Robust Simultaneous Localization and Mapping Using the Relative Pose Estimation of Trilateration UHF RFID Tags

We consider a mobile robot equipped with wheel encoders and a RFID reader, which measures the phase of the signal backscattered by a set of passive UHF-RFID tags, deployed in unknown position on the ceiling of the environment. We deal with a Simultaneous Localization And Mapping (SLAM) problem, where the position of the tags must be estimated to create a reference map, within which the robot will be localized. One of the contributions of the paper is the use of a special kind of tag, the TriLateration Tag (TLT), including three antennas close one each other. The solution is based on the range and bearing estimation of all TLT detected, performed through a set of Multi-Hypothesis Extended Kalman Filters (MHEKF), one for each TLT. Then, the range and the bearing information of the responding TLTs is used in an EKF-SLAM algorithm which solves the SLAM problem. The proposed approach is more robust and computationally efficient with respect to other approaches available in the literature and is particularly suited for large warehouses where RFID tags cannot be deployed too densely.