Supporting multi-layer modeling in BPMN collaborations

Abstract

In recent years, BPMN has acquired a clear predominance among the notations for modeling business processes. This is mainly due to its capability to close the communication gap between business and IT people. As a consequence, the quality of produced models is more and more important and, among the others, understandability plays a relevant role to permit to properly convey information in such a heterogeneous context. To improve understandability, it is generally recommended to not overwhelm models with to many details, and to use instead sub-process modeling elements to split collaborations into layers. However, the BPMN standard does not provide precise specifications on how the details, hidden at the given layer, should be included in the model, in particular considering message exchange. In particular, the consistency checking between collapsed sub-processes and their detailed representation is left to the modeler, and there is not much support to help him/her in this activity. In this paper, we analyze BPMN modeling tools with respect to their actual capabilities to support multi-layer collaborations. From the analysis we observed a general lack of support in the modeling environment. Then we propose a design methodology providing a set of guidelines to ensure consistency in multi-layer collaborations. These guidelines have been implemented in a stand alone tool, which enables their automated checking in any BPMN modeling tool.