Analysis of supply chains using system dynamics, neural nets, and eigenvalues

Supply chain management is a critically significant strategy that enterprises depend on in meeting the challenges of today's highly competitive and dynamic business environments. An important aspect of supply chain management is how enterprises can detect the supply chain behavioral changes due to endogenous and/or exogenous influences and to predict such changes and their impacts in the short and long term horizons. A methodology for addressing this problem that combines system dynamics and neural networks analysis is proposed in this paper. We use neural networks' pattern recognition abilities to capture a system dynamics model and analyze simulation results to predict changes before they take place. We also describe how eigenvalue analysis can be used to enhance the understanding of the problematic behaviors. A case study in the electronics manufacturing industry is used to illustrate the methodology.