Modeling of transport times in partly observable factory logistic systems based on event logs

Internal logistic systems are an important part of the modern industry. Their design involves as a rule elaborate simulations. However, during the lifetime of a factory, simulation models of the transport system often become outdated as components are added or removed and the laborious maintenance of the models is neglected. Nonetheless, for performance optimization of the transport system, accurate, up-to-date models are needed. This paper describes an approach to semi-automatically generate a transport time model based on event logs and elementary routing knowledge about a running factory. Both kinds of data are usually readily available and reflect the current system behavior. The purpose of this approach is to establish a basis for revealing optimization potential of transport systems and to lessen the effort of setting up a detailed simulation model.