Quantitative evaluation of the reliability of hybrid corrosion inspection and monitoring approaches

Conventional ultrasonic approaches for corrosion surveillance such as large-scale C-scans and continuous localised monitoring face inherent limitations in either temporal resolution or spatial coverage. Recent advances in sensor technology and low-cost autonomous robotics enables a hybrid approach that combine their strengths. However, critical questions remain about optimising such approaches, including determining the optimal inspection intervals, spatial coverage, and number of sensors. This study presents a multi-stage framework that addresses these challenges through integrated modelling: degradation simulation for spatiotemporal evolution of component surfaces, physics-based surrogate models for ultrasonic thickness measurements, data subsampling to replicate inspection and monitoring procedures, and reliability assessments against surveillance targets. Through Monte Carlo simulations on synthesised data calibrated to field measurements, we demonstrate the potential benefits of the hybrid strategy. Results show that strategic sensor placement and periodic repositioning can enhance reliability while reducing coverage requirements and extending inspection intervals. Importantly, this framework provides quantitative guidance for corrosion surveillance planning and can be adapted to different degradation phenomena across various asset types.