International Journal of Advanced Research in Computer and Communication Engineering

Abstract: An Internet of Things (IoT)-driven quality control mechanism within the manufacturing sector commences with the visualization of item movements throughout production lines, aimed at identifying critical stations and pathways that have a substantial impact on product quality. This preliminary phase yields valuable insights into the production flow and highlights potential bottlenecks or opportunities for enhancement. Feature Engineering is employed to derive pertinent information through the selection and transformation of features, thereby augmenting the efficacy of machine learning models. The performance of the model is evaluated in comparison to other classification methodologies, such as Support Vector Machine, Naive Bayes, Random Forest, and Gradient Boosting, predicated on the chosen features. Through the examination of the interrelations among features, stations, lines, and the response variable, a deeper comprehension of the most influential factors affecting product quality and defect occurrence is attained. By leveraging visualizations of production line movements, feature importance rankings, and classifier performance metrics, this IoT-driven framework furnishes actionable insights for manufacturers to enhance product quality and mitigate defects.

Keywords: Internet of Things, Sensors, Machine Learning, Classification, Manufacturing.

Downloads: | DOI: 10.17148/IJARCCE.2025.14833