Congratulations to Sen Hirano on passing his Phase II Exam with the paper “Exploring the Potential for Gas Sensors to Classify Odors in Ubicomp Applications.”
“The sense of smell is the primary method many animals use to infer information about their environment. Although human noses are less developed, it is still a sense on which we rely heavily. Ubicomp researchers have long been interested in sensing user activities in an unobtrusive manner. One way to do this sensing is through detecting the gases released during those activities. Previous research has shown that gas sensors can be used to classify odors when used in highly controlled experimental testing chambers. However, potential ubicomp applications require these sensors to perform an analysis in less controlled environments. In this paper, I present my design of uSmell—a gas sensor system for sensing smell in ubicomp environments. My system samples an odor fingerprint from a variable amount of metal oxide semiconductor (MOS) gas sensors every second. It then processes the time series data to extract three features that highlight how time and distance affect the gas sensors’ ability to react to the gas molecules released by an odor, for each five-second window. I evaluate uSmell through four experiments: basic efficacy, effects of airflow and distance, classifying bathroom activities, and tracking cooking state. These experiments demonstrated my system’s ability to classify a set of odors in a small container with 88% average accuracy, a set of odors in open air 0.5-2 m from the odor samples, bathroom activities with 93% accuracy, and doneness for waffles and popcorn with 73% and 85% accuracy, respectively. I close with a discussion on opportunities for optimizing classification, opportunities for using shapes and patterns for recognition, and future directions for using this technology to help guide cooks.”