Workflow Modeling Using Stochastic Activity Networks

Abstract

The essence of workflow systems is workflow patterns. The aim is to use an existing powerful formal modeling language with workflow systems. Stochastic activity networks (SANs) are a powerful extension of Petri nets. Having the SAN model of a system, one can verify the functional aspects and evaluate the operational measures, both on a same model. SANs have already been used in a wide range of applications. As a new application area, we have used SANs for modeling workflow systems. The results show that the most important workflow patterns can be modeled in SANs. In addition, the resulting SAN models of workflow systems can be used for model checking and/or performance evaluation purposes using the existing tools. In this paper, we will present the results of this work. For this purpose, we will present the SAN submodels corresponding to the most important workflow patterns. Then, the proposed SAN submodels are used in a case study for workflow modeling, which will also be presented in this paper. Finally, we will present the results of the evaluation of the model using the Mobius modeling tool.