A Wearable Multi-sensor Node for Detecting Anomalies on Worker Safety in Complex Scenarios
摘要
Smart Personal Protective Equipment (PPE) enhances workplace safety by continuously monitoring workers and environmental parameters, enabling real-time interventions in the event of accidents. This paper presents advancements in such devices, utilizing Internet of Things (IoT) technology for monitoring purposes. Specifically, it outlines improvements in a battery-powered sensor node over similar devices documented in scientific literature [1–3] and earlier prototypes [4, 5]. The IoT node can monitor chemical and physical parameters, including particulate matter, VOCs, O2, CO2, CO, as well as incorporating audio and inertial sensors, and a connectivity module for the LORAWAN™ network to transmit reports to a dashboard. It also features real-time location capabilities for both indoor and outdoor scenarios, which can localize a worker's location in the event of an accident, aiding in reducing response times for first aid. This version of the system can identify complex hazardous conditions by detecting simultaneous breaches of both minimum or maximum thresholds for relevant chemical parameters. For physical parameters, the system uses AI-driven processing to identify potentially dangerous worker activities. In audio detection, it recognizes not only threshold violations but also specific audio patterns indicative of hazardous situations. The progress includes the development of a compact sensor node with enhanced low-power features and an updated prototype sensor shield with latest sensors available in the market. Preliminary trials in simplified operational settings have assessed initial performance metrics, such as battery life, data delivery time to the dashboard, and location detection accuracy.