Minimizing Temperature Deviations in Rubber Extrusion Lines using Artificial Neural Network based Process Control

Rubber extrusion is a complex manufacturing process that poses challenges for process control due to the high number of control variables including extrusion parameter settings, rheological behaviour, compound viscosity and batch-dependent material variations. Already small deviations from the control variables can either influence the throughput stability or result in decreased surface quality of the extrudate, leading to increased scrap rates. To address these challenges, this paper presents an Artificial Intelligence (AI)-based approach for autonomous control of process parameters in rubber extrusion lines. The proposed method utilizes Artificial Neural Networks to enable early identification and prediction of batchspecific temperature deviations, allowing the system to adapt and improve with each new application. To maintain specific tolerance limits, the system leverages data mining techniques to identify cause-and-effect relationships between process parameters and product characteristics. The AI model calculates control variables based on specified tolerance limits and measurement variables as input, which are then validated by the data mining model and forwarded to a hardware interface to counteract deviations. The proposed method was evaluated on a rubber extrusion line in a real-world manufacturing setting. The results demonstrate the potential of this approach to maintain specific tolerance limits in rubber extrusion by identifying and addressing batch-specific temperature deviations, thereby enabling a more sustainable process with less scrap rates. This paper presents the system architecture of the AI-based process control for rubber extrusion and provides a comprehensive overview of the experimental results.