From a reactive, heterarchical to a holonic system: an application for optimizing flow in an automotive plant

The aim of this paper is to study the notion of overall optimization of vehicle flow in an automotive plant. The present system architecture for managing flows may be described as heterarchical reactive. In terms of optimization, this has been translated as a succession of independent local schedulers. With the aim of reducing costs whilst maintaining quality, the goal is to increase the scheduling power of each scheduler with a view to global optimization. The proposal presented here is to model this strategy not by using a reactive heterarchical architecture but by a holonic-type architecture. A mathematical model has been used to formalize each of these two types of architecture. The advantages of this new architecture in terms of global performance and robustness of scheduling have been highlighted.