Abstract:
With the increase of dependant people such as the elderly or those suffering from various
illnesses, researchers are increasingly interested in the implementation of "Smart Homes" (or
automated homes) to help them continue to live at home safely and independently. They
incorporate common devices that control the characteristics of the house. Originally, Smart
Home technology was used to control environmental systems such as lighting and heating,
but recently it has been adapted to many different areas and demands, such as ensuring the
comfort and safety of residents. The research in this dissertation focuses on the safety of an
Alzheimer's patient in the Smart Home. Our contribution consists in adapting a predictive
algorithm, which is generally used for comfort, associated with a multi-agent architecture,
in order to detect dangers and consequently make the right decision at the proper time.
Several simulations have been carried out which demonstrate the reliability, efficiency and
scalability of the proposed solution.