Agent-Based approach to supply chain: How unexpected changes in the factory affect the supply chain

Abstract

The production industry has to control the material flow from suppliers to customers; the supply chain is the name for the network that performs these activities, and the enterprises have to consider it as a key issue that must be continuously improved.

This master's thesis studies how the disturbances in a factory affect the supply chain. To do this, first a supply chain model is built using agent-based modelling software, and a set of experiments has been designed and executed to validate its general behaviour. Later, disturbances in factory have been modelled, and a new set of experiments has been developed to study how the system behaves under these different disturbances, The disturbances in the factory have impact on the whole supply chain and the results are dependent on the forecasting system used by the agents. Two different types of forecast have been used: Agent-based and centralized forecast. Because of bullwhip effect, the use of the agent based forecast leads to a higher total cost, and higher stock levels. After the disturbances, the factory needs a noticeable longer time to recover than the other agents in the supply chain.