Data-Driven Quality Assurance of the Dry Fibre Tape Laying Process

<p dir="ltr">This dissertation addresses the high labour intensity of carbon fibre composite production in the aerospace sector by advancing the Automated Dry Fibre Tape Placement process. It introduces a comprehensive quality definition by examining how gaps, overlaps, and undulations affect the material's mechanical properties. Furthermore, a real-time monitoring system employing laser triangulation and deep learning evaluation for accurate defect detection was developed to enable immediate corrective feedback. A novel training method for the neural network-based evaluation algorithm reduces data requirements while maintaining high performance. These innovations support cost-efficient, high-quality manufacturing of lightweight composite materials.</p>