I hear of a lot of startups working on image detection and sound detection, but I'm wondering why there isn't anything done for scent detection. Maybe A2D for scent doesn't really exist yet, but it would limit the need for drug sniffing dogs
> I'm wondering why there isn't anything done for scent detection ... it would limit the need for drug sniffing dogs
If police departments believe that they get the resources they want based on the Clever Hans Effect, what would be their incentive to use a more accurate mechanism for sniffing out drugs?
I did some work during my PhD on recognizing scenes from electronic-nose data. The problem is more on the hardware side of things than the algorithm side of things. Today's electronic noses are still pretty large devices in which a sample needs to be placed within the device. Unless the state-of-the-art has advanced significantly in the past 3 years, none of them are like wielding an artificial nose on a stick that you can just wave around sniffing things.
Those are things similar to a mass spectrograms. The tissue picks up particles, and is then analyzed by the machine to create a mass spectrogram of the napkin. If it contains masses similar to explosive powder residues, then you're in trouble.
Olfaction is really amazing. (Even the crappy human version, to say nothing of e.g. canines.) We can build cameras that see better than eyes, microphones that hear better than ears, but there is no technology that comes close to matching a nose. Sure there are some special-purpose sensors that measure the precise concentrations of particular gases, but nothing so general as a what's found in all sorts of different animals. It's not even clear how to quantify smell, for that matter.
This kind of tech has been around for decades in the CBRN community. The initial gotchas would be lack of effectiveness with the commonly small quantities of substances in the user end of the market, and the high cost of equipment. Also the common use of radioisotopes for the detection devices that require NRC compliance headaches.
I too have wondered this. It seems like it would be relatively simple to create a device that can detect drugs more accurately than a dog. The only problem might be that creating such a device might cost significantly more than training a dog.
