The goal is to demonstrate the potential of the use of machine learning techniques (a branch of artificial intelligence) for the predictive maintenance of rolling stock and in the railway sector in general. The phases defined are as follows:
o Development of predictive maintenance algorithms, in regard to the use case, based on a learning model that allows the prediction of the main parameters necessary for maintenance.
o Incorporation of available historical information of the use case, in a high-performance technological platform that allows the learning of the designed algorithms.
o Application of the resulting learning algorithms to the actual maintenance activity to predict, in real time as much as possible, behaviour of the parameters evaluated. This activity will be carried out in parallel with the existing methods in order to be able to validate the new method.