COSIBAS targets the next step in IoT-based applications and solutions, namely the integration of semantic and cognitive AI technologies. The Digital Transformation in the industrial domain is currently limited to the connectivity of devices, machines, tools, workers etc. The amount of data is rising and requires a sophisticated interpretation through analytics to generate business value in terms of faster detection, better forecasts, and improved decisions with an overall increased flexibility. Current IoT-stacks are frequently focused on handling of data or data streams – semantic properties and semantic descriptions or ontologies which can be utilised by cognitive services are generally missing.

COSIBAS focuses on existing IoT-technologies to create an open, standardised approach to define and develop cognitive enablers and cognitive services augmenting the IoT-stack. Enablers provide additional semantic descriptions or can translate semantic information into IoT-data. Depending on the capability of involved IT-structures, this may contain approaches like SDM leading to CEMs. Based on these enabling factors, cognitive services will be able to interpret and evaluate the status of the whole system to build and transact on a cognitive model, which is the basis for the interaction with the human blue and white-collar workforce providing cognitive assistants, decision-support as well as to support intelligent mechanisms to enable seamless interoperability respecting semantic.

Due to the high complexity of such scenarios COSIBAS will rely on existing IoT-technologies and standards (e.g. FIWARE) and evolve existing semantic and cognitive technologies into a set of enablers and services based on proven elements of AI approaches and components. As such, COSIBAS utilises knowledge-trees, ontology languages, and inference & rule engines in order to create accurate models and implementations to enhance IoT-related systems. Standardised management of the cognitive-APIs will ensure high reuse and marketability.