Modelling of Coordinating Production and Inventory Cycles in A Manufacturing Supply Chain Involving Reverse Logistics

In today’s global and competitive markets selling products at competitive prices, coordination of supply chain configuration, and environmental and ecological consciousness and responsibility become important issues for all companies around the world. The price of products is affected by costs, one of which is inventory cost. Inventory does not give any added value to products but must be kept in order to fulfill the customer demand in time. Therefore, this cost must be kept at the minimum level. In order to reduce the amount of inventory across a supply chain, coordination of decisions among all players in the chain is necessary. Coordination is needed not only for a two-level supply chain involving a manufacturer and its customers, but also for a complex supply chain of multiple tiers involving many players. With increasing attention being placed to environmental and ecological consciousness and responsibility, companies are keen to have a reverse supply chain where used products are collected and usable components remanufactured and reused in production to minimize negative impacts on the environment, adding further complexity to decision making across a supply chain. To deal with the above issues, this thesis proposes and develops the mathematical models and solution methods for coordinating the production inventory system in a complex manufacturing supply chain involving reverse logistics and multiple products. The supply chain consists of tier-2 suppliers for raw materials, tier-1 suppliers for parts, a manufacturer who manufactures and assembles parts into finished products, distributors, retailers and a third party who collects the used products and returns usable parts to the system. The models consider a limited contract period among all players, capacity constraints in transportation units and stochastic demand. The solution methods for solving the models are proposed based on decentralized, semi-centralized and centralized decision making processes. Numerical examples are used by adopting data from the literature to demonstrate, test, analyse and discuss the models. The results show that centralised decision making process is the best way to coordinate all players in the supply chain which minimise total cost of the supply chain as a whole. The results also show that the selection of the length of limited horizon/ contract period will be one of the main factors which will determine the type of coordination (decentralised, centralised or semi-centralised) among all players in the supply chain. We also found that the models developed can be viewed as generalised models for multi-level supply chain by examining the models using systems of different tiers from the literature. We conclude that the models are insensitive to changes of input parameters since percentage changes of the supply chain’s total cost are less than percentage changes of input parameters for the scenarios studied