<p>Injection molding is a widely adopted manufacturing process for producing plastic components with high precision and efficiency. However, real-time monitoring and management of defects remain critical challenges. This study integrates Digital Twin (DT) technology into the part separator mechanism of injection molding machines, providing a dynamic and data-driven solution for defect detection and process optimization. By leveraging real-time data acquisition and machine learning algorithms, the proposed DT model correlates operational parameters such as injection temperature and cycle time with defect rates to predict and minimize defects effectively. The integration of advanced predictive analytics enables the identification of high-risk operational conditions, allowing for proactive adjustments and enhanced quality control. Experimental results demonstrate a 92% accuracy in defect prediction, with significant reductions in waste and improved production efficiency. The proposed approach not only addresses existing gaps in injection molding quality assurance but also contributes to the broader adoption of Industry 4.0 technologies in manufacturing. Future directions include refining the predictive model and expanding DT applications to encompass the entire injection molding lifecycle.</p> Graphical abstract <p></p>