The radio frequency id ent i ficat i on (RFID) tec h nology has been studied and used to solve the localization and mapping problems in which most of studies use long-range active or passive RFID systems with multiple readers and dense tags. In this paper, we propose to use a short-range passive RFID reader to accomplish extended Kalman filter (EKF) based simultaneous localization and mapping (SLAM) with sparse tags in large indoor environments. We demonstrate that it is feasible to accomplish SLAM with sufficient accuracy using only odometry and short-range RFID data. The simulation and experimental results demonstrate the feasibility of the proposed system.
I. INTRODU CTIO NThe radio frequency identification (RFID) technology has been applied in many applications such as storage man agement, personal guarding and public transporting system. The RFID applications are flexible and feasible because of RFID's fast detection capability and contactless identification with perfect data association. As data association is one of the most challenging problems in robot perception, the RFID technology has been used to solve a number of daunting robot perception problems such as mapping, localization and activity recognition. With a predefined or calibrated map of RFID tags, localization can be performed [10] [8] [1]. Hahnel et al. [4] and Joho et al. [6] proposed to learn the map of the RFID tags with perfect localization, and then use the built map to perform localization in which mapping and localization are solved separately in the two stages. RFID is also used and integrated with other sensors. In [3], data from a camera and RFID mounted on a pan-tilt unit of a robot are fused to accomplish person tracking where a RFID tagged person can be followed by the robot. As RFID tags also have a limited data storage capability, RFID can be used to deliver information in tasks such as activity recognition [11], search and rescue [7], and human robot interaction [9].In this work, a short-range passive RFID system is used to accomplish extended Kalman filter (EKF) based simultane ous localization and mapping (SLAM). Fig. 1 (a) shows the NTU-PAL4 robot in which the RFID reader was mounted on the right side of the robot and the SICK laser scanners were used to collect ground truth data. Fig. 1 (b) shows the RFID system used in this work. The RFID system is the RWM600 This work was partially supported by grants fr om Taiwan NSC ((a) The NTU-PAL4 robot.(b) The short-range passive RFID reader and tag. Fig. 1. The RFID system is mounted on the right side of the NTU-PAL4 robot. The SICK laser scanners on the top of the robot were used to collect ground truth data.module for ISOIIEC 15693 spec working on 13.56MHz channel with 5Hz reading frequency and the sensible range is only 0.1 meter. As the RFID devices in this work only have a centimeter level of detection range, only odometry data are available most of the time in the situations that RFID tags are deployed sparsely in environments. A new featurellandmark is added to the SLAM state...