Digital Twin of Automotive Press Shop Logistics. The Stellantis Case.

This thesis work has been planned and realized thanks to a collaboration between Politecnico di Torino, Stellantis Press Shop&Dies in Mirafiori and Centro Ricerche Fiat (CRF). It is focused on the construction of a simulation model that aims to replicate, as closely as possible to reality, the internal logistics flows affecting finished products of the Stellantis stamping plant in Mirafiori. The automotive industry, where the stress produced by global market competition and simple and rapid demand fluctuation has prompted production plants to discover innovative solutions to adapt to the changing environment, is a perfect example of the greater responsiveness that manufacturing systems must face nowadays. The adoption of Industry 4.0 principles by automotive suppliers' operations is critical in the struggle for market share. The main purpose of these strategies is to save money so that big expenditures in value partnerships and technological advances in customer service may be made. To effectively integrate Industry 4.0 technology into each phase of the car-making process, a thorough examination of the current state of automotive factories is required. Progress in the field of simulation has now provided managers with new effective tools for determining the optimum solution for each goal. The purpose of this simulation study is to create a Digital Twin in order to test how the plant can respond to an introduction of some Industry 4.0 technologies. The work is part of a larger project to renovate a traditional plant to improve its efficiency and productivity. The project started in September 2021 with a careful study of the plant and an intensive data collection, with the aim of obtaining all the information useful to build the parameters of the model. Subsequently, a first draft of the logistic flows was presented to the work committee and the useful level of detail to be achieved for the plant was discussed. This first phase led to the identification of the most frequently handled parts, on which the times were calculated for the various phases that lead the metal sheets to become the parts that make up the structure of the new cars. In the following months, the implementation on the Plant Simulation software went step by step, starting from a more general model and then going deeper into the addition of parts, trying to achieve an effective consistency with the outputs of the real production days. Initially, the focus was on representing production lines that mimicked the speeds of the various parts. Then we moved on to the representation of the transport flows of the lines, the main protagonists of the analysis. Then, a series of simulation logics were introduced to imitate the behavior of human operators within the software. Finally, tests on past production days were carried out to verify the model thus constructed. The project is a first step towards the construction of a software application that can support the future strategic decisions that plant managers will have to face.