Des stratégies d'approvisionnement pérennes sont nécessaires pour les gestionnaires de la chaîne d'approvisionnement afin de faire face aux incertitudes d’approvisionnement et de demande. La diminution des niveaux de service et l'augmentation simultanée des coûts de stockage sont les impacts les plus importants de ces incertitudes. Les perturbations peuvent être causées par des discontinuités de l’approvisionnement, de l'instabilité politique, des catastrophes naturelles et des grèves des employés. Elles pourraient avoir un effet important sur la performance de la chaîne d'approvisionnement. Pour faire face à de telles perturbations, les modèles d'optimisation des stocks doivent être adaptés pour couvrir une structure de réseau multi-échelons et envisager des stratégies d'approvisionnement alternatives telles que le transport latéral (lateral transshipment) et plusieurs sources d’approvisionnement. Dans ce travail, une approche de modélisation basée sur des scénarios est proposée pour résoudre un problème d'optimisation multi-échelons des stocks. En prenant en compte la demande stochastique et les incertitudes sur les capacités de production, le modèle minimise le coût opérationnel total (coûts de stockage, de transport et de retard) tout en optimisant la gestion des stocks et les flux des marchandises. Afin de faire face aux incertitudes, plusieurs échantillons de scénarios sont générés par Monte Carlo et les exemples correspondants d'approximation (SAA) des programmes sont résolus pour obtenir une politique de réponse adéquate au système d'inventaire en cas de perturbations. De nombreuses expériences numériques sont menées et les résultats permettent d'acquérir des connaissances sur l'impact des perturbations sur le coût total du réseau et le niveau de service. / Supply Chain Management (SCM) is an important part of most companies and applying the appropriate strategy is essential for managers in competitive industries and markets. In this context, Inventory Management plays a crucial role. Different inventory systems are widely used in practice. However, it is fundamentally difficult to optimize, especially in multi-echelon networks. A key challenge in managing inventory is dealing with uncertainties in supply and demand. The simultaneous decrease of customer service and increase of inventory-related costs are the most significant effects of such uncertainties. To deal with this pattern, supply chain managers need to establish more effective and more flexible sourcing and distribution strategies. In this thesis, a “framework to optimize inventory decisions in multi-echelon distribution networks under supply and demand uncertainty” is proposed. In the first part of the research work, multi-echelon distribution systems, subject to demand uncertainty, are studied. Such distribution systems are one of the most challenging inventory network topologies to analyze. The optimal inventory and sourcing policies for these systems are not yet unknown. We consider a basic type of distribution network with a single family product through a periodic review setting. Based on this property, a two-stage mixed integer programming approach is proposed to find the optimal inventory-related decisions considering the non-stationary demand pattern. The model, which is based on a Distribution Requirements Planning (DRP) approach, minimizes the expected total cost composed of the fixed allocation, inventory holding, procurement, transportation, and back-ordering costs. Alternative inventory optimization models, including the lateral transshipment strategy and multiple sourcing, are thus built, and the corresponding stochastic programs are solved using the sample average approximation method. Several problem instances are generated to validate the applicability of the model and to evaluate the benefit of lateral transshipments and multiple sourcing in reducing the expected total costs of the distribution network. An empirical investigation is also conducted to validate the numerical findings by using the case of a major French retailer’s distribution network. The second part of the research work is focused on the structure of the optimal inventory policy which is investigated under supply disruptions. A two-stage stochastic model is proposed to solve a capacitated multi-echelon inventory optimization problem considering a stochastic demand as well as uncertain throughput capacity and possible inventory losses, due to disruptions. The model minimizes the total cost, composed of fixed allocation cost, inventory holding, transportation and backordering costs by optimizing inventory policy and flow decisions. The inventory is controlled according to a reorder point order-up-to-level (s, S) policy. In order to deal with the uncertainties, several scenario samples are generated by Monte Carlo method. Corresponding sample average approximations programs are solved to obtain the adequate response policy to the inventory system under disruptions. In addition, extensive numerical experiments are conducted. The results enable insights to be gained into the impact of disruptions on the network total cost and service level. In both parts of the research, insights are offered which could be valuable for practitioners. Further research possibilities are also provided.

