- Data from regular smart electricity meters allows for accurate detection of occupancy in residential buildings

Detecting when a household is occupied by its residents is fundamental to enable a number of home automation applications. Current systems for occupancy detection usually require the installation of dedicated sensors, like passive infrared sensors, magnetic reed switches or cameras. Using data from smart electricity meters that are currently being installed in millions of homes makes additional hardware obsolete, thus reduces cost and effort to set up such systems.

We collected electricity time-series data from five households over a period of eight months, together with ground truth data on occupancy. The data has been used to train several standard machine learning algorithms.

The results show that by using common classification methods, it is possible to achieve occupancy detection accuracies of more than 80% in practical settings – despite the fact that many electrical appliances are also activated when homeowners are away, such as fridges, washing machines, and heating systems.

Non-Intrusive Load Monitoring: Do NILM-Researchers Promise too much?