Digital Twins for Aircraft Structural Inspections: Enhancing Dent Detection

Abstract

This paper outlines the development of Dent and Buckle Digital Twins (D&B DT) for aircraft structures aimed at enhancing the accuracy and efficiency of damage inspections. By utilising 2D and 3D data sources from different scanning technologies, detailed digital models of aircraft structures are generated for the damage analysis. Key features in this methodology include the development of dent detection algorithms, localisation, and assessment of dents as well as the creation of D&B DTs. Integrating computer vision techniques enables automatic highlighting of dents and missing rivets, narrowing down the region of interest. Commercial-Off-The-Shelf (COTS) software offers a proprietary algorithm for the damage characterization that works with proprietary 3D structured light scanners. This study addresses this constraint by proposing a solution using open-source software for dent detection able to use more efficient, less labour-intensive scanning outputs. The Digital Twin (DT) functions as a digital representative and its capacity for multi-source data fusion overcomes the deficiencies of single technology analysis. A necessity for this is the identification of meta data for the DT assembly. The presented case study includes a detailed parameter extraction, automated dent highlighting, and DT generation for an aircraft wing structure with impact damages.