Data management in digital twins: a systematic literature review

The Internet of Things (IoT) and continuous advances in data-gathering devices and techniques have significantly increased the amount of relevant data that can be leveraged for innovative real-time, data-driven applications. Digital Twins (DTs) are virtual representations of physical objects, which are fully integrated and in which the automatic data exchange occurs in a bidirectional way. Modern DTs follow a five-component architecture, which includes an explicit Data Management (DM) component that acts as a bridge between the other systems. However, there is no clarity on its role and functionalities. This article presents a Systematic Literature Review on DM solutions proposed in the DT context. We analyzed DM under the Big Data chain of activities to add value to data, highlighting key issues to be addressed: data heterogeneity, interoperability, integration, data search, and quality. In addition to surveying existing solutions for handling these issues, we contextualized them in the domain and function for which the DT was proposed, the type of data dealt with, and the technological infrastructure. Our main findings revealed that the maturity level assumed for the DM component is at an early stage. The most mature solutions were proposed for the industry domain, and many of them assume humans as the ultimate information consumers. Data integration is the prevalent DM issue addressed due to the bridging role of the DM component, and cloud computing is the key implementation technology. Among the research opportunities are reference data management architectures, adoption of industry standards and ontologies, interoperability among distinct DTs, the development of agnostic standard implementations, and data provenance mechanisms.