Real-Time Processing of Heterogeneous Data in Sensor-Based Systems

Nowadays the applications for real time processing of large amounts of data are encountered increasingly more frequently, as there are lots of system's types that can generate large comprehensive information in a relatively short time. In this paper we focus on sensor-based systems. Such systems may be found in several important domains, such as smart farming, medical field, water management, or smart cities. The proposed solution in this paper has the capacity to analyze data streams from different sensors but also considers historical data, in order to provide alerts or invoke different services. This is a new approach, as, to our knowledge, none of the existing stream-processing solutions support combining streaming with batch processing data. We tested our solution with data from sensors and actuators, using a smart farm test scenario.