Sequential Extended Kalman Filter for Simultaneous Ultrawideband Localization and Synchronization: Theory and Application

Autonomous transport vehicles require robust navigation solutions to operate seamlessly across diverse weather conditions. Traditional satellite-based navigation (GNSS), however, is limited under extreme conditions and does not work well without line-of-sight satellite connection. We addressed this problem by implementing a stationary ultrawideband (UWB) localization system to enhance data acquisition of vehicle position at docking ramps in a hub-to-hub material transport solution. We devised an algorithmic approach—based on a sequential extended Kalman filter (EKF) for simultaneous estimation of synchronization and localization states—to explore four configurations for tracking a vehicle’s 2-D pose (i.e., position and orientation). The filter processes each reception separately using univariate measurements, thus reducing computational complexity. We show the equivalence of this approach to multivariate processing. Our measurement results in a real-world logistics environment demonstrate that estimation accuracy is significantly enhanced by both geometrically interlocking the tags and incorporating a vehicle motion model into the localization algorithm.