Characterisation and evaluation of identicality for digital twins for the manufacturing domain
Shuo Su, Aydin Nassehi, Ben Hicks, Joel Ross
Abstract
This paper proposes a comprehensive characteristic for digital twins (DTs) for the manufacturing domain, called “identicality,” to evaluate the capability to represent their physical counterparts in terms of their specific purpose. It is characterised by four attributes: completeness, trueness, precision, and latency. Specifically, completeness refers to the proportion of physical information that has been represented in the DT. Trueness and precision are used to evaluate the agreement between representative information in the DT and the accepted reference value in the physical counterpart. Latency reflects the degree of bidirectional synchronisation in the twinning process. Regarded as the representative information, data, model(s), and parameters in the digital representation are considered. The evaluation method for identicality is then driven by the specific purpose of the DT for the manufacturing scenario. The approach involves two stages: manufacturing scenario analysis and identicality evaluation. Evolved from a proposed manufacturing ontology, an information model with purpose-based weights is developed in the first stage. Based on it, four attributes are evaluated by integrating metrology principles and human knowledge. The characteristic and its evaluation method are demonstrated using a specific use case of estimating the dimensional accuracy of a printed artefact during the material extrusion (MEX) process.