BigDataGrapes D8.4 - Integration and Operation with real-life Practices

The deliverable D8.4 “Integration with existing real-life Practices” showcases the application of BigDataGrapes technologies in data-intensive and critical operations related to the Natural Cosmetics Pilot and the Food Protection Pilot.

More specifically in the case of Natural Cosmetics Pilot, the ultimate goal was to prepare a software platform with the form of a dashboard that will expose the required functionality to practitioners in the grapevines and the end-users of relative cosmetic industries, during realistic operations and processes of these stakeholders. The dashboard uses the results of predictive analysis over data samples managed by expert users and showcases the ability to help decision-making by end-users based on a small subset of exhibits in comparison to the amount that a human should manually check. In this final version of the deliverable the scope is to showcase the initial hypothesis that BigDataGrapes (BDG) software stack could serve for the prediction of the Biological Activity (BA) parameters by incorporating their correlation analysis to Satellite-based Spectral Vegetation Indices (SVIs), while, as an additional feature, to incorporate also Weather Data (WD) for correlation analysis with BA parameters. This final version of the deliverable presents the development of the dashboard and its visualisation developed during the lifetime of the project, and its enhancement with additional input data and visualisation outputs, according to end-user needs, to address their critical decisions in natural cosmetic industries.

D8.4, “Integration with existing real-life Practices”, is based on the piloting plan of the Natural Cosmetics Pilot (SYMBEEOSIS) and the BA data collected from samples all around Greece, with GEOCLEDIAN providing the SVIs datasets, SYMBEEOSIS also collecting the WD from meteorological stations of Institute for Environmental Research (National Observatory of Athens), CNR undertaking the data correlation analysis, Ontotext the data modelling, Agroknow the data management and their appropriate transformation for uploading to the software stack, and KU Leuven the visualisation of the dashboard.

In the case of Food Protection pilot, the goal was to deliver to the food safety and quality assurance (FSQA) expert an online platform, namely FOODAKAI, that a) can monitor risks associated with any supplier, any ingredient or any product, b) can be customized to serve everyone in your safety, quality & compliance teams and c) will reduce by 50% the time devoted to food risk monitoring & assessment tasks. The FOODAKAI platform uses the Big Data software stack developed in Big Data Grapes to collect large amount of food safety incidents, to process this data, to enrich and to build prediction model that can help the FSQA experts to prevent food safety incidents in the food supply chain. D8.4 “Integration with existing real-life Practices” presents how the FOODAKAI platform was integrated into the real-life practices of the FSQA departments of food companies and how it can help them to move from reaction to prevention using the food safety analytics and risk predictions.