AN ONLINE ODOR CLASSIFICATION SYSTEM FOR ELECTRONIC NOSES

Abstract

The design of technical instrumentation based on biological principles, which are utilized by human perception senses to interact with the environment, has been a principal subject of research in several engineering disciplines for decades. However, only some of these artificial senses, such as machine vision, have comparatively reached a sufficient level of maturity so as to allow their regular use in nowadays industry. The challenge of artificially replicate the biological olfactory sense has resulted in compact electronic systems commonly known as electronic noses or e-noses. These systems require, on the one hand, the use of sensors specifically designed to detect odors, and secondly, the application of artificial intelligence software techniques for data processing and classification in a real-time basis. In this paper, a novel artificial olfaction system based on web services, which enables the implementation of distributed networked online access systems, is introduced. The proposed system allows online access to the software, which is responsible for data processing and recognition, regardless of the location of the terminal detector, thus facilitating the design of smaller electronic noses, suitable for use in connected smart, connected IT environments.