A Survey on Autonomous Damage Detection on Aircraft Surfaces using Deep Learning Models

Abstract

The employment of contemporary technology in aircraft maintenance and breakdown operations has increased in the aviation sector in recent years, as the application area of artificial intelligence has expanded. Surface faults produced by dent, corrosion and cracks, and stains from oil spills, grease, and soil deposits, are all detected during an aircraft surface examination. The traditional way for evaluating aircraft skin is human visual inspection, which is time-consuming and wasteful, but robots with onboard vision systems can scan the aircraft’s skin safely, quickly, and correctly. In aviation maintenance, deep learning can be utilized to automate visual exams. This can help improve damage detection accuracy, save aircraft downtime, and prevent Inspection errors. This work uses deep learning to investigate the visual detectability of various types of in-service damage in laminated composite aviation structures, based on a review of more than five studies. The application of deep learning-based visual inspection systems to aviation maintenance and repair is examined. This discovery will facilitate future research and development in the field of autonomous visual inspection of composite structures, potentially lowering costs, eliminating health and safety concerns, and reducing downtime associated with integrity assessment.