An algorithm for automatic grasping an UHF RFID passive tag

This paper proposes an automatic system which allows a mobile robot to reach a tagged object in an unknown environment, exploiting the Radio Frequency IDentification (RFID) technology. It is assumed that the robot has an odometry system (e.g., wheel encoders) and an RFID reader which provides it with phase-shift measurements of the RFID signal backscattered by the tagged object. Since phases are ambiguous measurements of the tag-reader distance, a Multi-Hypothesis Extended Kalman Filter is proposed to on-line extract, by fusing phase measurements with odometry readings, the range and the bearing of the unknown tag. Once the relative position of the tag is known, a control algorithm is applied to drive the robot toward the desired destination. Numerical results are reported to illustrate the approach. © Published under licence by IOP Publishing Ltd.