Product-oriented modelling, simulation, and analysis of multiple-variant production in supply chains (diss.)
Researcher:
Financed by:
The Professorship’s own resources
Motivation:
In recent years many companies have tried to increase their customer orientation and to differentiate from their competitors by expanding the range of products they offer. In addition, a global market presence often demands product modifications to match different local requirements. This has led to an exponential growth of the number of product variants. The persistent shortening of product life cycles adds to this effect even more.
Consequently, companies face a number of challenges, as product variety results in an increasing complexity of products, processes, and production systems. Often, the only possible way to answer questions in this context is by utilizing appropriate models. Discrete event simulation can then help to analyze the network, taking into account dynamics and uncertainties. One of the main challenges thereby is to limit the amount of effort required for building and analyzing such models. Many of the existing approaches suffer from the fact that the acquisition of all the necessary data throughout the network is very difficult and time-consuming. Other approaches focus on a limited number of aspects, resulting in an optimization of one aspect at the expense of others.
Objectives:
The objective of the dissertation is to develop an efficient approach to modeling supply chains with product variants. This comprises identification of necessary information, representation of this information in a concise model of the supply chain, and guidance of the decision taker through the analysis. A software prototype will be developed and validated in several industry cases.
Activities completed:
Existing modeling approaches for supply chains were analyzed and clustered with regard to their functionalities, information requirements, robustness, and tangibility. It was found that taking the product and its bill of materials as a starting point for building the supply chain model can reduce the modeling effort significantly. This approach was then extended using knowledge from product configuration and process configuration. The resulting model allows for a structured and easy mapping of product families and the corresponding production processes. The modeling approach was implemented in a software prototype based on Java 1.5 and the Eclipse Rich Client Platform.
Activities in progress:
The model will be connected to a discrete event simulation. Further, specific analysis functionalities to support decisions concerning efficient management of product variety in a production network will be developed. The software prototype will constantly be extended in the course of the project.