Complexity-Based Evaluation of Production Strategies Using Discrete-Event Simulation

Abstract

Today’s networks of production and logistics are often characterized by a large structural and dynamical complexity. As a consequence of their nonlinear and potentially unstable dynamics, an efficient planning and control is hardly possible, resulting in economic risks. The solution of the corresponding problems requires an overall understanding of the complex behavior of such systems. For this purpose, discrete-event simulation is applied to study networks consisting of only a few cooperating manufacturers. Based on these simulations, it is possible to identify dynamical mechanisms which make the dynamics of inventory levels strongly irregular even without the presence of stochastic factors. Many of the underlying dynamic instabilities can be attributed to an imperfect logistic synchronization, which emerges due to an improper mutual adjustment of lot sizes, transportation and processing times. In order to approach more quantitative results, a combination of different concepts from nonlinear time series analysis (such as symbolic time series analysis and recurrence quantification analysis) is suggested that provides characteristic measures for the complexity of inventory variations. The corresponding methods allow a systematic evaluation and potential improvement of the performance of different concepts and strategies for the control of material flows in manufacturing networks.