Desenvolvimento de uma pulseira inteligente para detecção de quedas: uma aplicação voltada à segurança de idosos e pessoas com deficiência.

Abstract

This work presents the development of a low-cost smart wristband for fall detection, focusing on the safety of elderly people and individuals with disabilities in domestic environments. The proposal aims to fill the gap between expensive commercial solutions and less ergonomic alternatives by offering a wearable, autonomous, and accessible system. The prototype was built using the ESP-01 microcontroller, MPU6050 inertial sensor, and a piezoelectric buzzer, powered by two 18650 batteries regulated by an AMS1117. The embedded logic performs automatic calibration and detects falls based on three criteria: reduced acceleration (free fall), sudden impact, and variation in arm inclination. The physical structure was 3D-printed and attached to the wrist with velcro, aiming for comfort and practicality. Controlled tests were conducted with simulations of falls in different directions and common daily activities. The results showed a high detection rate for real falls and a low occurrence of false positives, validating the effectiveness of the three-stage algorithm. The system’s autonomy reached approximately 8 hours of continuous use, limited by thermal instability of the AMS1117 regulator rather than battery depletion. It is concluded that the proposed system is technically feasible and performs well in simulated scenarios. Future work includes improving energy efficiency and integrating communication modules for remote alert transmission. The proposal contributes to the advancement of accessible assistive technologies with the potential to positively impact the safety and autonomy of vulnerable populations.