Realizing the vast, largely untapped potential of the Internet of Things involves more than simple equipment asset monitoring, predictive maintenance, and accumulating distributed sensor data in the cloud.
The IoT’s full value proposition is delivered with real-time automation between decentralized IT systems for singular deployments, like rapidly processing data from smartphones, customer profiles, and real-time weather conditions for personalized discounts when customers near smart price tags.
Such timely action derived from IoT systems in retail, smart cities, and the Industrial Internet requires interoperability: system-wide information exchange and use without express knowledge of collaborating systems. The three relevant levels of interoperability include standards-based, semantic, and sustained interoperability.
Of the three, semantic interoperability is pivotal to optimizing the Industrial Internet of Things because it enables machine understanding of all system data, while leveraging standards-based interoperability to underpin sustained interoperability.
Standards-based interoperability is responsible for the homogeneity in which data are aligned, despite their inherent heterogeneous natures. While currently adopted mainstream standards only rise to structural definitions, the universal standards of this interoperability level expressly solidify uniform reference models for business areas and relevant application activities. In retail, for example, standards-based interoperability allows for connected refrigerators to communicate with food and beverage suppliers for refills. Standards-based interoperability is designed for inter-system exchanges predicated on common models. Examples of universal semantic standards include RDF, OWL, and SPARQL.
Semantic interoperability facilitates data exchanges with unambiguous, machine understandable meaning. This level improves upon the interoperability standards of the previous level by giving explicit meaning to the data involved. The first level allows for disparate systems to communicate; the semantic level enables them to understand what the data are ‘saying’. The semantic level is necessary for intelligent data inferencing, data federation, and true machine intelligence.
This semantic understanding of data is possible because data are modeled uniformly with ontologies—semantic data models based on business comprehension of what data mean. Ontologies use vocabularies and taxonomies (the underlying terminology) to specify the concepts, properties and interrelations of data’s meaning. Because they’re machine understandable in terms business users understand, they’re vital to the inter-machine automation necessary for high value IoT or Industrial Internet use cases.
Sustained interoperability is the final interoperability layer reinforcing network harmonization when changes to models or semantics occur. It involves different layers for detecting such changes, creating strategies to compensate for them, and selecting and applying action to restore network harmony. Semantic interoperability is necessary for this interoperability level to understand what changes have taken place.
In smart cities, for example, there are numerous opportunities to capitalize on these interoperability levels. When civic events occur, intelligent traffic management systems can adjust speed limits and light signals dynamically to accommodate traffic flows, smart grids in the Industrial Internet can modify street lights in the area accordingly, and dynamic pricing can be utilized for real-time changes to germane lots to maximize revenues, transportation efficiency, and public safety.
Semantic interoperability is crucial for communicating these factors between systems for action like raising parking rates or delaying the timing of relevant traffic lights. Sustained interoperability ensures that changes to traffic patterns due to vehicular accidents, for example, are addressed before negatively affecting the event.
Automating the IoT
Overall, standards-based interoperability unifies the methodology of inter-system communication, semantic interoperability implements machine intelligence of the data used, and sustained interoperability keeps IoT networks going. Semantic interoperability, however, is directly responsible for the machine intelligence at the core of the automation described in the above use cases and, without which, the IoT is just another application of big data sans the timely action necessary to make good on it