AI-BASED SMART COATING DEGRADATION DETECTION FOR OFFSHORE STRUCTURES

Abstract

This study presents a comprehensive systematic review of advanced artificial intelligence (AI)-based approaches for smart coating degradation detection in offshore structures, with a particular focus on real-time sensor fusion, machine learning models, digital twin integration, and simulation-assisted analytics. Given the harsh marine environments in which offshore infrastructure operates—exposed to salinity, humidity, UV radiation, and mechanical stress—traditional coating inspection methods such as visual assessments and manual testing often fall short in detecting early-stage corrosion and subsurface anomalies. As a result, there has been a growing body of research leveraging AI-driven technologies to automate and enhance the accuracy, speed, and reliability of corrosion detection. Following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines, this study systematically reviewed and synthesized findings from 76 peer-reviewed articles published across major databases including Scopus, Web of Science, IEEE Xplore, and ScienceDirect. The review reveals that Convolutional Neural Networks (CNNs) are widely adopted for image-based surface inspection tasks, offering superior performance in detecting rust, blistering, cracking, and delamination. Time-series models, particularly Long Short-Term Memory (LSTM) networks, are effectively used to forecast degradation trends based on continuous sensor inputs. Sensor fusion strategies—combining data from visual, acoustic, thermal, and chemical sensors—further improve detection reliability, especially in dynamic offshore environments where single-sensor systems are prone to errors. The integration of digital twin technology enables real-time simulation and virtual monitoring of coating performance, while simulation-assisted learning allows the generation of synthetic datasets to overcome the challenge of limited field data. Despite these advancements, challenges such as energy efficiency, data synchronization, sensor drift, and environmental noise persist and need to be addressed for large-scale implementation. The findings of this study collectively highlight the potential of AI-enhanced monitoring frameworks in transforming traditional corrosion inspection methods into predictive, intelligent, and automated systems tailored for the complex demands of offshore infrastructure.