RFID ZONE SENSOR POSITIONING AND CAD DIGITAL DOUBLE SYNCHRONIZATION ACCURACY USING A KNN SUPERVISED MACHINE LEARNING ALGORITHM - A DESIGN SCIENCE STUDY Curtis Shull, DCS Curtis.Shull@Proxigroup.com Proximus Software Systems Inc, DBA Proxigroup, ProxiLabs LLC, 901 Broadway # 24210 Nashville, TN 37202-3990 USA Samuel Sambasivam, PhD Samuel.Sambasivam@Woodbury.edu Computer Science Data Analytics Woodbury University, Burbank, CA, USA ABSTRACT Aim/Purpose The problem that was addressed was that existing RFID software technology presents numerous challenges regarding automated accuracy, efficiency, and ac- curate coverage, affecting the total cost of ownership (TCO) regarding purchas- ing RFID asset tracking solutions. The core problem was that no automated paradigm exists to determine an accurate Received Signal Strength Indicator (RSSI) that presents signal strength based on the RFID tag coordinates and RFID reader distance in a dynamic 3D digital double infrastructure. Background The research question that drives the study is, what is the accuracy of a CAD- based infrastructure digital double model and RFID sensor location geometry derived from XML? The research looks into the precision of the physical ge- ometry of a digital double created using computer-aided design (CAD) within the context of Cartesian XML data-driven positioning and visualization of RFID antennae and zones to expand upon existing results. Methodology The research used a quantitative approach within the domain of design science to provide a theoretical framework for problems stemming from multiple data sets in computer-aided design (CAD) and radio frequency identification (RFID) technology. Mathematical analysis was applied to the DSR study of RFID syn- chronization with CAD. This structure incorporates the conceptual groundwork for CAD RFID studies. The notions that form the basis of the conceptual framework are drawn from the relevant literature and the results of prior stud- ies. The study is void of research samples.