Representing adaptation options in experimentable digital twins of production systems

Abstract

ABSTRACTSimulations are powerful tools for decision support during factory adaptation processes. In order to provide the most valuable form of decision support, a tool must take into account the decision problem and relevant decision alternatives. Today’s simulation tools, however, only accept simple notions of variability, like numeric parameter ranges. Thereby they ignore most of the variability of production systems, and do not utilise their full potential to aid decisions. To improve this situation, a new concept on how to systematically model the variability of production systems in Digital Twins of production entities is proposed. It combines Model-Based Systems Engineering and Variability Management to model different variants of production systems, and utilises capabilities of production equipment to make Digital Twins modular and reconfigurable. Each valid combination of variants results in a directly 3D-simulable Digital Twin for the whole production system, allowing automatic validation testing and fast feedback loops during system development. The presented concept is not only a very important basis for managing variants. As the variants model can be used as a search space for optimisation algorithms, the concept is an important stepping stone for a more powerful simulation-based optimisation of production systems.