Digital twin technology aims to increase the accuracy of objects' and processes' representation and predictions using fine-grained models and real-time synchronization. Considering runtime data integration, the multi-physical and multi-scale nature of different models used to simulate all the complex structures and processes, the digital part of a twin becomes vast and hardly manageable. This research aims to develop the architecture of a digital twin of the road system using a single underlying model - an approach dedicated to increase a system's manageability by storing all information about the system and creating specified partial concern-based representations - views. Software product lines are similar to a Digital Twin in modularity, real-time data integration, and runtime reconfiguration. Therefore, it is feasible to reuse the experience of constructing and maintaining software product lines in digital twins. Variability management allows explicitly representing variability in software artifacts throughout the lifecycle, managing dependencies among different variabilities, and supporting the instantiations of those variabilities [1]. Three system views are used for this representation: - The structural view emphasizes the system's static structure using objects, attributes, operations, and relationships. - The behavioral view shows the system's dynamic behavior by showing collaborations among objects and changes to the internal states of objects. - Contextual view represents context awareness of the system, i.e., the capability to change its behavior according to conditions arising during execution [2]. [1] Lianping Chen, Muhammad Ali Babar, and Nour Ali, Variability management in software product lines: A systematic review. Proceedings of the 13th International Software Product Line Conference, pp. 81–90, 2009. [2] Mahsa Varshosaz, Harsh Beohar, and Mohammad Reza Mousavi, Basic behavioral models for software product lines: Revisited. Science of Computer Programming, Vol. 168, pp. 171–185, 2018.