Using Non-Intrusive Environmental Sensing for ADLS Recognition in One-Person Household
Long Niu, Sachio Saiki, Masahide Nakamura · International Journal of Software Innovation · 2018
This article describes how pervasive sensing technologies are promising for increasing one-person household (OPH), where a system monitors and assists a resident to maintain healthy life rhythm. Automatic recognition of activities of daily living (ADLS) has been a hot research topic in pervasive computing. However, most existing methods have limitations in development cost, privacy exposure, and inconvenience for residents. To cope with the limitations, this article presents a new ADL recognition system especially for OPH. To minimize the development cost as well as intrusions to user and house, the system exploits an IoT-based environment sensing device, called autonomous sensor box (sensorbox) which can autonomously measure 7 kinds of environment attributes. The system then applies machine-learning techniques to predict 7 kinds of ADLS. Finally, this article conducts an experiment within an actual apartment of a single user. The result shows that the proposed system achieves the average accuracy of ADLS recognition with about 88%, by carefully developing the features of environment attributes.