One of the critical technologies for many of our modern applications is the availability of effective sensors and transducers. They are so widespread and embedded that we often take them for granted, or don't fully realize the vital role they lay. We can detect--with high sensitivity and precision-- light (even individual photons), sound, pressure, acceleration, position, current, voltage, temperature, electric fields, magnetic fields, and more. Many of these sensors are low in cost, high in performance, small, and use little (or no) power.
But there's one area which has eluded our ability to produce a low-cost, easy-to-apply sensor: smell (and taste, which is closely related). I thought about this last week when I saw two uncorrelated items the same day. The first was a TV show about dogs and their sense of smell, with a focus on the bloodhound. The show explained how the bloodhound's nasal passage and structure allows for a long air-flow path, while also diverting the exhaust breath from interfering with the air intake. The second was news that one of the co-winners of the 2004 Nobel Prize in Physiology or Medicine, for deciphering the workings of the sense of smell, has retracted two scientific papers after she and her colleagues were unable to repeat the findings, here. It looks like smell is still a big mystery, despite what we have learned. Even insects with their small brains, such as mosquitoes, can sense specific scents and know where to go, as well as what to avoid.
Certainly, we do have complex instruments, such as liquid or gas chromatographs, which can replicate the functions of the nose to some extent. We also have molecule-specific sensors which can detect individual scents, such as targeted hydrocarbons, using localized physical and chemical interaction at the sensor site.
But we don’t have a wide-ranging, general-purpose scent transducer—at least, not yet. There's lots of work being done in this area, using MEMS devices, nanochemistry, and other advanced technologies, but there's a long way to go.
I wonder: will there be some sort of breakthrough in this area? Or will improvements in scent and smell sensors be a long, tough slog? Will someone put an entire chromatography system on a chip, so we can have an easy-to-apply device?
Sometimes it's good to stop thinking about and working on the day-to-day problems, and think about the bigger, more complex, less tractable issues. If nothing else, it's a great reason to procrastinate! ?