Effects of sensor cleaning on the classification rate for an electronic nose

The electronic nose (E-Nose) is an instrument for measuring odors in air samples. In many applications processing samples rapidly to reduce testing costs is a goal. In this study, we examine the impact of eliminating the sensor cleaning step that is customarily inserted between measurements. We have found that cleaning the sensors does improve classification accuracy, but not significantly in some cases. Future studies to characterize optimal operating sequences for E-Nose devices are planned.