A framework for planning underwater hull inspections based on computer vision and degradation assessment

Traditional underwater hull inspections, which rely heavily on human divers, face significant challenges such as safety hazards and dependency on expert judgment. Recent technological advancements, such as remotely operated vehicles and artificial intelligence, offer promising alternatives to address these limitations and enhance the efficiency and safety of underwater inspections. This paper proposes a framework for planning underwater hull inspections based on computer vision and degradation assessment. The proposed modeling includes three main processes: Detect degradation, Assess degradation, and Perform maintenance decision-making. The degradation detection process utilizes Convolutional Neural Networks for computer vision to identify structural degradations such as corrosion and cracks through automatic image analysis. Then, the degradation assessment process assesses the hull degradation based on measurements such as material loss to provide a comprehensive understanding of structural integrity. Finally, the maintenance decision-making process guides the decision on maintenance tasks based on the Remaining Useful Life estimates. Through a case study, the proposed framework was demonstrated considering the operational context of Floating Production Storage and Offloading (FPSO). As a result, the proposed framework showed to be consistent in identifying different types of structural degradations based on a U-Net architecture and supporting underwater hull inspection planning.