Increasing accuracy RFID Dock Door Discrimination with Naive Bayes Classifier
B.W. Berenschot (TU Delft - Mechanical Engineering)
Mark Duinkerken – Mentor (TU Delft - Transport Engineering and Logistics)
R. R. Negenborn – Graduation committee member (TU Delft - Transport Engineering and Logistics)
Walter Romijn – Mentor (Mieloo & Alexander)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
Errors during truck loading at dock doors lead to unwanted wrong deliveries in logistics. Due to the falling price of RFID tags, RFID dock door discrimination is now being used for product registration. The problems that arise with the current system for RFID Dock Door Discrimination are cross-reads and miss-reads. The purpose of this research is to increase the accuracy of product registration by the proposed Dock Door Discrimination method with Naive Bayes Classifier (NBC). A hardware design, including 4 RFID antennas at three adjacent dock doors and 1 added antenna at the staging area, and software design, including the implementation of the NBC, are proposed to improve the RFID Dock Door Discrimination. The Experimental Setup and Plan were used to gather data to compare the current RFID transition system with the new proposed NBC system in six scenarios, three with and without noise at other gates. For each individual scenario, the accuracy improved most with NBC with one input feature. The accuracy for all scenarios combined for the collected data improved from 82.1$ (current) to 93.6(NBC).These results mean that there is a solid improvement in implementing the Naive Bayes Classifier over the current RFID transition system for Dock Door Discrimination.