Gerka, Alexander; Lübeck, Felix; Eichelberg, Marco; Hein, Andreas
VDE-Kongress 2016 – Internet der Dinge. Technologien / Anwendungen / Perspektiven.
Smart meters have been developed to improve the efficiency of metering procedures and to enable novel applications such as appliances making use of flexible tariffs based on the current availability of energy from renewable sources. This paper presents a novel application for smart meters in private households: The identification of health parameters and needs of dementia patients based on their usage of electrical appliances. The system presented in this paper aims at identifying odd behavior patterns of dementia patients. It improves the ability of a caretaker to identify long-term changes in the behavior of their patients. For this purpose, a smart meter, motion detectors and a water consumption sensor are installed in the flat of persons with dementia. The high sampling rate of the smart meter used in this setup ensures an accurate detection of electrical appliances in the flat. The data from the sensors is processed by machine learning algorithms on three levels. The system is developed and tested in the context of the QuoVadis research project.