Integrating machine learning and material feeding systems for competitive advantage in manufacturing

Abstract

In contemporary business environments, manufacturing companies must continuously enhance their performance to ensure competitiveness. Material feeding systems are of pivotal importance in the optimization of productivity, with attendant improvements in quality, reduction of costs, and minimization of delivery times. This study investigates the selection of material feeding methods, including Kanban, line-storage, call-out, and kitting systems, within a manufacturing company. The research employs six machine learning (ML) algorithms—logistic regression (LR), decision trees (DT), random forest (RF), support vector machines (SVM), K-nearest neighbors (K-NN), and artificial neural networks (ANN)—to develop a multi-class classification model for material feeding system selection. Utilizing a dataset comprising 2221 materials and an 8-fold cross-validation technique, the ANN model exhibits superior performance across all evaluation metrics. Shapley values analysis is employed to elucidate the influence of pivotal input parameters within the selection process for material feeding systems. This research provides a comprehensive framework for material feeding system selection, integrating advanced ML models with practical manufacturing insights. This study makes a significant contribution to the field by enhancing decision-making processes, optimizing resource utilization, and establishing the foundation for future studies on adaptive and scalable material feeding strategies in dynamic industrial environments.