Household Behavior Observation System Utilizing Mesh Sensor Network Consisting of Smart Taps

Abstract

In recent years, the devices focusing on the smart home have rapidly been spreading and have been expected to be applied to the services that support daily life of the residents such as the observation of daily activity of the elderly people living alone and controlling the energy efficiency. To realize the smart home, the system that can recognize daily activities of the residents with high accuracy and low cost is indispensable. The existing studies have proposed the recognition methods using various sensors such as cameras and ultrasonic sensors for estimating the living behavior. However, the existing methods have problems in terms of the privacy and installation cost. Therefore, in this study, we propose and develop a new behavior observation system that estimates daily activities of the residents by placing smart taps supporting the wireless communication on various places at home. The smart tap has a function of measuring and analyzing the power consumption of the connected devices and the received signal strength of radio waves between the smart taps. In addition, the system constructs a mesh network of the smart taps using the Thread, a type of short-range wireless communication, and the location and behavior of the residents are estimated by collecting and analyzing the measurement data. By analyzing only the measurement results of the power consumption of appliances and the received signal strength of the radio wave, the proposed system achieves accurate behavioral estimation with no privacy issues and low installation cost.