Designing sensor systems capable of differentiating children from adults

Abstract

Introduction Injury prevention systems intended to prevent children from entering hazardous locations (or at least alert caregivers if that occurs) often respond to every instance of a person's presence, regardless of whether the intruder is a child. This performance results in a high nuisance alarm rate that sometimes causes adults to disable or circumvent the safety system. If a child safety system can accurately identify intruders as adults or children, nuisance alarm rates can be decreased. Method This analysis selects three human factors (height, foot length, and cognition) amenable to adult/child differentiation and describes likely sensor strategies, advantages, and disadvantages. Results Preliminary testing of prototypes systems shows that simple sensor systems are capable of acquiring adequate data for adult/child differentiation. The discussion addresses requirements for discriminator systems and the effects of various sensor combinations on overall performance.