TOWARDS PROACTIVE HEALTHMONITORING: LOCATION-BASED STATISTICAL BEHAVIOR MODELING

As the elderly population increases, more elderly persons are living alone. The well-being of the elderly population is becoming an important social issue. It is envisioned that the sensor network technology can provide necessary assistance to elderly persons living alone. Sensor networks can be used to collect useful behavioral and biological information to monitor the wellness of an elderly person. With its easy installation, expected low product cost, low maintenance cost, and non-invasive monitoring scheme, sensor networks are capable of radically changing our society. Allowing elders to stay at home with the assistance of sensor networks, rather than elders going to retirement or nursing homes, helps decrease the annual expenditures on health care and allows the elder to live freely and economically. The sensor networks used in this project was the passive infrared (PIR) sensors and trio modes, which are the receivers that collect the data. All of the equipment used in this experiment was wireless. The goal of this project was to build a statistical behavior model of an elderly person living alone with the PIR sensors. Experiments are performed using a miniature model home which is consisted of five rooms with one PIR sensor in each room, a dining room, living room, bedroom, bathroom and kitchen. The PIR sensors detect movements throughout the home. After we collected all the sensor data we developed an algorithm which can detect irregularity in behavior, such as not going to the kitchen or the bathroom or even staying in one room all day. 1