Robust flexible real-time monitoring of workflow processes

Business processes show unpredictable and unstable behavior. This requires workflow systems to support monitoring functions with the ability to flexibly and robustly adapt to the changing activities of the workflow at runtime. In this paper flexible distributed realtime workflow monitoring is realized by a layered architecture. The architecture consists of: (1) middleware components supporting real-time and fault-tolerant monitoring and (2) different controllers for controlling workflows and performance data, view models for the specification of the graphical user interface and quality of service (QoS) requirements. A software system was implemented as a client-server application in Java using real-time and fault-tolerant CORBA middleware components. The prototype shows that the separation of process models, data models and view models can lead to flexible monitoring of workflows in real-time and thus guarantees robust and valuable information about business processes at runtime.