Automatic prediction of maintenance intervention types in transport linear infrastructures using machine learning

This article presents a methodology to automate the prediction of maintenance intervention alerts in transport linear infrastructures. It combines current and predicted asset conditions with operational and historical maintenance data, to predict the needed tasks to avoid later severe degradation. By means of data analytics and machine learning models, a prioritised listing ranked on severity level corresponding to the alerts generated for all assets of the infrastructure is inferred. The scientific part presents: a discussion on relevant data to train machine learning algorithms in order to generate reliable predictions of the interventions to be carried out in further time scenarios, a schematic flow chart of the automatic learning procedure, and the self-learning rules from automatic learning from false positive/negatives. The empirical part describes a road network pilot case, the available historical data information, measurements, maintenance interventions, and a selected set of outcomes.