A wearable sensor for AAL-based continuous monitoring

Continuous monitoring of safety and health conditions are among the primary goals of Ambient Assisted Living technologies. Effective solutions, aimed at fostering independent life of elderly persons, need for carefully balancing system intrusiveness, perspicacity, reliability and cost features. Heterogeneous networks, including different combination of environmental and personal (wearable) sensors can be used, with relevant value coming from data fusion: analysis techniques, aiming at inferring safety and health-related information in an indirect fashion from behavioral features are being deeply investigated with this aim. Device cooperation and interoperability are thus key factors: in this paper, the development of a wearable sensor suitable for broad range AAL application is introduced, addressing features specifically oriented to behavioral analysis. First, the device itself is capable of analyzing different features of the user motion patterns, synthesizing high-level information (simple task identification, energy expenditure) on board. This result in better battery management (less data transferred over the radio link) and interoperability (thanks to data abstraction). Second, by means of a suitable operating protocol, it cooperates with environmental sensors (e.g., a toilet sensor) providing the latters with user identification information, and thus allowing to exploit related data even in a multi-user context. This avoid the need of more expensive and complex indoor localization techniques or of more intrusive identification technologies (e.g., NFC/RFID tags).