Development of Digital Twins for Extrusion Lines of PVC Tubes

The increasing demand for plastic products requires innovative methodologies to enhance efficiency and reduce the environmental impact. This research explores the integration of Artificial Intelligence (AI) and Industry 4.0 technologies, emphasizing developing and deploying Digital Twins (DT) for real-time process optimization in the extrusion of flexible PVC tubes. This study incorporates real-time data acquisition through Industrial Internet of Things (IIoT) devices and machine learning (ML) algorithms to predict and dynamically control process parameters. Central to the research is the novel application of a DT, which serves as a reflective model of the physical extrusion process and a predictive tool, enhancing operational efficiencies and reducing material wastage. Extensive experimental trials were conducted under varied operational conditions to validate the efficacy of the proposed approach. By implementing the DT, the system can provide a real-time estimation of the most critical quality indexes of the tube (Soft-Sensor-based DT) and optimize operational parameters to reduce the startup time of the machines and ensure product quality. Image processing techniques were used to acquire data for training the ML regression models, reducing the need for manually collecting experimental data. Novel insights into the dependency of energy usage on extrusion parameters and material properties facilitated the development of a digital model for monitoring and optimizing power consumption during extrusion. Furthermore, by integrating traditional LCA methodology with the DT framework, a new approach was proposed for assessing the environmental impact of the process. The results indicate significant improvements in the sustainability of the production process, reducing environmental impact and energy consumption. The research highlights the transformative potential of ML and DT technologies in advancing the plastic extrusion industry by providing deeper process insights and encouraging a shift toward autonomous systems. This work advances theoretical understanding and demonstrates practical applications that can lead to substantial economic and environmental benefits. Future work will focus on scaling the solution across different production lines and exploring its applicability in other manufacturing sectors.