Abstract:
The nature of complex interactions between machine, material and process parameters frequently leads to quality inconsistencies in injection molding processes. The present paper introduces an autonomous quality control framework based on AI and having real-time data on cavity pressure and temperature sensors of industrial injection molding cycles. Process data in large quantities are used to train learning models which learn to detect quality-sensitive cavity conditions and predict deviations which cause defective products. The system dynamically modifies the input parameters to the machine based on a real-time comparison with the learned reference patterns to stabilize in-cavity conditions. The feasibility of the proposed approach shows how AI-based control can help to ensure the stability of the quality of the products in the case of industrial injection molding.
Keywords: