Understanding Distributions of Environmental Parameters for Thermal Comfort Study in Singapore

Abstract

The National Science Experiment (NSE) is a nationwide project that began in 2015 in Singapore. As apart of this project, school students in Singapore carry the wearable device SENSg which measures environmental data such as temperature, humidity, light intensity, ambient noise, and air pressure. In this study, we analyzed NSE data collected during the year 2016 and examined how factors such as mode of transportation, location, and academic performance relate with various environmental parameters. We studied the relationships between environmental parameters such as temperature, humidity and noise with location, mode of transport, time, and student academic achievement. We studied three out of the five transportation modes, those that involved the user action in directly interacting with the corresponding settings (AC settings, humidity, etc.): not walking, walking, and traveling by car. We clustered them corresponding to a particular mode and by location in order to gauge mode specific and geographical patterns of environmental parameters in Singapore. For these first two tasks, we used the Expectation Maximization Algorithm in order to fit a n-component Gaussian Mixture Model to each of the three smaller datasets - where n was dependent on the mode of transportation or the geographical location. Finally, we compiled a list of the top 150 highest ranked secondary schools in Singapore in order to identify patterns of thermal comfort in schools with high rankings. We envision that the results of this study will provide insight into distribution of environmental conditions across Singapore.