A MIMO UHF-RFID SAR 3D Locating System for Autonomous Inventory Robots

As ultrahigh frequency (UHF) radio-frequency identification (RFID) is a cost-efficient, reliable technology, its use for Internet of Things applications is very attractive. The goal of three-dimensional (3D) product maps for smart warehouses is very challenging, but can be achieved with UHF-RFID. Therefore, we equipped a mobile robot platform with a commercial RFID reader and added a switching matrix to integrate a multiple input-multiple output (MIMO) synthetic aperture radar based localization concept (SARFID) with the use of eight antennas. The mobile robot platform and the MIMO localization technique together demonstrate good tag localization results. We evaluate both single-dimensional and two-dimensional synthetic aperture radar (SAR) trajectories. Their shapes are the main contributor to the achievable accuracy. The robot position is acquired by the robot's odometry data. Despite the substantial error in the robot position, we generate 3D product maps with tags labeled on objects and achieve a 3D root mean square error (RMSE) of 3.8 cm.