risk assessment of dropped objects on corroded submarine pipelines using machine learning algorithms

This paper proposes a probabilistic model based on machine learning algorithms to estimate the risk associated with different levels of damage (per dnv-rp-f101) due to a dropped-object impact on subsea pipelines. the model is generalized by considering a wide range of pipeline geometric and mechanical specifications, corrosion conditions, and various possible impact scenarios. multiple machine learning algorithms—including linear regression, decision tree, random forest, k-nearest neighbors, support vector machine, and gradient boosting—were evaluated, with random forest demonstrating the highest accuracy. the analysis of how pipeline characteristics influence the probability of different damage levels provides a basis for decision-making on implementing preventive measures to reduce damage probability during the pipeline design stage