A scenario-based evolutionary scheduling approach for assessing future supply chain fleet capabilities

Abstract

When assessing land vehicle fleet capabilities in off-shore military operations, it is necessary to develop and apply a reliable mechanism for creating and testing different fleet structures. In this paper, the problem of optimizing vehicle fleet mixes in the settings of military deployments is introduced, the complexity of this problem is discussed and a fleet-optimization system that optimizes multiple objectives while satisfying the system constraints is introduced. The system is based on a set of heuristics that can answer a number of key questions required for long-term capability planning such as utilization of current fleet, mix of different vehicle and modular units for a given scenario, and overall fleet structure. It is evolutionary and uses a relaxed version of the concept of Pareto-dominance to identify the set of different options available for possible use by military analysts. The effectiveness of our system through a case study based on a simple example dataset is demonstrated. The novelty in this work is that we optimize a problem where scheduling is tightly coupled with routing and bin packing. In addition, land-based operations are also constrained with different standard operating procedures that complicate the scheduling problem. This paper is a first attempt towards solving this complex problem. © Springer-Verlag Berlin Heidelberg 2007.