Detecting Elevators and Escalators in 3D Pedestrian Indoor Navigation

Abstract

In the field of indoor navigation, inertial measurement units (IMUs) are commonly used to track pedestrians. Unfortunately, in conjunction with dead reckoning localization approaches, the position accuracy is degraded by the accumulation of sensor errors. Most pedestrian dead reckoning systems are therefore adapted to human gait patterns to constraint this error growth. These interventions however make them ignore any user displacements different from doing steps, such as by elevators and escalators, and as a result they fail to track the full user trajectory indoors. In order to avoid such estimation failures, an elevator and escalator detection algorithm is proposed that is capable to detect such incidences in order to enhance the position estimation in an elevator/escalator ride.