An Approximate Analysis of Expected Cycle Time in Business Process Execution

The accurate prediction of business process performance during its design phase can facilitate the assessment of existing processes and the generation of alternatives. In this paper, an approximation method to estimate the cycle time of a business process is introduced. First, we propose a process execution scheme, with which Business Process Management Systems (BPMS) can control the execution of processes. Second, an approximation method for analyzing its cycle time, based on queueing theory, is presented. We consider agents as queueing servers with multi-class customers and predict the response time of the agents. The cycle time of the whole process is calculated using the expected response time and process structure, taking into account parallel process execution. Finally, the results from the analytical approximation are validated against those of a simulation. This analysis can be used to obtain an optimal process execution plan.