Ubiquitous Fall Hazard Identification With Smart Insole.

Falls are leading causes of nonfatal injuries in workplaces which lead to substantial injury and economic consequences. To help avoid fall injuries, safety managers usually need to inspect working areas routinely. However, it is difficult for a limited number of safety managers to inspect fall hazards instantly especially in large workplaces. To address this problem, a novel fall hazard identification method is proposed in this paper which makes it possible for all workers to report the potential hazards automatically. This method is based on the fact that people use different gaits to get across different floor surfaces. Through analyzing gait patterns, potential fall hazards could be identified automatically. In this research, Smart Insole, an insole shaped wearable system for gait analysis, was applied to measure gait patterns for fall hazard identification. Slips and trips are the focus of this study since they are two main causes of falls in workplaces. Five effective gait features were extracted to train a Support Vector Machine (SVM) model for recognizing slip hazard, trip hazard, and safe floor surfaces. Experiment results showed that fall hazards could be recognized with high accuracy (98.1%).