I. Introduction
In recent years, with the increasing complexity of internet of things (IoT) systems, there are many lifecycle stages and systems engineering artifacts, such as concepts, development and maintenance[1]. Throughout the IoT system development, stakeholders make use of specific tools to address their concerns [2]. These tools come from different tool venders with different data structures which rely on different databases, platforms and UIs and play an important role during IoT system development [3]. However, the data of the entire lifecycle between such specific tools is isolated from each other, resulting in the inability to achieve data interoperability across domains. Moreover, during the maintenance stage, real-time data from IoT system implementations is also isolated from the development data. This results in two challenges of data integration: 1) to integrate physical space and virtual space of ioT systems which requires more efforts on consolidating heterogeneous data across the whole lifecycle; 2) to enable big data analysis and machine learning for future IoT system development.