The possibility of using electronic noses (ENs) to measure odor intensity was investigated in this study. Two commercially available ENs, an Aromascan A32S with conducting polymer sensors and an Alpha M.O.S. Fox 3000 with metal oxide sensors, as well as an experimental EN made of Taguchi-type tin oxide sensors, were used in the experiments. Odor intensity measurement by sensory analysis and EN sensor response were obtained for samples of odorous compounds (n-butanol, CH 3 COCH 3 , and C 2 H 5 SH) and for binary mixtures of odorous compounds (n-butanol and CH 3 COCH 3 ). Linear regression analysis and artificial neural networks (ANN) were used to establish a relationship between odor intensity and EN sensor responses.The results suggest that large differences in sensor response to samples of equivalent odor intensity exist and that sensitivity to odorous compounds varies according to the type of sensors. A linear relationship between odor intensity and averaged sensor response was found to be appropriate for the EN based on conducting polymer sensors with a correlation coefficient (r) of 0.94 between calculated and measured odor intensity. However, the linear regression approach was shown to be inadequate for both ENs, which included metal oxide-type sensors. Very strong correlation (r = 0.99) between measured odor intensity IMPLICATIONS Odor pollution is a significant issue throughout the developed world, and several legislative bodies have regulated or are in the process of regulating odorous emissions in the environment. The measurement of an odor is not an easy task, and existing methods suffer from major drawbacks. It thus appears that there is a real need for the development of new reliable odor measurement techniques. This paper investigates the possibility of using ENs to measure odor intensity. The proposed procedure, which comprises training ANN to predict odor intensity based on EN sensor response, was applied with success for simple mixtures of odorous compounds.and calculated odor intensity using the ANN developed were obtained for both commercial ENs. A weaker correlation (r = 0.84) was found for the experimental instrument, suggesting an insufficient number of sensors and/or not enough diversity in sensor responses. The results demonstrated the ability of ENs to measure odor intensity associated with simple mixtures of odorous compounds and suggest that ANN are appropriate to model the relationship between odor intensity measurement and EN sensor response.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.