Precise Location Tracking of Livestock using RFID Phase and Nonlinear Filtering

Ultra-high frequency (UHF) RFID offers longer read range and better immunity to interference when compared to low frequency technology for livestock identification. However, the longer UHF read range can create an over-range problem by reading tags in local "hot spots" well outside of the intended read zone. With UHF it is not enough to know that a tag has been read, but you must know where the tag is. This application proposes a new particle filter for precise tracking of UHF RFID tags. The computationally efficient algorithm may achieve tracking accuracy on the order of one inch for animals using standard RFID readers and passive tags, but we must first remove phase effects due to the reader and tag. We will validate a calibration procedure to remove these nuisance parameters based on a theoretical model of the tag and reader, then we will experimentally measure the algorithm's tracking performance at Clairvoyant's lab and at a beef research facility at Kansas State University.The anticipated result is a proof of concept tracking algorithm implemented offline on a computer. The commercial applications are substantial, enabling a cost effective eID retrofit for auction markets and dairies without slowing down or changing the flow of livestock. Successful implementation creates new opportunities in many industries, including cross-read tag filters in logistics and retail, vehicle positioning in free flow tolling, simultaneous location and mapping algorithms in robotic systems, factory automation, baggage handling systems, and early detection of sick animals in livestock health monitoring systems.