Supply chain modelling and its analytical evaluation

Abstract

A model-based analysis of supply chains (SCs) is proposed as a means of estimating performance measures like lead times and resource utilisations. SCs are regarded as discrete event dynamic systems (DEDS), described as process chains in an application-specific formalism, and analysed by state-of-the-art techniques developed for queueing networks (QNs) and Petri nets (PNs). The theoretical approach is accompanied by a toolset that provides automatic translations between a process-chain-specification front end and analysis engines for QN and PN models. In addition, the analysis technique also supports a hybrid approach, in which dedicated sub-models are mapped into a PN model, numerically analysed, and replaced with an aggregate. Finally, the resulting QN models may be analysed using analytical as well as simulative techniques. As an example application, we investigate the impact of an additional SC channel between a manufacturer and web-consumers on the overall performance of an SC. The example was inspired by the trend towards retail trade via the internet and illustrates the new challenges of electronic commerce for SC management.