Automated Digital Twins Generation for Manufacturing Systems: a Case Study
Giovanni Lugaresi, Andréa Matta
Abstract
The recent industrial scenario was defined by the emergence of digital twins and cyber physical systems as key elements for manufacturers leadership. Digital models can perform good in terms of production planning and control decisions if they are correctly representing their physical counterparts at anytime. Discrete event simulation can be considered as established digital models of manufacturing system, thanks to the proven capabilities of correctly estimating the system performances. Automated simulation model generation techniques can significantly reduce model development phases and allow for using simulation models for short term decisions in production. Application studies and test cases are scarce in the literature. In this paper, we present the application of a digital model generation method. The test case is done exploiting a lab-scale model of a manufacturing system composed by six stations. We investigate how the model generation works online, during the transient phase of a manufacturing system. Results confirm the real-time applicability of the approach provided that sufficient data points are available from the production event logs.