Abstract

This paper proposes an internet of things application with intelligent data processing, which can be applied in several industrial sectors. The designed application allows the data collection, storage and visualisation. An IoT gateway is implemented on a Raspberry Pi and functions as a local server. The IoT gateway allows access to the information from the sensor nodes, even without an internet connection. The information processing is performed on a big data server in the cloud, using Apache Spark, Apache Kafka and the IoT ThingsBoard platform. In previous work, the IoT application architecture was implemented in tomato crop greenhouses and a database was obtained. The objective of this research is to evaluate the potentialities of processing the acquired data using the proposed IoT application architecture. Additionally, two public databases are used for experimental tests of the intelligent data processing. Different machine learning algorithms available in Spark was applied and a selection of those that provide the best results is made.