Dynamic Balance Measures and Sensing for Humans in Fall Detection and Prevention

Abstract

Balance Aid aims to enhance balance by detecting unstable postures and providing feedback to prevent a fall. This thesis presents the development of a shoe sensor, a method to detect anomalous gait behaviours, and a measure of dynamic balance.Ground reaction sensors must avoid disturbing natural gait. A lightweight, compliant, wearable shoe sensor was developed that measures ground reaction in real-time and can be used in fall detection and prevention. Falls can be prevented by reducing risk factors. An anomalous gait behaviour detection system was developed to detect fatigue, eyes-closed, and cluttered gait using force plates and the ground reaction sensors. In addition to reducing risky behaviours. Traditional measures of balance do not apply well to human gait and detect steps as falls. An angular momentum based measurement was found to be a dynamic measure of balance that is able to differentiate between a fall and a step.iii