LoRa-Based Wearable Monitoring System for Elderly Vital Signs and Emergency Alerts

Abstract

This study presents a LoRa-based wearable system for monitoring the vital signs of older adults and generating emergency alerts. The system integrates a wearable node, an SX1262 LoRa link, a gateway, Google Sheets logging, a web dashboard, and LINE Messaging API caregiver notification. The urgency-aware DATA/ALERT architecture separates routine physiological monitoring from emergency management. Heart rate, SpO2, and MAX30102-derived temperature-related values were sent as DATA packets, whereas fall and SOS events were processed at the wearable edge and sent as ALERT packets. The prototype was evaluated through vital-sign comparison, fall and activities of daily living (ADL) testing, LoRa field measurements, and battery-duration and alert-latency tests. For 20 paired measurements, MAPE was 5.08% for heart rate, 0.99% for SpO2, and 0.52% for the temperature-related value. Fall testing achieved 94.17% sensitivity over 120 trials, while 180 ADL trials yielded 97.78% specificity; the overall accuracy and F1-score were 96.33% and 95.36%, respectively. DATA packets remained receivable at 820 m under LOS with a mean PRR of 76.00%, and 81.33% at 450 m under NLOS. The battery durations were 1 h 59 min and 3 h 21 min for 12 s and 60 s DATA intervals, respectively. Across 45 NLOS SOS trials, all alerts reached the gateway, LINE, and web dashboard, with mean end-to-end latencies of $3.67 \pm 1.15$ s and $10.36 \pm 1.69$ s for LINE and the dashboard. These results demonstrate the feasibility of integrating physiological monitoring, web visualization, and emergency notification in a LoRa-based wearable prototype for elderly care.