Assessing exposure to slip, trip, and fall hazards based on abnormal gait patterns predicted from confidence interval estimation

Abstract

Monitoring workers' exposures to slip, trip, and fall (STF) hazards is critical to preventing STFs at construction sites. This study developed a model to assess workers' exposures to STF hazards by predicting abnormal gait patterns from a series of steps. The model was then evaluated and validated through a field experiment. Gait variability features were extracted from a waist-worn inertial measurement unit (IMU) and converted into Mahalanobis distance. Bidirectional long short-term memory models were used to predict abnormal gait patterns using confidence interval estimation. The model generated an Unweighted Average Recall (UAR) of 93.0% (normal walking: 93.0% and exposure to STF hazards: 93.0%), which demonstrates that workers' exposures to STF hazards can be continuously and remotely monitored, potentially helping to prevent STFs on construction worksites.