Due to the low-cost and contactless way of communication, radio-frequency identification (RFID) technology provides a solution to overcome the difficulties (e.g. occlusions) that the traditional line of sight sensors (e.g. cameras and laser range finders) face. In this thesis, we address the applications of using passive ultra high frequency (UHF) RFID as a sensing technology for mobile robots in three fundamental tasks, namely mapping, path following, and tracking.
In this thesis, we address the problem of recovering from mapping failures of static RFID tags and localizing non-static RFID tags which do not move frequently using a particle filter. The usefulness of negative information (e.g. non-detections) is also examined in the context of mapping RFID tags. Moreover, we present a novel three dimensional (3D) sensor model to improve the mapping accuracy of RFID tags. Second, we present a novel approach that combines RFID fingerprints and odometry information as an input of the motion control of a mobile robot for the purpose of path following in unknown environments. Last, we address the problem of tracking dynamic objects for mobile robots using RFID tags. To achieve this, we combine a two stage dynamic motion model with the dual particle filter, to capture the dynamic motion of the object and to quickly recover from failures in tracking. The state estimation from the particle filter is used in a combination with the VFH+ (Vector Field Histogram), which serves as a local path planner for obstacle avoidance, to guide the robot towards the target.