Data driven multi-agent m-health system to characterize the daily activities of elderly people

With the continuous growing of aging population, the society is facing new challenges, namely the implementation of healthcare services for older people, as well as the promotion of the active aging and well-being. These challenges imply the optimization of these services through biomedical, physical, psychological and socio-environmental interventions. ICT technologies can support the implementation of these healthcare services, e.g., the use of wearable devices to collect the physiological data, cloud technology to store big amounts of data and advanced data analytics algorithms to extract valuable conclusions and actionable knowledge. This work proposes a multi-agent system driven data analysis approach to characterize the daily physiological conditions of a group of elderly people institutionalized in a nursing home, supporting the healthcare professionals to monitor their behavior and promote an active aging. The individual' data were collected by a set of Fitbit Charge HR wristbands and analyzed by clustering algorithms, running in the distributed autonomous agents, allowing to identify and characterize the individuals' daily habits and physiological conditions.