Process Model for the Data-driven Identification of Machine Function Usage for the Reduction of Machine Variants
IEEE International Conference on Industrial Engineering and Engineering Management (IEEM 2022)
Contribution to Conference
Current research has shown that a data-driven approach to identify customer usage patterns provides sufficient indication to cluster, predict and derive decisions. However, specific problems and questions arise due to the peculiarities of complex mechatronic systems. This research aims at the development of a process model to support a systematic reduction of machine variants by analyzing field-gathered data concerning machine function usage. To analyze factors influencing the data-driven identification of machine usage data, 21 semi-structured expert interviews are conducted. Based on the core statement given by the participants, six factors have derived that influence a data-driven identification of machine function usage. Further, based on the derived requirements, the CRISP-DM process model and the SPALTEN problemsolving methodology, an initial process model is developed. This model is evaluated by the conduction of multiple data analyses. To evaluate the designed process-model the effect of the derived analysis results on the potential reduction of variants is examined. The exclusion of two identified unused machine functions indicated a theoretical bisection of the variety of machine variants.