LEARNING ABOUT THE BULLWHIP EFFECT USING COLORED PETRI NET SIMULATOR

Abstract

Colored Petri Net (CPN) simulator for study and analysis of supply chain processes is described in this paper and some experimental results obtained by its use are presented. This CPN simulator is one version of the well-known beer game and overcomes some shortcomings we have noticed playing other beer games. The beer game is a role-play simulation game that lets students (or managers as well) experience typical coordination problems of supply chains.  Supply chain consists of four stages (or co-makers): retailer, wholesaler, distributor and manufacturer. The simulator allows calculation of different supply chain performances. It is developed using a timed, hierarchical colored Petri Net and CPN Tools software package. The CPN structure developed and presented in this paper consists of one top page and 17 sub-pages in four hierarchical levels, which models decomposition of observed supply chain in sub-processes. For each stage, demand forecasting methods, replenishment policies, production and delivery times, inventory costs, and customer’s demand may be defined and given as input data for simulation run. In CPN Simulator we have used three strategies for demand forecasting. First strategy is based just on the past demand and second and third strategies are based on adaptive time series method: Moving Average and Exponential Smoothing Methods. We conducted three groups of experiments, every group for a different strategy of demand forecasting and for every group we experimented with several different parameters. The simulation results are exported to Excel and its visual presentation and expressive reporting capabilities are used. This package is aimed for evaluation of different management strategies in a supply chain, as well as for educational purposes. This developed model can efficiently be used for teaching the bullwhip effect not only to undergraduate and graduate students, but managers as well.