Digital twin as a reference–decision model for engineering analysis and system-level decision support

In recent years, the concept of the digital twin has gained wide popularity in engineering, computer science, and spatial planning; however, its interpretation remains ambiguous and is often detached from metrological rigour. In practice, a digital twin is often equated with a visual 3D model or an extended GIS platform, leading to conceptual dilution and limiting its applicability in decision-making processes involving high responsibility. The aim of this paper is to propose a geodetic and cartographic interpretation of the digital twin as a new generation of measurement-based products.The paper presents the evolution of geodetic and cartographic products—from static maps, through GIS systems and three-dimensional models, to spatio-temporal (4D) models that provided the foundation for the development of the digital twin concept. Against the background of existing definitions, an original definition of the digital twin in a geodetic-cartographic framework is proposed, emphasising unambiguous spatial referencing, explicitly defined accuracy and measurement uncertainty, and the decision-support function of the model.A typology of digital twins is also introduced, encompassing visual, analytical, and reference–decision digital twins (DT-G), of which only the latter fully satisfies the criteria of a geodetic-cartographic product. Particular attention is devoted to the role of geodetic control networks, reference systems, data calibration, and validation in ensuring the reliability of dynamic models.It is demonstrated that in systems characterised by large spatial extent, limited field accessibility, and high decision-making responsibility, the reference–decision digital twin is no longer an optional solution but becomes a prerequisite for the reliability of spatial analyses.The paper has a conceptual character and does not include empirical validation based on operational measurement data.The proposed concept is illustrated with using examples of linear critical infrastructure, such as railway networks, power transmission systems, and state borders.The paper formulates the thesis that any digital twin used in decision-making processes should be treated as a formal measurement model, firmly rooted in the conceptual frameworks of geodesy and cartography.From an engineering perspective, the proposed reference–decision digital twin is treated as a system-level analytical model that supports engineering analysis and decision-making for large-scale technical systems. In this sense, the digital twin is positioned within analysis and modelling, rather than as a visualisation or a purely geodetic product.