This paper introduces a system for nursing homes, where daily activities of elderly persons are monitored by pervasive sensors. In this system, the sensors detect and save position of an elderly person as his life log and these logs are analyzed for obtaining the evidence of his activities of daily living. The collected evidence enables caregivers to provide effective support for the elderly. In our experiment, the system accumulated position data for a month and a half and we analyzed them to obtain his daily activities models, such as the relation between the time of day and the probability of visiting the lavatory. This paper presents the concept and the overview of the system and explains further some experimental results.