Combining Process Mining And Simulation In Production Planning

Abstract

The conditions for industrial companies are changing due to increasing customer demands for individualised as well as sustainable products. Furthermore, companies are confronted with technological change by digital transformation. Therefore, production planning has to address various structural, procedural and organisational changes. Planning projects are often characterised by a high degree of complexity. In order to master the associated challenges, simulation models are used in production planning. In contrast to mathematical-analytical methods, simulation models examine and assess especially complex production systems and support improvement measures. A major difficulty during the model initialisation and the determination of the planning variables is the capture of data and the assurance of sufficient data quality. Both are associated with a high expenditure of time. At this point, manufacturing companies are faced with a conflict of objectives between the reduction of the planning time and the development of reliable simulation models. Process Mining (PM) can be used to capture data from central information systems and to uncover social and organisational networks and map them in a process model. This can create a well-founded data basis for simulation models. To support simulation models within the planning process, a methodology linking process mining and simulation has been developed. This methodology improves the database within the planning process and renders it usable for rescheduling production systems. Potentials that can be achieved in the areas of data acquisition, data quality and model building are systematically analysed. The approach is validated on the basis of a use case from the pharmaceutical industry.