Intelligent Machine Olfaction

M. Kermit, Å.J. Eide (Norway), T. Lindblad, and K. Agehed (Sweden)


Electronic nose (ENose), Machine olfaction, Pattern Recognition, Detection of explosives


The recent technological advances in the field of gas sensors and artificial intelligence has made it possible to develop systems mimicking the mammalian nose. De vices that respond to what a human will interpret as an odour are termed artificial or electronic noses. With a proper pattern recognition subsystem attached to the instrument, such a mechanism has a great potential and a large number of possible applications. Unfortunately, commercially available electronic noses are expensive and predominantly for laboratory use. In this paper, we propose a robust and low-cost design of an artificial nose using inexpensive and standard available gas-sensors. Starting with hardware requirements and electronic circuitry, a complete system with proper data analysis and recognition is described. A practical example using measurement data acquired from different types of explosives is also considered.

Important Links:

Go Back