A Finite State Machine-Based Fall Detection Mechanism on Smartphones

This paper presents a detection mechanism that utilizes the accelerometer in a smart phone carried by an individual to measure the human movement and hence determine if a fall event has occurred. The model of the fall activities are characterized as a finite state machine, which transits from one state to another according to the data generated from the accelerometer. The presented detection mechanism utilizes the finite state machine to identify different types of falls, including forward falls, backward falls, and lateral falls. Experiments were conducted to evaluate the performance of the presented mechanism. The results show that the mechanism can effectively distinguish between actual fall events and normal activities such as squatting, and walking up and down stairs.