Developing a predictive model using multivariate analysis and Long Short-Term Memory (LSTM) to assess corrosion degradation in mining pipeline thickness.

Pipeline corrosion has significant impacts on the human,economic, and natural environment. To help betterdetect and prevent it over time, in this paper, we proposea multivariate approach using machine learning.More precisely, we propose to study the evolution ofthe thickness of the mining pipeline using a multivariateapproach and to implement a predictive model usingthe Long Short-Term Memory (LSTM) artificial neuralnetwork. Indeed, LSTM is a specific recurrent neuralnetwork (RNN) architecture designed to model temporalsequences. The proposed predictive model achievedan accuracy of 80% and a loss of 0.01 and was able topredict variations in eight thickness measurements overone hundred days.