Active-localization methods for mobile robots in a coarsely structured environment with floor-embedded RFID tags and indoor GPS

Abstract

Two active-localization methods for mobile robots, namely, “rhythm-based adaptive localization” for floor-embedded RFID tags and “real-time kinematic Doppler positioning” for indoor GPS (an indoor messaging system, or IMES), are proposed. Both methods increase the positioning accuracy of a robot by moving the sensor attached to the robot. The results of a positioning experiment with floor-embedded RFID tags show that the rhythm-based adaptive-localization method can achieve a positioning accuracy of 25 cm under the conditions that noise is intentionally added to the value of a wheel encoder, the installation interval of tags is 30 cm, and the tag installation is incomplete (i.e., some tags are missing). On the other hand, the results of a positioning experiment using IMES show that the real-time kinematic Doppler-positioning method can achieve centimeter to decimeter-level positioning accuracy even if the number of visible transmitters is only one (i.e., trilateration is not used).