Quality Prediction in Injection Molding Using Machine Learning Methods

In order to achieve high efficiency in the plastic injection molding (PIM) process and ensure exceptional service quality for customers, it is imperative to identify the variables that influence the final product's quality and continuously monitor the intricate and dynamic changes these variables undergo during the PIM process. Furthermore, attaining the desired quality level necessitates the development and implementation of intelligent systems harnessing Artificial Intelligence (AI) and Machine Learning (ML). These systems have the capability to detect and even anticipate defects, shifting the paradigm from reactive actions to preventive measures. Consequently, we can intervene proactively before producing defective components, thereby reducing waste generation.In line with these considerations, this article presents a study on the feasibility of utilizing artificial neural networks (ANN), specifically the multilayer perceptron (MLP), as well as k-nearest neighbors (KNN) and support vector machines (SVM), to formulate quality prediction models. The accuracy of these models is compared, using data collected from industrial operations conducted within the Tangier free zone in Morocco.