Signal Transduction
The NOSE is based on a multi-element array of volatile organic compound (VOC) sensors, each based on ABTECH's Interdigitated Microsensor Electrodes (IMEs). The IME's are covered with a thin, VOC-responsive conductive polymer membrane that may be formed from either of polyaniline, polypyrrole or polythiophene. When a VOC or mixture of VOCs is passed over these polymers the vapors are absorbed by the polymer network, causing measurable changes in electrical impedance. These changes in impedance are directly detected across the array of eight sensor elements and recorded as signal data.  In addition, temperature and relative humidity are recorded at the same time.

The Sensor Response Data
Each sensor element is based on a different polymer membrane acheived by modification of the three basic conductive polymers. Although each sensor is unique, none is specific to any particular vapor.  No one sensor element would be able to deliver useful information about the gas. However, the array of sensor elements, taken as a whole, contains all the information. This information is embedded in the group response of the array and isn't readily interpreted. Thus, some form of sophisticated data analysis is necessary.

Data Analysis
To extract the gas composition and concentration level within a target vapor stream, data from the sensor responses (the signal data) is passed to an artifical neural network (ANN). The ANN has been trained to recognize specific vapors and their concentration levels, over a specificed range of relative humidity and temperatre, within the air. As long as the vapor has been trained for, the ANN will be able to identify and quantify the concetration of that vaopr, or mixture of vapors, in the air down to its ppm detection limit.