Addressing RFID Misreadings to Better Infer Bee Hive Activity

Abstract

This paper proposes a method to address misreadings and consequent inadequacy of radio-frequency identification data for social insect monitoring. Six-month worth field experiment data were collected to demonstrate the application of the method. The data are transformed into a linear combination of the Gaussian model and curve-fitted using an evolutionary algorithm. This results show that the proposed method allows us to improve the quality of data that infer honey bee behavior at the colony level.