A System for Predicting Unprecedented Injury by Spatiotemporally Superimposing Children’s Normal Behavior

Abstract

Predicting injuries in daily life is important in the field of product safety design and risk assessment. However, in the case of children, it is usually thought that unprecedented injuries are difficult to predict because they are caused by “irregular” child behavior. Despite the prevalence of this belief, this study proposes a new injury prediction system based on the view that unprecedented injuries can, in fact, be predicted by identifying high-risk combinations of “normal” behaviors and environmental states. In this article, we also propose an injury prediction system based on spatiotemporally superimposing normal child behavior. The proposed system enables us to consistently predict injury processes consisting of the situation leading to the injury, the impact occurrence, and the resulting injury. This paper also presents an example of a system application for predicting potential injuries around a swing set in an actual park. To prove the effectiveness of the proposed system, we compare the patterns of accident processes predicted by the system with those of actual incident processes found in our observations of normal behaviors.