Stüven, OleOleStüvenGeiger, FelixFelixGeigerMoenck, KenoKenoMoenckSchüppstuhl, ThorstenThorstenSchüppstuhl2026-10-232026-10-232024-09MHI Kolloquium 2024https://hdl.handle.net/11420/65389In-line inspection and quality assurance are essential activities in manufacturing high-value and safety–critical products. In the scope of manufacturing aircraft fuselages, the high-quantity operation drilling is one of the critical steps, in which even slight improvements concerning quality or lot times are majorly impactful. Data recorded during drilling can be used to infer information about, e.g., the quality of the hole. Then, Artificial Intelligence (AI) methods can infer information about the quality of the process. More complex AI models have the advantage of more advanced predictive capabilities, which comes at the cost of being opaque even to experts. That hinders the deployment of the models, especially if the human must still be part of the process. Here, methods of Explainable AI (XAI) can augment AI predictions with explanations to enable humans to make more informed decisions. In this work, we propose a process framework that uses XAI to enable efficient in-line inspection of drill holes.dehttps://creativecommons.org/licenses/by/4.0/DrillingAircraft manufacturingIn-line inspectionQuality assuranceXAITechnology::629: Other Branches::629.1: Aviation::629.13: Aviation EngineeringTechnology::658: General Managament::658.5: Of ProductionComputer Science, Information and General Works::006: Special computer methods::006.3: Artificial IntelligenceConference Paper10.1007/978-3-031-91463-8_2110.15480/882.18956