Dans le contexte économique actuel, les entreprises cherchent à développer de nouvelles stratégies de distribution pour leurs performances logistique. Dans cette quête de performances, les entreprises doivent adapter les stratégies de distribution misent en place avec les typologies de leurs produits. Plusieurs stratégies de distribution existent dans la chaîne logistique de grande distribution. Ces stratégies sont choisies sur la base des caractéristiques des produits, et /ou l'impact sur les performances logistiques. Dans cette thèse, nous étudions l'impact de trois stratégies de distribution, à savoir: stockage traditionnel, cross-docking pick by line et le cross-docking pick by store, sur trois performances de la logistiques, à savoir: le niveau de service, les coûts et le bullwhip effect. En outre, nous analysons l'impact des caractéristiques des produits sur les performances des stratégies de distribution et enfin proposer un cadre pour le choix de la stratégie la plus adaptée pour chaque produit. La chaîne logistique étudiée est composée de trois échelons: Centre de distribution du fournisseur, Centre de distribution du distributeur et les magasins. Basé sur un cas réel, nous effectuons une modélisation des processus, qui nous permet de développer un modèle déterministe de coût Macro et un modèle de simulation. Le modèle de coût macro permet d'évaluer l'impact des stratégies de distribution sur des coûts de la chaîne logistique. Après l'analyse macro des coûts, nous développons un modèle de simulation où nous intégrons les données relatives aux produits (la demande, le volume, etc.). Ce modèle permet une simulation dynamique du système la stratégie la plus adaptée pour chaque produit en fonction de ses caractéristiques et de l'impact sur les performances. A la fin de cette recherche, nous présentons une matrice de choix pour la segmentation des produits et choix de la stratégie de distribution. / Nowadays companies must look to develop new distribution strategies in order to achieve the required performance from their supply chain. In this quest, companies wonder about the consistency of their distribution strategies with the products they are selling. Several types of distribution strategies exist in the retail supply chain. These strategies are chosen based on the products characteristics, and/or the impact on the supply chain performances. In this research, we study the impact of three distribution strategies, namely: traditional warehousing, cross-docking pick by line and cross-docking pick by store, on three supply chain performances, namely: service level, cost and bullwhip effect. In addition, we analyse the impact of the products characteristics on the performances of the distribution strategies and propose a framework for choosing the right strategy for each product. The supply chain studied is composed of three echelons: Supplier Distribution Centre, Retailer Distribution Centre and Stores. Based a real business case, we perform a process modelling, that allows us to develop a deterministic Macro cost model and a simulation model. The macro cost model allows to evaluate the impact of the distribution strategies on the supply chain cost performance. After the macro cost analysis, we develop a simulation model where we integrate the data related to the products (demand, volume, ordering quantities etc.) in the model. This model allows a more dynamic simulation of the system in a large time period and determines the right strategy to select for each product depending on its characteristics and the impact on the performances. At the end of this research, we present a framework for product segmentation and distribution strategy selection.

Worldwide globalization has made supply chains more vulnerable to risk factors, increasing the associated costs of outsourcing goods. Outsourcing is highly beneficial for any company that values building upon its core competencies, but the emergence of the COVID-19 pandemic and other crises have exposed significant vulnerabilities within supply chains. These disruptions forced a shift in the production of goods from outsourcing to domestic methods. This paper considers a multi-echelon supply chain model with global and domestic raw material suppliers, manufacturing plants, warehouses, and markets. All levels within the supply chain network are evaluated from a holistic perspective, calculating a total cost for all levels with embedded risk. We formulate the problem as a mixed-integer linear model programmed in Excel Solver linear to solve smaller optimization problems. Then, we create a Tabu Search algorithm that solves problems of any size. Excel Solver considers three small-scale supply chain networks of varying sizes, one of which maximizes the decision variables the software can handle. In comparison, the Tabu Search program, programmed in Python, solves an additional ten larger-scaled supply chain networks. Tabu Search’s capabilities illustrate its scalability and replicability. A quadratic multi-regression analysis interprets the input parameters (iterations, neighbors, and tabu list size) associated with total supply chain cost and run time. The analysis shows iterations and neighbors to minimize total supply chain cost, while the interaction between iterations x neighbors increases the run time exponentially. Therefore, increasing the number of iterations and neighbors will increase run time but provide a more optimal result for total supply chain cost. Tabu Search’s input parameters should be set high in almost every practical case to achieve the most optimal result. This work is the first to incorporate risk and outsourcing into a multi-echelon supply chain, solved using an exact (Excel Solver) and metaheuristic (Tabu Search) solution methodology. From a practical case, managers can visualize supply chain networks of any size and variation to estimate the total supply chain cost in a relatively short time. Supply chain managers can identify suppliers and pick specific suppliers based on cost or risk. Lastly, they can adjust for risk according to external or internal risk factors. Future research directions include expanding or simplifying the supply chain network design, considering multiple parts, and considering scrap or defective products. In addition, one could incorporate a multi-product dynamic planning horizon supply chain. Overall, considering a hybrid method combining Tabu Search with genetic algorithms, particle swarm optimization, simulated annealing, CPLEX, GUROBI, or LINGO, could provide better results in a faster computational time.