Extending two-level information modeling to the internet of things

Interoperability is a major challenge for the Internet of Things (IoT). The real potential of the IoT lies in facilitating largescale sharing of high-quality context-rich information through systems-of-IoT-systems, rather than IoT systems that operate as isolated technology silos. Real large-scale interoperability requires layers of standards, and each layer addresses different interoperability challenges. The SensorThings API data model seeks to tackle data interoperability at the data and informational layers of IoT platforms. SensorThings API is aligned to the ISO/OGC OM data standard, and like OM it is semi-structured. Semi-structured models allow for variance within implementations for different use-cases, which is both necessary and detrimental to systems interoperability.In this paper we propose that the SensorThings API data model should be defined as a set of archetypes, used to capture extensible domain concepts using a two-level modeling IoT systems design approach. Extending two-level modeling to the IoT using the SensorThings API as a base for domain concepts definition allows for a powerful framework to manage variance within systems implementation and maintaining semantic interoperability within systems-of-IoT-systems across diverse use-cases.