Slashdot Mirror
Researchers Design a Self-Powered Digital Camera
Jason Koebler writes: Researchers at Columbia University have designed a fully electric digital camera that powers itself using ambient light. Put in a well-lit room, it would work indefinitely. The camera's image sensor does double duty. It measures the light needed to make the photograph, and it also takes excess light and uses it to power a capacitor (it has no battery) that runs the camera (PDF). The research team says the technology can be used to create self-powered cameras that can live on the internet of things.
But what about low light performance?
How can I believe you when you tell me what I don't want to hear?
Trying to make it work off just light, is sheer gimmickry. You run just run power over ethernet in a wired environment, in a wireless setting running the transciever will really hurt your capture rate especially if there is variation in illumination.
I didn't know Columbia University had such a strong community of voyeurism fetishists in the engineering department. Otherwise, why would anyone want low-quality, discrete, remotely accessible and insecure (IoT) cameras that don't require maintenance? Everyone already has smartphone cameras that are readily available and can be connected.
I recently visited an old friend, and saw that he'd installed a PV grid on his roof -- three rows, with thirteen panels in each row, and separate control/monitor circuitry for each one. It took me about two minutes to say "you know, put up a big board with a hole in the middle, and you could do imaging with that array."
It was a dumb joke, not a profound engineering insight.
Yes, I'm sure this camera can self-power. No, there's no way to make it cheaper or more effective than putting bigger, dumb panels on the outside of the camera, outside the lens, where they'll collect light over a much larger solid angle. Heck, you'd probably do better by putting a semitransparent solar cell in front of the lens, and a conventional sensor behind.
I looked at the photo on TFA and that thing is HUGE. Seriously, it's got to be a foot square. :-)
It must have been something you assimilated. . . .
It looks like an array of solar cells on which you focus light. So the voltage level at each cell both powers the camera and determines the exposure for that pixel.
Kind of interesting, kind of pointless at such low resolutions (40x30).
The Daddy casts sleep on the Baby. The Baby resists!
A solar panel the size of a CCD sensor would probably suck. This way, at least it has a lens in front of it :-)
CLI paste? paste.pr0.tips!
In figure 3 Q1 is drawn as a FET (and the circuit implies it is one) but they refer to the "emitter" of it when speaking of the drain.
And obviously the goal of high resolution is counter to needing large cells to capture charge for harvesting.
The design would seem to imply that the device cannot be self-starting. That is, if it runs out of charge, it has no way to activate the harvesting and get it self running again. Ah, I see in there it say they had to start with a charged supercap.
It's still an interesting experiment.
http://lkml.org/lkml/2005/8/20/95
Basically it is just a really low resolution camera made out of discrete photodiodes. It's pretty cool that you can put something like this together quite easily now (though that is a lot of soldering) and a fun project. As for practical applications, well, that's just not really going to work. The active light collection area of the camera is the aperture size, which you have to make really small if you want any sensible depth of field, so the amount of useful energy you can collect is tiny and when you factor in colour filters it just gets silly tiny. You could have a dot sized solar cell next to the lens and it would generate more power than any practical system for much less effort.
to create self-powered cameras that can live on the internet of things.
Anyone who uses the term internet of things (IoC) when talking about a product should be shot on sight. Things DO NOT need to be connected to the internet.
If we can't secure the basic things already connected, important things such as power plants, traffic signals or government computers, wtf do you think will happen when crap like this is thrown in the mix?
We will bankrupt ourselves in the vain search for absolute security. -- Dwight D. Eisenhower
This would be perfect for powering an eye implant like the one recently demonstrated in the news.
Are they spray-able? Will they stick to a General Products hull?
The word "selfie" has been pre empted in the digital camera world to mean a different thing.
sed -e 's/Chuck Norris/Rajnikant/g' joke > fact
welcome to the internet of things
You had me until you said "internet of things."
https://www.eff.org/https-everywhere
It's not powered by itself. Of course, this can be accomplished with most current digital cameras by adding solar cells. .with much better results.
-- these are only opinions and they might not be mine.
Could be usable in wearables where power is at a premium.
Custom electronics and digital signage for your business: www.evcircuits.com
I think the point is you just stick this thing on a pole somewhere, tell it when to trigger and collect the photos months later. It could be a whole lot of separate parts, like a computer can have a monitor, keyboard, mouse and speakers attached - or it could be in one box like a laptop is.
It means the people that want to deploy remote gear can get something prepared earlier instead of assembling a system.
And this is why we needed Operation Dark Storm to blacken the sky when the Machines tried to take over. "We don't know who struck first, us or them, but we know that it was us that scorched the sky. At the time, they were dependent on solar power and it was believed that they would be unable to survive without an energy source as abundant as the sun." - Morpheus
This is a selfselfie. Or a selfie 2.0, if you prefer.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."