The H2020 DEMETER project uses advanced AI and IoT technologies in order to promote the sustainability of the agricultural sector, to decrease the costs and the environmental impact, while maintaining yields at the same time. To achieve this it developed a service-oriented architecture where various interoperable enablers (i.e., services) can be combined in order to create apps that optimize the usage of resources and yield at the same time. This paper first presents a specific use-case that optimizes rice and maize irrigation and fertilization management processes. Then, it focuses its main part in the development of a data fusion enabler that, combined with appropriate sensor, data analytics, decision support and actuation enablers was combined into the development of DEMETER apps for this use case, that achieved reductions of the used resources (water and fertilizer) while maintaining the crop yields. For the development of this fusion enabler, several preexisting fusion techniques were examined, and the most appropriate of these were used and combined to form the fusion enabler.