Seamless human life monitoring and tracking all-day long

Human life monitoring systems utilizing wireless sensors networks (WSNs) and/or smart phones became a hot topic for the evaluation of the life quality. A daily life of a human can be divided into three main parts as outside, home and sleep. There are various systems that monitor the lifestyle of a human either inside or outside the home. Yet, the challenge is to develop a system that covers all the activities of a person for 24 hours. Considering the fact that people spend one-third of their lives sleeping, sleep is another important activity to monitor. While sleep studies mainly focus on the sleep quality of a person, the effects of life style and ambient factors on the sleep quality are usually neglected. In this thesis, we propose a seamless human life monitoring system that covers 24 hours of a person's life including the sleep activity. The proposed system utilizes a WSN and a smart phone and collects life-log and sleep data from multiple users. The WSN collects nocturnal ambient and sleep data via various sensors. On the other hand, the applications running on the smart phone collect daily performed activities with their durations and locations. In the data collection phase, we employed nine people for fifteen days. The system is designed to provide the unobtrusiveness and respect the privacy of the users. By using the collected data, we extracted the sleep behavior and the life style choices of the users. In order to figure out the factors affecting the sleep quality of a person, we applied three feature selection algorithms namely the decision tree, the correlation coeficient, and the sequential feature selection to the collected data. The results indicate that two features namely the Sleep Onset Latency and the Leisure Activity Duration are reported as important features in all of three algorithms for their effects on the sleep quality.