it is; it's still going to make vehicle maintenance expensive.

If you want a device like this, chances are you're either a) out and about on back roads and want animal detection and/or b) spend a lot of time on the highway/interstate/lonely stretches of road where hitting animals or having a HUD would be useful - like a roadtrip. These are scenarios where you're more likely to get rocks thrown at your windshield, causing a crack.

A windshield repair costs, what... $150? And a new one costs several times that. Replacing a windshield with a built-in display is bound to cost more than a decent small, used car.

By any standards, that's not "cheap" - but it is "cheaper". I really hate the word "cheap" because it's come to be almost meaningless: it's got half a dozen meanings and many of them are almost contradictory.

Re:Toyota

[Anonymous Coward]

If you KNOW deer are a problem, then drive more slowly at night during the times of year
when the deer are most mobile. I live a an area with a huge deer population, and this strategy
has saved numerous collisions. What I am talking about is driving the speed limit or slightly
below the speed limit. And BTW, a FLIR system won't help when the deer jumps
out of thick brush directly into the path of your car. Such thick brush would prevent the deer
from being seen even in daylight.

Some problems don't require tech to solve.

Some problems just require common sense.

Re:Toyota

[Amouth]

based on the exceptionally short article - and a random guess ..

i wonder if this isn't done by energizing the OLED array so that when IR hits them they energies (taking from their vacuum tube comment i see this acting like the phosphorus layer on a CRT)

if that is the case.. this is nothing more than a way of receiving IR on one side of the film and illuminating on the other - so more of an IR screen instead of an object detection and alert system

if that is true this would be useful and far cheaper - but given the expectational short story.. i doubt any of us will remember it when the tech hits the market.

also would mean you could blind passing drivers with an IR transmitter.. something only drivers with this screen would see/notice..

Re:Toyota

[Nadaka]

Also, on sufficiently low traffic roads, it can be safer to drive down the middle at night as the odds of a deer, horse or cow jumping in your path at the edge of the road is vastly higher than someone coming the opposite direction without headlights.

what about cheap boobies ?

[dario_moreno]

As someone who has demonstrated a FLIR in a science fair, I can tell you that far infrared is quite an interesting wavelength to explore body shapes...Too bad that even the cheapest 160X200 model is at least $2K (a drop of a factor 10 in 5 years, but still). It allows to pinpoint the location of certain points quite accurately, the more so if the person imaged starts to blush. Sweaty areas come out as black as much colder than the surroundings. So basically the question is : at which wavelength does this magic film operate ? Does it correspond to black body (Planck was a perv) radiation at ambient temperature ?

Re:Toyota

[JWSmythe]

I've looked at this for a variety of applications. I'm currently building out an RV myself, and wanted to have all the bells and whistles that you can't get anywhere else. I wanted three cameras facing forward. FLIR, light intensifying, and normal color/IR illuminated.

      FLIR is simply way too expensive. I wanted it for the advantage of being able to see well beyond the illumination of the headlights for possible problems ahead. It's one thing to stop or swerve in your 3,300 pound Corvette. It's another thing to stop or swerve in my 35,000 pound RV.

      Until the prices comes down, normal color/IR illuminated with IR flood lights will be used to help visibility beyond the headlights.

    Down at the bottom of the story, they mentioned that they only had a 1cm square sample that worked. It's an awful long way from a HUD type display or full windshield thermal imaging.

Re:Toyota

[JWSmythe]

When driving remote interstate roads in the middle of the night (like when even the truckers have stopped for the night), I always sit in the left lane. Well, unless a car comes up behind me, which is very rare at those hours. It's been to my advantage quite a few times. I've seen animals wander out onto the road, and I've had more time to react because they've been more visible.

    A friend of mine wasn't so lucky on a 2 lane country road years ago. A deer jumped out and landed perfectly to intersect with the nose of his car. All the body metal in front of the doors had to be changed, as well as the windshield (damaged by the hood shoving back into it).

    I'd prefer to hug the center line than to actually be across it. That's all you need is a bored LEO sitting on the side of the road to see that and assume you're an easy DUI catch. At very least, you'll get ticketed. Or worse would be a driver turns out because it looks like you're in your lane, and suddenly it's a head-on accident where you're at fault.

Re:Toyota

[GrpA]

FLIR won't be that useful... It's a nice idea limited by the practical application and usefulness of the technology. It can actually be unsafer to have, given the problems with background noise.

On the other hand, IITs (Image Intensifier Tubes) provide a much more useful view and you can get IIT's fused with LWIR (Long Wave Infra Red) if you really want and you can turn your headlights off...

Why? Because it's the headlights that actually draw the animals into the road in the first place... Because they get scared and the only place they can SEE that's available to run to just happens to be directly in front of your car...

But without headlights, they will likely stay off the road and with decent passive night vision, you can see them quite some distance off - Add some extra IR if you like and better still, modulate it and use a gated IIT so you don't get blinding feedback from near reflections.

None of this technology is new though.. All been around since WW2 or thereabouts.

But even if it's safer, I'd love to see someone try to explain it to a LEO why they were driving at night with the headlights off...

IR conversion into a HUD, however, IS effective. Especially if you can get visibility of SWIR (Short wave infra-red ) up around the 1500nm region.

Though sensitivity is something that's missing from all of the articles I could find and I didn't find mention anywhere confirming that this was related to LWIR as most of the articles seem to suggest.

GrpA

Re:Toyota

[wagnerrp]

Certainly there would be governors put on such a thing to prevent too high a light output from physically blinding the driver, but you would render the NV film useless for as long as you were pointing at it.

Kind of like these

[SteeldrivingJon]

People who work with non-visible IR or UV lasers can use special tools that fluoresce visibly when contacted by a laser of the appropriate wavelength.

There are specially coated cardboard cards, as well as small disks that can be placed in lens holders on an optical breadboard.

http://thorlabs.com/NewGroupPage9.cfm?ObjectGroup_ID=296

Perfect

[ThatsNotPudding]

Just don't ever turn on the defroster.

Cheap or inexpensive?

[AP31R0N]

There is a difference.

Re:Toyota

[Amouth]

that is kinda the point.. if it is supposed to be a supplemental overlay on the windshield for night driving.. having it white/green out would be a serious hazard - blocking view and disrupting any sense of night vision they would have.

Re:Toyota

[nametaken]

Agreed, I'm sure they'll prevent wash-outs. Besides... I can could blind you with a sufficiently bright flashlight TONIGHT... and yet there doesn't seem to be a huge problem with people running around blinding drivers at night.

Re:Toyota

[Fred_A]

But even if it's safer, I'd love to see someone try to explain it to a LEO why they were driving at night with the headlights off...

But they can't see you from Low Earth Orbit if your headlights are off !

Re:Toyota

[mr_mischief]

Forget the cops. Try explaining it to the oncoming car that didn't see you because you were driving dark. Headlights are a safety device in more than one way.