A Three-Level Decision Support Approach Based on Multi-objective Simulation-Optimization and DEA: A Supply Chain Application

Abstract

Decision-making in multi-objective problems is a complex problem with several approaches existing in the literature. Many such approaches typically combine simulation and optimization methods to achieve a robust tool capable of providing useful information to decision-makers. While simulation helps decision-makers to test alternative scenarios and allows uncertainty to be considered, optimization enables them to find the best alternatives for specific conditions. This paper proposes an alternative approach consisting of three levels, wherein we use simulation to model complex scenarios, simulation-optimization to identify the scenarios of the Pareto-front, and Data Envelopment Analysis to identify the most efficient solutions, including those not belonging to the Pareto-front, thereby exploiting the benefits of efficiency analysis and simulation-optimization. This can be useful when decisionmakers decide to consider scenarios that, while not optimum, are efficient for their decision-making profile. To test our approach, we applied it to a supply chain design problem. Our results show how our approach can be used to analyze a given system from three different perspectives and that some of the solutions, while not optimum, are efficient. In traditional approaches, such scenarios could be overlooked, despite their efficiency for specific decision-making profiles.