Model-driven development of interoperable communication interfaces for FAIR sensor services

Abstract

To gain long-term knowledge on product and process quality the sustainable collection, processing and storage of sensor data is an essential requirement. Whereas the application of the FAIR principles as a guideline to provide all collected data with rich metadata that allow for finding, accessing, interoperating with and reusing the data effectively is constantly growing within science and research, the usage in an industrial scenario is largely unexplored. Today, collected industrial data is often stored in arbitrary data formats without appropriate metadata describing the data content and is therefore lost for future reuse because crucial information on how to find, access and interoperate with the data is missing. Moreover, insufficiently described or missing data can lead to wrong decisions. In order to implement the FAIR principles for industrial sensor data, this article derives the major deficits and challenges of making industrial sensor data FAIR. Furthermore, the authors propose a three-layer architecture of FAIR sensor services, which handles the discussed challenges, to acquire FAIR sensor data at the time of measurement. The conceptual draft is evaluated by a prototypical implementation of a FAIR sensor service.