Page 21 -

6 W. van der Aalst et al.
techniques used for the analysis of business processes originate from operations
research. Students taking courses in operations management will learn to apply
techniques such as simulation, queueing theory, and Markovian analysis. The focus
mainly is on performance analysis and less attention is paid to the correctness of
models. However, verification is needed to check whether the resulting system is
free of logical errors. Many process designs suffer from deadlocks and livelocks that
could have been detected using verification techniques. Notions such as soundness
can be used to verify the correctness of the systems.
Finally, models can be used to enact processes. In the context of a BPM system,
models are often used for enactment, that is, based on a model of the process, the
corresponding runtime support is generated. In a workflow management system, a
model of a process suffices to generate the corresponding system support. In other
environments, the set of processes is often hard-coded. For example, although ERP
systems like SAP have a workflow engine, most processes are hard-coded into the
system and can only be changed by programming or changing configuration param-
eters. As a result, modifications are either time-consuming (because of substantial
programming efforts) or restricted by the set of predefined configuration parameters.
Figure 1.2 shows another view on the role of (process) models. Models can be
used to model and analyze operational processes without explicitly considering the
information system. Consider, for example, the use of business process simula-
tion in environments where Information Technology (IT) plays only a minor role.
However, models can also be used to configure and/or implement systems. The
basic idea of workflow technology is to generate systems on the basis of models.
Figure 1.2 emphasizes the role of event logs. Information systems record events
and this information can be used when designing or analyzing processes. In fact,
workflow management systems provide excellent facilities for logging events at

Fig. 1.2 On the role of (process) models in a BPM setting