Improving RFID tag Reads

Radio Frequency Identification (RFID) is used in various ways for the optimization of today’s supply chains. One problem in dense reader environments are false-positive RFID tag reads i.e., tags that have been read unintentionally by an RFID reader. This project proposes to apply low-level readers such as Received Signal Strength Indication, time stamps of reads and used antennae. An algorithm has been developed and verified in a productive RFID environment.
Improving RFID tag Reads


  • Machine learning based approach that makes use of the low-level reader data collected when reading tags to detectsuch false-positives
  • Evaluation in a productive RFID enabled distribution center with 7month of data describing 53,998 pallet events
  • Results: More than 95.5% of the real world data was classified correctly


Selected Publications