Abstract
Missteps are a common cause of falls from ladders, but the contributing factors are unclear. Destabilizing events like tripping on a ladder rung could result in falls. Movement quality, quantified by stability, smoothness, or symmetry is known to affect fall risk in other contexts and may be relevant to ladder trips. This work generated predictive models for rung tripping events using accelerometer data from 25 individuals, which were collected while they climbed a ladder. Two moderately predictive models were created: one that predicted a person’s likelihood of tripping on a rung from both foot and pelvis accelerometry data (area under the curve or AUC = 0.71) and one that predicted someone’s likelihood of tripping from only pelvis data (AUC = 0.73). These results indicate that a person’s climbing mechanics, particularly at the pelvis, are associated with their likelihood of experiencing a trip. These models, intended for deployment in the field as a tool for training and monitoring misstep risk, could help reduce the frequency of ladder falls.