ABSTRACT: In this paper are discussed some results related to an industrial project oriented on the integration of data mining tools into Enterprise Service Bus (ESB) platform. WSO2 ESB has been implemented for data transaction and to interface a client web service connected to a KNIME workflow behaving as a flexible data mining engine. In order to validate the implementation two test have been performed: the first one is related to the data management of two relational database management system (RDBMS) merged into one database whose data have been processed by KNIME dashboard statistical tool thus proving the data transfer of the prototype system; the second one is related to a simulation of two sensor data belonging to two distinct production lines connected to the same ESB. Specifically in the second example has been developed a practical case by processing by a Multilayered Perceptron (MLP) neural networks the temperatures of two milk production lines and by providing information about predictive maintenance. The platform prototype system is suitable for data automatism and Internet of Thing (IoT) related to Industry 4.0, and it is suitable for innovative hybrid system embedding different hardware and software technologies integrated with ESB, data mining engine and client web-services.
KEYWORDS: ESB, Data Mining, KNIME, Industry 4.0, Predictive Maintenance, Artificial Neural Networks (ANN), MLP neural network.