Machine Olfaction

Abstract

It has been a longstanding goal of many research groups to replicate human olfactory sense with instruments. Sensor technology aims not only to replace the traditional analytic methods that are mostly focused on individual chemical identification and quantitation, but also to predict the human perceptions of smell, odor recognition and odor hedonics, thus replacing human sensory evaluation. Sensors have progressed from early gas sensors, to e-noses and e-tongues to biosensors and bio-e-noses that utilize elements from natural signal transduction to gain better sensitivity and selectivity. There has recently been a rapid increase in research and development of advanced sensor technologies and enabling technologies such as nanotechnology, cellular biology, wireless communication, and neural computing methods that have helped overcome the sensitivity, selectivity, portability and recognition problems of early sensor systems. Much of this development comes in response to global bioterrorism and other security threats. The activities in the various areas enabled by machine olfaction are poised to impact many industries not only as potential enablers of competitive advantage, but also through international standards development and enforcement. However, while machine olfaction instruments and sensors systems have been under development for more than 30 years, they still cannot completely replace the human senses for sensitivity, selectivity, and speed. While complete replacement of human sensory perception is not yet possible, certain sensor arrays provide fast, cheap, portable, networkable, low-expertise alternatives in some applications where simple detection is required. Nevertheless, current machine olfaction devices can provide a low-sample preparation approach that significantly reduces the amount of human sensory and advanced chemical testing needed.