D. Lehner, J. Pfeiffer, S. Klikovits, A. Wortmann, M. Wimmer: A Method for Template-based Architecture Modeling: and its Application to Digital Twins, Journal of Object Technology, Vol. 22, No. 3, 2024. Doi: 10.5381/jot.2024.22.3.a8


Digital Twins (DTs) have recently emerged to support domain experts in engineering and operating Cyber-Physical Systems (CPSs). As a result, software vendors started to create DT services offering advanced functionality to support the development and operation of DTs in industry. However, the integration of services into a DT architecture is challenging. Services typically rely on specific software and modeling languages that are often not interoperable with other services. Hence, they have to be manually integrated which requires a significant, repetitive effort. Thus, currently, it is tedious to extend the DT’s underlying architectures with new services or exchange individual service implementations. In this paper, we propose a tool-supported method for architecture modeling and its application for digital twins. The presented method provides several steps to manage the complexity of current DT architectures. First, DT reference architectures are assembled by connecting DT templates, which provide an abstraction for a set of similar DT services. Second, dedicated DT modules are used to wrap existing services which provide concrete realizations of the DT templates. Third, a product-line-based generator supports the configuration of reference architectures into concrete architectures by selecting an appropriate set of modules for the used templates which are finally used in the derived integration solution. The transition from reference architecture modeling to the product-line-based configuration is supported by a dedicated model transformation. Our evaluation shows that the proposed DT templates enable the efficient modeling of different DT reference architectures and integration of new DT services into already existing architectures.

A Method for Template-based Architecture Modeling and its Application to Digital Twins