High Coverage and Privacy-Preserving IoT-Based Fall Detection System With Two-Thermal-Sensor Configuration
提出一种基于双热传感器和窄带物联网的跌倒检测系统，用于私密区域（如洗手间）的人员状态监测，具有高覆盖、低成本、隐私保护的特点，已在商业场所验证。
This study proposes a simple Internet of Things (IoT)-based fall detection system with high coverage and privacy preservation to detect people’s status in private areas, such as washrooms. Human detection is crucial to prevent treatment delays because falls can lead to serious injuries or death. The proposed system utilizes a two-thermal-sensor configuration with top and side sensors to detect humans and perform fall detection. The number of top and side sensors used depends on the size of the detected area. The system incorporates the narrowband (NB)-IoT technology to facilitate communication between servers and sensors as a bridge in this system. The fall detection circuit includes an NB-IoT module, a microcontroller unit (MCU), and a two-thermal-sensor configuration. This low-cost MCU connects to the sensors and performs computations to determine if a private area is occupied or not. Apart from high coverage and privacy preservation, a two-thermal-sensor configuration enhances the accuracy of the fall detection system. A convolutional neural network (CNN) can be added as an additional fall detection metric. This fully integrated system was already implemented in different commercial places, and four commercial places are demonstrated in this article. Experiments were performed via five cases to verify the performance of the proposed fall detection system. This system, compared with other technologies, has simple installation, low cost, high coverage, and privacy-preserving features.