Constructing Multi-level Credibility Models for Digital Twins: A Benchmark for Credibility Evaluation

Digital twin (DT) technology holds significant potential for industrial applications. However, their widespread adoption depends on the credibility of DT models, necessitating robust credibility evaluation methods. Existing DT credibility evaluation methods are often validated only on high-credibility models, which is insufficient to comprehensively demonstrate their effectiveness. To address this limitation, it is essential to construct tailored model repositories for specific DTs, providing diverse test cases for evaluation. This paper proposes a methodology to construct a multi-level benchmark model repository that covers different levels of credibility for DTs and illustrates its application using an industrial robotic arm as a case study.