Slashdot Mirror
Artificial Nose Works By Color
Alien54 writes: "As reported here in the Science Daily News, chemists Kenneth Suslick and Neal Rakow at the University of Illinois have developed an artificial nose that is simple, fast and inexpensive - and works by visualizing odors. Called "smell-seeing" by its inventors, the technique is based on color changes that occur in an array of vapor-sensitive dyes known as metalloporphyrins - doughnut-shaped molecules that bind metal atoms. Metalloporphyrins are closely related to hemoglobin (the red pigment in blood) and chlorophyll (the green pigment in plants) Smell-seeing arrays have many potential uses, such as in the food and beverage industry to detect the presence of flavorings, additives or spoilage; in the perfume industry to identify counterfeit products; at customs checkpoints to detect banned plant materials, fruits and vegetables; and in the chemical workplace to detect and monitor poisons or toxins. The full text is available as a PDF file (but is recommended for chemistry geeks only)."
Add that to the machines that analyze the "aura" of heated air that surrounds our bodies, and you can get a stinkometer the likes of which has been heretofore confined to the dark recesses of deodorant company imaginations. Till then, it looks like a cool approach to the problem of identifying smells electronically for all kinds of other purposes.
Actually, this is by no means a new idea. The East German Stasi had a program where smells were gathered through a variety of means and stored in sealed jars. The jars could be opened months, even years later and the scents (usually preserved on a piece of fabric) presented to bloodhounds and the like. Surprisingly enough, this actually worked pretty well.
References to this practice can be found in a number of places, including reporter's notes (search for "smell samples") archived at the Federation of American Scientists.
So it's fairly clear that people and governments are not only capable of doing this sort of thing, but, in fact, it has been done (and will be done again if it's deemed valuable, I have no doubt). However, it must be pointed out that the usefulness of electronic bomb sniffers and smell based weapons locators is huge, and, properly applied, would likely save many lives.
--
Behold the Power of Cheese!
Till then, it looks like a cool approach to the problem of identifying smells electronically for all kinds of other purposes.
I am not sure we might soon have some Stinkometer tomorrow as, if I understand that a device using this technology could detect some smell, and maybe recognize some that stink, there will still be issues about mixes, for example, if you take a flower perfume and some food smell, they could independantly smell good but their mix could smell awful (Roquefort cheese + flowers, for example).
Except in few cases when mixing odorous gas will induce molecular changes (that could reasonably make a sensor react properly), these smells will consist of sets of smell clouds which could independantly be analyzed but may probably not (yet) be classified as a unique smell type.
So, I accept the idea of an electronic nose though we may all agree that we need (1) enough sensors to detect all these subtilities (2) a brain in order to analyze the resulting smell by confronting all the sensor results simultaneously.
I don't even mention cultural facts that'll make (for example) alcohol, smell better for non-Muslims than for Muslims and thus will require different classification schemes.
A typical application of this study (IMHO) would be to detect gas (c3h8, c4h10) leaks in houses in order to avoid explosions. Cheers
PS: BTW, it is funny not to have a single occurrence of the word smell in the PDF file...
--
Trolling using another account since 2005.
A system somewhat like this is half of what is needed to make the recording of odor/flavor precise enough for legal action--the other half is an understanding of what goes on in our brains in response to a given set of chemicals.
Note that if done properly, this will allow the digitization of odor. To play back an odor you'd spray a mixture of simple compounds that would have the same neurologic effect as the compounds being measured.
(Technical note: metalloporphyrins can potentially be used to inexpensively identify a large number of compounds simultaneously; mass spec requires a very expensive machine to do this.)
Okay, I suck at chemistry, so feel free to step in and correct me, but isn't "smell by color" a pretty obvious method? The fact is that "smell" isn't really a sense, in the way that seeing and hearing are senses. Your body actually takes in some particles, and your brain looks them up in and determines what they are, then feeds you a sensation based on the response. (I know I've probably over-computer-metaphored this, but the point I'm trying to make is that all of this is abstract and methaphoric, since senses are just electro-chemical responses in your brain, and the computer doesn't have a chemical brain)
So the computer has to identify the chemical, then the process of recognizing it is the actual "smelling." The only two ways I can think of to identify a chemical are by physically contacting it with chemical receptors (this is the way noses do it, I believe, but this is far from practical with computers) or by looking at the molecules. Once you're looking at the molecules, the wavelength/frequency of light emitted (color) is the best way to recognize what molecules you're looking at.
So smell by color isn't nearly as odd as it sounds.
20 posts and not one fart joke. I'm ready to weep tears of joy. I knew that /. was slowly improving due to Taco's inspired moderation technology and now we have proof.
/. fart-joke free and on-topic.
Thank you Taco and thank you slashdotters. Let's work together to make
Together we can make it happen.
--Shoeboy
scan, wipe, scan, compare
I don't know what's wrong with this, it's what I do with my real nose. Not to mention my ass. Is there something wrong with modeling artifical organs off of real ones?
--Shoeboy
It just seems like something that sounds cool, but may not be all that practical in the real world.
The area where I see where this having the most popularity is in spying^H^H^H^H^H^Hlaw enforcement. We already have police dogs that sniff out criminals; electronic smell recognition seems to be a (un)natural extension of this. Can't you imagine the FBI or NSA with a huge database of peoples' "smells" and matching them with gloves left behind at a crime or sperm samples from a rape? Of course, you wouldn't have to register your smell with them -- after all, you don't really need that driver's license!
It's amazing how quickly technology is eradicating whatever notions of privacy that people still had. We already have our appearance, blood type, and actions recorded and disseminated all around the world; now we're going to have our smells tracked too? What's next, our skin texture?
Sorry I don't have a login, so I'll just post this with my name at the end.
This is just one example of a new field in biochemistry called "porphyrins". I've been working on them for my doctorate thesis for the last three years, but I never expected to see them mentioned on Slashdot.
The way metalloporphyrins (and the larger group of porphyrins) work is that when a particular molecule hits a pre-prepared surface coated with the dye, the metalloporphyrin alters its structure through movement in the electron shells to mimic the molecules form, producing an altered isotope of the metalloporphyrin that has different optical properties, reflecting only a select sequence of wavelengths. This is caused by the electron shells collecting energy from the photons as they arrive, and re-emitting them in accordance with Precontelli's principle of refraction.
Porphyrins have much in common with the family of chromopolymers, as well (at least in their macroscopic physical interactions. It's a very interesting field of research, and I hope that some future chemists might become involved.
-- Chris Ellendale
Ph.D candidate, University of California
For more information, see the following links:
Porphyrin Co., Ltd.
Summary of Porphyrins and Metalloporphyrins
Metalloporphyri ns
Photosensisizing effects of metalloporphyrins