RISK MANAGEMENT IN SUPPLY NETWORKS USING MONTE-CARLO SIMULATION Léa A. Deleris Feryal Erhun Department of Management Science and Engineering Stanford University Stanford, CA 94305 U.S.A.
ABSTRACT Trends such as (1) globalization, (2) heavy reliance on transportation and communication infrastructures, and (3) lean manufacturing have led to an increase in the vulnerability of supply networks. Due to a large number of interrelated processes and products, disruptions caused by these vulnerabilities propagate rapidly. Firms, however, can partially control the robustness and resilience of their supply networks through strategic and tactical decisions. Therefore, a decision-support tool that assists managers to evaluate the risk exposure of their supply networks can considerably increase the robustness/resilience of these networks. In this study, we present a Monte Carlo simulation based tool designed to assess uncertainty in supply networks. We describe its application and discuss the possible drawbacks of our approach. 1 INTRODUCTION
Events within the past decades have shown the extent to which companies, and subsequently their supply chains, are vulnerable to adverse events (Deleris, Erhun, and PatéCornell 2004). This observation should be an expected outcome of recent developments in supply networks. For instance, researchers comment on increasing reliance of corporate supply chains on transportation, utilities, and communication infrastructures (NACFAM report 2003 and Cranfield University report 2002). Trends such as globalization and offshoring further increase the complexities and interdependencies in supply chains which are now predominantly characterized by a large number of interrelated processes and products. However, the strategic and tactical decisions of a firm influence the extent to which its supply network is flexible in the face of uncertainty and able to mitigate – or on the contrary exacerbate – the consequences of these adverse events. Therefore, it is important to provide decision-makers at all levels of an organization with appropriate information for decision support. The study that we present in this paper precisely addresses this issue.
Our study originates from discussions that we had with managers at Seltik, a disguised high-tech company based in the Silicon Valley. Seltik operates a global supply network in order to produce several hundred SKUs (Stock Keeping Units). Our initial analysis revealed that managers at Seltik were concerned by the overall vulnerability of their supply network. They were particularly worried about whether the strategic development of their supply network – in terms of their choice of partners and geographical area – was adequate. We present a simulation-based tool designed to help Seltik in the risk assessment of their supply network. Because of the size and intricacies of supply networks, simulation is considered to be a suitable approach for their analysis. For example, Ingalls (1998) points out that unlike optimization, simulation enables to identify robust solutions. The author further notes that robustness, not optimality, is the main concern of senior management when dealing with supply chains. Hicks (1999) describes a fourstep method based on both simulation and optimization aimed at supply chain strategic planning. In this method, simulation is used to describe the dynamic behavior of a given supply chain structure and to assess the benefits of supply chain policies, such as inventory policies. Ingalls (1999) describes a simulation-based tool for supply chain analysis implemented at Compaq, which incorporates demand forecast errors. Deleris, Elkins and Paté-Cornell (2004) use a Monte Carlo simulation of a dynamic stochastic process to determine the losses caused by fire hazard within a large manufacturing network. Our approach for supply...