Digital Twin Fidelity Requirements Model for Manufacturing

Abstract

The Digital Twin (DT), including its sub-categories Digital Model (DM) and Digital Shadow (DS), is a promising concept in the context of Smart Manufacturing and Industry 4.0. With ongoing maturation of its fundamental technologies like Simulation, Internet of Things (IoT), Cyber-Physical Systems (CPS), Artificial Intelligence (AI) and Big Data, DT has experienced a substantial increase in scholarly publications and industrial applications. According to academia, DT is considered as an ultra-realistic, high-fidelity virtual model of a physical entity, mirroring all of its properties most accurately. Furthermore, the DT is capable of altering this physical entity based on virtual modifications. Fidelity thereby refers to the number of parameters, their accuracy and level of abstraction. In practice, it is questionable whether the highest fidelity is required to achieve desired benefits. A literary analysis of 77 recent DT application articles reveals that there is currently no structured method supporting scholars and practitioners by elaborating appropriate fidelity levels. Hence, this article proposes the Digital Twin Fidelity Requirements Model (DT-FRM) as a possible solution. It has been developed by using concepts from Design Science Research methodology. Based on an initial problem definition, DT-FRM guides through problem breakdown, identifying problem centric dependent target variables (1), deriving (2) and prioritizing underlying independent variables (3), and defining the required fidelity level for each variable (4). This way, DT-FRM enables its users to efficiently solve their initial problem while minimizing DT implementation and recurring costs. It is shown that assessing the appropriate level of DT fidelity is crucial to realize benefits and reduce implementation complexity in manufacturing.