Industrial Machine Data Generation and Artificial Optimisation for Blow Molding Extrusion Machines

This paper introduces a novel framework developed under the AIDEAS project for automating the generation of simulation data and optimizing blow molding extrusion machine designs using evolutionary algorithms. The framework comprises three core components: a Machine Data Generator (MDG) for synthesizing extensive simulation datasets augmented with real or historical process data, a Machine Design Optimizer (MDO) leveraging AI-based techniques for design optimisation, and a CAx plugin enabling seamless integration with common CAD tools. By integrating digital twin technology, computational fluid dynamics (CFD) and evolutionary algorithms, the system enables iterative refinement of complex machine configurations, such as spiral distributors for blow molding, to achieve optimal performance. Engineers can efficiently define, simulate, and visualize machine designs in real-time using user-friendly front-end components and cloud-based services. Preliminary results demonstrate significant reductions in iteration cycles by up to 75%, material waste up to 66% and manual tuning efforts in manufacturer trials. Future developments will expand parametric capabilities and adapt the solution to additional polymer types and manufacturing processes, enhancing its applicability across the plastics industry.