Using ML techniques to build systems support in the industrial processes

This thesis project is based on the application of Machine Learning (ML) techniques on industrial processes. This work was carried out in the R&D department of a leading tire manufacturing company, Pirelli. The proposed project consists of two different objectives to be achieved through the use of ML: Prediction of curing cycles, the curing cycle is a process that brings the green tire (the tire in its embryonic form) into its final form, the one we all know, with distinct mechanical properties. Our goal is to predict the combination of the three thermal variables that define this process: pressure, temperature, and time. Prediction of Mooney viscosity, a critical parameter for quality control of the mixing process, as knowing the viscosity of the batches will allow minimizing the viscosity variation of the blend in subsequent processes.