Discrete event simulation for supply chain management

MELE, FERNANDO D.; ESPUÑA, ANTONIO; PUIGJANER, LUIS

Indianápolis, EEUU

Conferencia; AIChE Annual Meeting 2002; 2002

AIChE

This work proposes a dynamic approach for Supply Chain Management (SCM) based on the development of a discrete event-driven system model of a Supply Chain (SC) contemplating several companies. The interaction between these companies may be very complex, so it is difficult to mathematically analyse in general terms, making it necessary to explore the scenario through simulation techniques. The proposed model includes all the entities that belong to the SC as independent and well-defined blocks, each of these blocks being represented by a collection of states and transitions. The aim of this work is to represent and to model in a functional way (capable to be used in a simulation and a decision-making environment), the interactions between the components of the chain, to finally obtain a decision-making tool for Supply Chain Management.This approach models the material and information flow within the Supply Chain. Since these flows have intricate interrelationships, it is difficult to easily formulate a compact model for simulation. Therefore, in this paper, the functions needed for SC simulation are expressed modularly and they are interconnected to generate a whole SC model. The system envisaged considers the SC as decentralised system where there is no global co-ordinator and every entity in it takes decisions locally. A variety of scenarios including several aspects and modes of operation (inventory control policies, choosing between suppliers, demand uncertainty, etc.) are tested in order to analyse their impact on the behaviour of the chain and to demonstrate the validity of this kind of representation. The system performance is evaluated by means of certain indicators such as transport cost, inventory cost, and/or customer satisfaction level.The way in which the model works is illustrated with several examples, including industrial scenarios.The results obtained provide information about the trade-offs found in real systems and give valuable insight for future work underway to consolidate SCM based on discrete event models and simulation.