Analysing bottlenecks in production processes with the help of process mining

Bottlenecks are detrimental to the output rate of a process, causing extended waiting times and increased costs of production. This paper will be looking at process mining, a technique used to create a process model using the event log. Multiple methods will be discussed that will be able to determine bottlenecks in systems, and their precision and recall rates will be evaluated to find the best model. As most studies stay in the detection of bottlenecks,this paper will also explore the methods for predicting bottlenecks. Using predictive process modelling and transition systems, it is possible to predict time left for an item to be finished and if the item is up to quality.