An Integrated Framework for Predictive Quality in Injection Molding: Combining Explainable AI and Time Series Analysis in a German Industry Case Study

In the era of Industry 4.0 and the emerging vision of Industry 5.0, ensuring consistent product quality is crucial, particularly in complex manufacturing processes like injection molding. This study integrates Explainable AI (XAI) with time series analysis using real-world data from a German injection molding facility to enhance predictive accuracy and process interpretability. Results demonstrate that combining explainability techniques, such as SHAP, with time series features improves model performance, reducing the Mean Squared Error (MSE) from 0.01025 to 0.00251 and increasing the R-squared from 0.9886 to 0.9972, while revealing hidden patterns in process dynamics Global SHAP analysis identified key factors influencing quality, while local SHAP insights highlighted the role of setting parameters those directly adjustable by operators in mitigating deviations. Time series analysis further enhanced decision-making by enabling proactive interventions before process fluctuations compromised stability. By structuring decision-making into key steps identifying influential parameters, prioritizing adjustable ones, and incorporating temporal insights this study provides a roadmap for integrating XAI into quality control. The findings reinforce the value of human-centric AI, ensuring transparency and empowering operators to optimize industrial processes effectively. Copyright (C) 2020 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)