Statistical Analysis on Readability of RFID Gen2 Passive Tags Using Bayesian Information Criterion and 2-Level Factorial Design

RFID holds a large potential for changing how a process is manipulated and controlled. RFID is a valuable improvement in many organizations, whether it is used for monitoring logistics and supply chains, securing access within the facility to locate and validate inventory. Most of the time, the main technical concern with RFID is whether the tags can be read in the environment they are being incorporated in. This paper examines the effect of different factors such as distance from an antenna, tag size, the surface on which tag is placed, tag location & velocity/speed on the readability of the RFID tag. In this paper, we developed an effective model analysis of the factors that affect the readability of the RFID tag using the Bayesian Information Criterion (BIC) & Design of Experiment (DOE) technique. Experiments were conducted and readings were taken to consider the effect of different factors and identify the critical factors that affect the readability, and a redundant regression model was developed using BIC. Design of Experiments (DOE) was used, specifically 2k factorial design and 2(k-1) fractional factorial design to perform statistical analysis and understand the source of variation in the model. The two designs were compared and the best out of the two was chosen to develop a model. Statistical analysis was performed using Minitab software to validate. The developed model intends to read the intensity of the tag with 97% accuracy.