The BOND Nose
There has been a great deal of research centred on how the olfactory sense is processed in the brain, with many surprising revelations forthcoming from that research on how scent is ordered into orchestras and multiple signals from different receptors make up detection of each smell. This is of course an area of vital research if we are ever going to add scent into virtual environments without relying on physical odour release. But, there is as very different angle also to be considered.
What if, instead of working on how the brain interprets scents, we work on creating artificial olfactory bulbs? Approach the problem from the other end, and create an artificial nose as capable as a human, or maybe even a canine nose, at differentiating different odours? Such would give us a wealth of data on how the olfactory bulbs themselves, have4 to function, and lead to a plethora of augmented reality applications for technological noses as powerful as organic ones.
Two projects, one building upon the results of the other, have been trying to address this. Both are multi-university and private laboratory, heavily funded efforts. The first, the SPOT-NOSED (Single Protein Nanobiosensor Grid Array) project had a goal of constructing artificial olfactory bulbs by fusing organic tissue and electronic components into a single device. The second, the BOND (Bioelectronic Olfactory Neuron Device} project is taking those results far further, integrating them into a massively parallel array of sensors to be built into a hand held odour detector a little bigger than a gesture interface wand.
The original SPOT-NOSED project used compound nanobiosensors, made by coating a microelectrode with a layer of olfactory receptor proteins taken from cells derived from animal noses and grown in solution. These living proteins bond with passing scent molecules, and the electrical reaction produced, is picked up by the attached electrode. In short, the actual olfactory bulb is still doing much of the work, but the challenge was to interface the technology with it, in such a way as reliable detection was possible, close to 100% of the time, and in concentrations too small for human noses to detect.
This was successful, creating a functioning array of proteins in a langmuir-blodgett film, and an artificial nose that could detect a single scent at a time was born.
The researchers checked every conceivable method they could think of to get a successful signal from the proteins in a reliable manner, including I-V measurements, impedance spectroscopy. Capacitance and signal noise measurements. They succeeded in being able to identify the waveform of protein activation, to be able to identify it every time, and detect when a pulse was about to be sent.
BOND's primary goal is thus obvious; to take the new artificial noses created by SPOT-NOSED, to differentiate them such that each sensor responds to a different odour, and to determine the ideal placement of such protein-sensors on an integrated chip, as to maximise the range of odours identified. This is the precise opposite of those efforts to determine how the signals are deciphered in the brain, and is likely to prove just as invaluable, as we learn the reason why the olfactory bulbs in nature, are arranged just how they are.
The initial design calls for sensors to be arranged in groups of sixteen (for perhaps obvious processing-related reasons) each group connected to it's own processor microchip. Many, many different designs will have to be trialled, in order to produce an optimal design for even a moderate range of odours. One of the greatest challenges facing the multi-use sensor system, will be how to unbond odours from the proteins, in order to avoid a sensor becoming increasingly clogged throughout its lifetime.
SPOT-NOSED summary (pdf)