A Monitoring System for Assessing Life Pattern of the Elderly Living Alone
Seonwoo Lee, Seong-Ho Song · INTERNATIONAL JOURNAL ON Advances in Information Sciences and Service Sciences · 2011
This paper describes a monitoring system to assess the daily activities of the elderly who live alone by using a circadian model. The developed system is composed of an in-house sensing system and a server system. The in-house sensing system installed in their own houses is a typical wireless sensor network system including three kinds of wireless sensors. The server system has a database server and an assessment server. The assessment server builds a circadian (24 hour) model that represents everyday life behavioral rhythms of a monitored resident living alone. The proposed model is composed of fuzzy sets which reflect degree of repetition of daily activities such as staying of each room, going-out,and sleep behavior. Such activities can be considered as feature components of a life pattern of the resident. Based on the model we define a measure called ‘degree of anomaly’ that can indicate behavioral deviations from usual habits. In order to evaluate the performance of the developed system and the assessing method, we installed the systems into three elderly houses, collected data during over three months averagely, then validated the modeling and anomaly detection method.