Toward a context-aware and automatic evaluation of elderly dependency in smart homes and cities

In this work, we propose a framework for automatic dependency evaluation that can notify the changes of the elderly dependency, hence providing persons with required services in time. In order to validate our approach, we target the French dependent population. Especially, we focus our proposal on the AGGIR dependency model used in the French medical field. Our objective is to provide a flexible architecture and an extensible model linked to opened data that can refer to a wide variety of services (hardware sensors and software; simple or composed services).