Device to Device Collaboration Architecture for Real- Time Identification of User and Abnormal Activities in Home

Abstract

Activities of Daily Living (ADL) are indicators for evaluating individual health, ability of independence and daily living, and degenerative brain disease of old people. Therefore, many researches are actively underway to measure user's ADL data by constructing Internet of Things (IoT) based smart home. However, general smart home solutions for measuring user's ADL only focus on collecting user's activity data, appliance usage and home environment data. Such simple ADL data cannot be used as an indicator for early recognition of the above-mentioned symptoms of the elderly people. Intuitively speaking, the ADL data we want to collect should be to know who the user is, and whether the device has been successfully used or misused. In this paper, we propose device-to-device collaboration architecture to identify the user, device to use, and success or failure of the device usage in real-time. By designing and implementing the proposed architecture, we can record the ADL data on the user's wearable device without any user intervention. In addition, as another advantage of the proposed concept, it is possible to easily check and record the physical moving ability of the user between two fixed spaces. The collected ADL and abnormal behavior may help a user or guardian to determine the user's dementia symptoms, activeness and daily living skills.