A Structured Methodology for Semantic Integration of Physical and Cyber Assets in Industrial Digital Twins

This paper addresses the lack of structured modelling approaches for Digital Twin (DT) systems that coherently integrate both physical and cyber assets in industrial environments. Existing standards such as the Asset Administration Shell (AAS) provide foundations for representing tangible components, but offer limited support for non-physical entities and semantic consistency across complex systems. To overcome these limitations, we propose the Y methodology—a convergent design methodology that unifies physical and cyber modelling streams into a single ontological backbone. Grounded in AAS semantics and implemented using the Digital Twin Definition Language (DTDL), the methodology enables scalable, standards-aligned deployment on Microsoft Azure Digital Twins. Its applicability is demonstrated through a real-world implementation in a discrete manufacturing line for melamine board production, where the resulting semantic graph supported modules for simulation, predictive analytics, maintenance assistance, and visualisation. The paper discusses lessons learned, remaining challenges, and future directions, including the formalisation of intangible asset modelling and the automation of semantic pipelines.