Proactive Decision Support During Business Process Execution

The execution of business processes produces lots of events that can be used by complex event processing systems to analyze and improve the processes. Typically, events are stored in an event log repository that can be used to identify meaningful patterns of events, such that it is possible to react to them during process execution. However, in many cases it is beneficial to deal with those events before they actually occur to avoid an undesirable outcome, e.g., machine failure or poor performance indicator. Therefore, we present a system architecture connecting the operation of a process engine with a proactive framework. The framework forecasts events, provides the best corresponding action and generates appropriate business rules that can be used by the process engine to make optimal decisions during process runtime. An elaborated example demonstrates the utility of our concept.