Low-cost and Device-free Activity Recognition System with Energy Harvesting PIR and Door Sensors

Abstract

Progress of IoT and ubiquitous computing technologies has strong anticipation to realize smart services in households such as efficient energy-saving appliance control and elderly monitoring. In order to put those applications into practice, high-accuracy and low-cost in-home living activity recognition is essential. Many researches have tackled living activity recognition so far, but the following problems remain: (i)privacy exposure due to utilization of cameras and microphones; (ii) high deployment and maintenance costs due to many sensors used; (iii) burden to force the user to carry the device and (iv) wire installation to supply power and communication between sensor node and server; (v) few recognizable activities; (vi) low recognition accuracy. In this paper, we propose an in-home living activity recognition method to solve all the problems. To solve the problems (i)--(iv), our method utilizes only energy harvesting PIR and door sensors with a home server for data collection and processing. The energy harvesting sensor has a solar cell to drive the sensor and wireless communication modules. To solve the problems (v) and (vi), we have tackled the following challenges: (a) determining appropriate features for training samples; and (b) determining the best machine learning algorithm to achieve high recognition accuracy; (c) complementing the dead zone of PIR sensor semipermanently. We have conducted experiments with the sensor by five subjects living in a home for 2-3 days each. As a result, the proposed method has achieved F-measure: 62.8% on average.