Digital Twins as Engines of Industrial Digital Transformation: Enabling Technologies, Applications, and Adoption Challenges

The digital twin, a virtual representation of a physical asset or process kept synchronized with its counterpart through a live flow of data, has moved in little more than a decade from a conceptual curiosity to one of the central enabling technologies of industrial digital transformation. This review synthesizes the state of the field for an automation and digital-transformation audience. We first clarify the concept and the widely used taxonomy that distinguishes a digital model, a digital shadow, and a full digital twin by the degree to which data flow between physical and virtual is automated. We then examine the technologies that make twins possible, the Internet of Things, cloud and edge computing, simulation and physics-based modelling, big data, connectivity, and above all machine learning, and we map the principal application domains, from manufacturing and predictive maintenance to smart cities, healthcare, energy, and supply chains. Across these domains the recurring value proposition is the same: a twin turns data into foresight, allowing operators to monitor, predict, and optimize rather than merely react. We give particular attention to predictive maintenance, where the integration of machine learning with digital twins has produced the clearest returns, and to the human-centric turn of Industry 5.0, in which twins increasingly model people and their interaction with machines. Finally, we analyse the barriers that keep adoption uneven, especially among smaller enterprises: data quality and interoperability, the absence of shared standards, model fidelity, cost, a scarcity of skills, and cybersecurity and privacy risk. We argue that the technical foundations are now mature enough that the frontier of value has shifted from feasibility to disciplined implementation, standardisation, and organisational readiness. The review is narrative rather than exhaustive and is weighted toward manufacturing, where the literature is deepest.