Development of an IoT-Based Fall Detection System for Older Adults Using Fuzzy Augmented Double Thresholding

Falls are a significant health issue among older adults, and may result in physical injury, delayed medical assistance, and post-fall complications. IoT-based wearable systems offer a potential solution for real-time fall detection and emergency notification. This study aimed to develop and evaluate an IoT-based fall detection system for older adult health monitoring using the Fuzzy Augmented Double Thresholding (FADoTh) method. This work was based on systems engineering. It was implemented using an MPU-6050 sensor and an ESP32 microcontroller, NEO-6M GPS module, and buzzer alarm, with Telegram Bot notification system. Functional tests were performed on three healthy young adult participants for simulated fall scenarios and Activities of Daily Living (ADLs). Sensor data were processed using the FADoTh algorithm and analyzed using Microsoft Excel. The system could detect simulated fall events in different directions and distinguish most ADL occurrences from fall events. Emergency reports with location details were successfully transmitted via Telegram Bot when a fall was detected. Nevertheless, false alarms due to stair ascent and descent movements mimicking an actual fall incident were observed. The developed IoT-based fall detection system has potential to support continuous health monitoring, caregiver response, and elderly safety. Further evaluation involving larger participant groups and real-world elderly monitoring scenarios is needed to improve reliability and clinical applicability.