Developing an intelligent library gate to detect unauthorized borrowing and to minimize electrical power consumption

Recently, many libraries have introduced a UHF-band RFID tag system because of its convenience. However, the UHF-band RFID reader has wide coverage area and this can reduce the management efficiency by unintentionally reading unnecessary tags. In this paper, we propose an intelligent library gate by using machine learning algorithm. The intelligent gate takes into account relative angle and distance between RFID tags and the reader and controls the intensity of the radio wave emitted from the reader by using machine learning algorithm. This enables the reader to maintain detection accuracy and minimize electrical power consumption. In order to validate the effectiveness of the gate, SVM is used to predict RSSI. The result of the prediction shows RMSE as 7.67 and R-squared as 0.77. Electrical power consumption is reduced to 59 % compared to a case when the radio wave is emitted at maximum intensity. The result shows that the intelligent gate which maintains detection accuracy and minimize electrical power consumption can be developed by selecting minimum necessary radio wave intensity predicted in the experiment.