Fall detection system for elderly by MEMS accelerometer and SMS alert

Abstract

A major problem in the elderly are the negative effects of prolonged unattended stay after a fall. The proposed solution helps to reduce the negative consequences of falls through a portable noninvasive low-cost electronic system. The system, properly attached to the person, has the capacity to detect falls using MEMS accelerometers and generate an alert by sending a short text message through the public cellular network. The project includes the full design and the implementation of a biomedical portable prototype equipment using a reliable Atmel microcontroller, a medical-grade Bluetooth module and functional tests to evaluate the global effectiveness. The results were excellent, with sensitivity of 1 and specificity of 0.77 in several typical fall and not-fall sceneries.