Detecting anomalous elderly behaviour in Ambient Assisted Living

The increasing availability of sensors and intelligent objects enables new functionalities and services. In the Ambient Assisted Living domain such technologies can be used for monitoring the elderly behaviour, and reasoning about it to detect possible anomalous situations, which could be a sign of the next onset of chronic illness or initial physical and cognitive decline. In this paper we propose a solution that exploits task models describing expected user behaviour, and a context manager able to detect relevant contextual events and conditions determined by the actual elderly behaviour. Planned and actual behaviour are compared to detect if any deviation occurred. The resulting environment is able to generate multimodal actions such as reminders and alarms aiming to provide useful support when such anomalous behaviour is detected.