Recognition System of Activities in Daily Living Based on Deep Sensing of Water Usage
Keidai Hori, Hiroshi Yamamoto · 2021
In recent years, the number of elderly people living alone has been increasing with the accelerated aging of society, and shortage of the labors of the elderly care sector becomes a serious problem. In order to solve this problem, the activity observation and recognition systems for the remote elderly care have been proposed. Many of these services often use cameras, and image analysis technology enables highly accurate recognition of activities. However, the possibility of invading the privacy of users and the high cost are regarded as problems.Therefore, in this study, we propose a new recognition system of activities in daily living based on sensing of water usage in the house because the usage pattern of the water is deeply related with the activities of the user in the house (e.g., washing, cooking, taking a bath). In order to avoid privacy issues, the proposed system obtains only information about time change of the water usage at a specific place in the house and estimates the type of activities of the people living in the house by analyzing the time-series data of the water usage utilizing the machine learning technologies. In addition, the proposed system builds a mesh network consisting of a large number of sensor nodes installed in the house using Thread communication.