Robust Parameter Setting of Supply Chain Flexibility Measures Using Distributed Evolutionary Computing

Abstract

Today's supply chains are challenged by volatile customer demand. Demand for a wider product choice, shortened product lifecycle and expected high availability add to the already complex, dynamic and uncertain business environment. Operating under such conditions poses difficulties to a company to uphold their supply chain's performance. Flexibility is required to be able to adapt to unanticipated changes in supply or demand and to diminish their repercussions. Miscellaneous flexibility measures, e.g. safety stocks or flexible capacities, are widespread used to compensate demand fluctuations. The selected measures’ parameters, e.g. range of flexible capacity, must be configured ahead of the implementation in practice. The flexibility parameters determine the scope of action a flexibility measure enables. This paper seeks to address conceptually the issue of setting robust flexibility parameters using a simulation-based optimization approach. Genetic Algorithm and Particle Swarm Optimization are used in a distributed island approach to optimize the flexibility parameters.