CYBELE is a H2020 EU funded project, aiming to generate innovation and create value in the domain of agri-food, and more specifically in the sub-domains of Precision Agriculture (PA) and Precision Livestock Farming (PLF). The technological solutions proposed by CYBELE will be demonstrated in 9 real-world industrial cases ('demonstrators') in the agri-food, viticulture, livestock and fish-farming sectors.
Taking into consideration that agriculture is a high-volume business of significant social and financial impact with low operational efficiency, CYBELE aspires at demonstrating how the convergence of HPC, Big Data, Cloud Computing and the IoT can revolutionise farming, reduce scarcity and increase national food supply levels, leading to enormous social, economic and environmental benefits. CYBELE intends to safeguard that the stakeholders involved in the agri-food value chain (research community, SMEs, entrepreneurs, etc.) have integrated access to a vast amount of large-scale datasets of diverse types, and that they are capable of actually generating value out of this data. For this purpose, secure and unmediate access to large-scale HPC infrastructures supporting advanced data discovery, processing, combination and visualisation services, solving computationally-intensive challenges modelled as mathematical algorithms requiring very high computing power and capacity.
CYBELE will develop large-scale HPC-enabled test beds and deliver a distributed big data management architecture and a related data management strategy providing 1) integrated, unmediated access to interoperable large-scale datasets (including their metadata) of diverse types (sensors, satellite, aerial images, etc.) from a multitude of distributed data sources, 2) a data- and service-driven virtual research HPC-enabled environment supporting the efficient execution of multi-parametric agri-food-related impact model experiments, optimising the specific features of processing such datasets and 3) a bouquet of domain-specific and also generic services on top of the CYBELE virtual, industrial experimentation environment facilitating the elicitation of knowledge from big agri-food related data, addressing the issue of increasing the responsiveness and empowering automation-assisted decision making, empowering the stakeholders involved to use resources in a more environmentally responsible manner, improve sourcing decisions, and implement circular economy solutions in the food chain.
CYBELE aims to expand the traditional boundaries of PA and PLF by performing climate simulations requiring fusion and analysis of satellite-derived earth observation time-series together withe climate forecasts, video streams, sensor information, plant genomics and more diverse data sources, within a short period of time. Moreover, it aspires to tackle issues, which can only be solved by exploiting the storage, computing memory and throughput capacity of HPC infrastructures and not using conventional intensive computing resources. It intends to to facilitate the extraction of advanced and more accurate insights, empowering the execution of HPC enabled computationally and memory intensive big data analytics, including quasi-Newitn optimisation techniques like BFGS, parallel matrix multiplication, deep convolutional neural networks (CNNs) , Hidden Markov Modelling combined with non-parametric methods.