Data-centric middleware based digital twin platform for dependable cyber-physical systems

Abstract

The concept of digital twin, a kind of virtual things with the precise states of the corresponding physical systems, is suggested by industrial domains to accurately estimate the status and predict the operation of machines. Digital twin can be used for development of critical systems, such as self-driving cars and auto-production factories. There, however, will be so different digital twins in terms of resolution, complexity, modelling languages and formats. It is required to cooperate heterogeneous digital twins in standardized ways. Since a centralized digital twin system uses too big resources and energies, it is preferable to make large-scale digital twin system geographically and logically distributed over the Internet. In addition, efficient interworking functions between digital twins and the physical systems are required also. In this paper, we propose a novel architecture of large-scale digital twin platform including distributed digital twin cooperation framework, flexible data-centric communication middleware, and the platform based digital twin application to develop a reliable advanced driver assistance system.