The development of a logistics state representation using industrial performance indicators

This study takes place in the new definition of logistics performance, with appropriate indicators and try to enhance the decision-making tools with a specific representation of these indicators. In the paper, we attempt to represent the variations in time of result indicators (i.e. global indicators) from a discrete states representation and we outline how these states can be used to identify the most pertinent action variable at a decision level. For a given global indicator, several local indicators are involved. Each of them has to be appraised, and should be characteristic of a specific task and closely tied-up with an specific (re)action in malfunctioning situations or when objectives become unreachable. The logistics states can be considered as vectors with Boolean components, each of them is coded 0 or 1 depending on whether the current value of the observed variable is lip or down according to a predefined threshold. It is thereby possible, using a Junctional model of the supply chain and a simulation model, to evaluate the probability of each state, to follow up the state evolution of the system day by day, and to identify the states which are ascribed in Junctional diseases and bad performances and therefore the corresponding action meanings. We present an industrial example and discuss about results obtained by simulation and about perspectives of this kind of representation.