Designing and Evaluating an Avatar for On-Person Screening

Abstract

In aviation security, avatars are generic human figures that are used to display alarms provided by on-person screening systems. One critical feature of these avatars is that they provide no body detail unique to an individual traveler. However, the generic nature of these avatars leaves few landmarks that can be used to map the location of an alarm on the avatar to a passenger. We manipulated two features of an avatar, body detail and grid lines, to create 6 avatars to investigate how design influences estimation of target location. Body detail was manipulated at three levels: no joints, some joints, and direct outline of the passenger. Grid lines were manipulated at two levels: grid lines or no grid lines. The results of the study showed that security screeners were nearly 20% closer to the true target location when the avatar featured landmarks that can be found on a typical passenger.