A hybrid framework for industrial data storage and exploitation
Konstantinos Grevenitis, Foivos Psarommatis, Alice Reina, Weifeng Xu, Ioannis Tourkogiorgis, Jovana Milenkovic, J. Cassina, Dimitris Kiritsis · Procedia CIRP · 2019
In this paper a hybrid framework is illustrated, with a software and hardware integration strategy, for an industrial platform that exploits features from a Relational Database (RDB) and Triplestore using the blackboard architectural pattern, ensuring efficient and accurate communication concerning data transfer among software applications and devices. Specifically, “Raw Data Handler”, manages unstructured data from IoT devices that are kept in an Apache Cassandra instance, while “Production Data Handler” acts on structured data, persisted in a MySQL database. Filtered data is transformed into knowledge and persisted into the Triplestore database (DB) and can be retrieved by expert systems at any time. The proposed framework will be tested and validated within Z-Fact0r project.