Production planning of mechanical plants in coordination with vehicle assembly plants.

Summary The objective of this thesis is to propose a tool for the production planning of a supplier and to study the means to improve the coordination between this supplier and the customer plants. Applied to the automotive industry, this thesis is based on the case of mechanical factories (engine, gearbox, and ground connection parts factories) that supply vehicle assembly plants.By analyzing the demand of customer factories, we highlight the factors of variability of this demand and propose ways to improve coordination with the supplier through a more efficient sharing of information. For the supplier's production planning, we propose a mathematical model in the form of a linear integer program. This model allows to define a detailed production plan by respecting the existing constraints and by aiming at four objectives: to satisfy the forecasted demand, to reach the safety stocks, to balance the stocks between the products, and to smooth the production. The interest brought by this model is demonstrated by tests performed on real data. Knowing that the proposed model is intended to be used in the framework of a rolling horizon planning, we analyze by simulation the influence of two parameters: the planning frequency and the length of the frozen horizon. We give recommendations on the choice of these parameters in order to ensure a good compromise between the service rate, the stock level, and the stability of the planning. Finally, to improve the coordination between the client factories and the supplier, we study two avenues in particular: the improvement of the reliability of the forecasts communicated to the supplier and the extension of the time allowed to the supplier to satisfy the demand. The interest of implementing these proposals is evaluated by simulation.