Use of Big Data to optimize the volume of wine production.

The PREDIVÍ project aims to improve decision-making related to wine harvests through Big Data.

The growing variability of the volume and quality of wine production means that the investment of resources and the dedication of technical teams to obtain harvest predictions is also increasing. Currently, the technical teams use multiple systems (sampling, maturation controls, gauging, etc.) to ascertain in advance the aforementioned variables of volume and optimal harvest time, but the reliability of the results that these systems provide has a lot of potential for improvement. The large number of variables that affect both the quality and quantity of production (meteorology, plot characteristics, production areas, etc.) makes it very complex to obtain reliable predictions with traditional approaches.

In this sense, Big Data makes it possible to combine a large volume of hyperlocalized meteorological variables (both historical and predictive) and combine them with historical records of production and ripening controls, among other variables. In this way, it is possible to create prediction models based on machine learning that are capable of improving the predictions currently made by technical teams. Other sectors are using prediction models and they have been shown to be able to improve human-made predictions.

Being able to reduce the uncertainty of what the volume of production (whether at the level of the plot, production area, variety, etc.) or what the evolution of the different qualitative parameters will be will help to optimize the planning of the harvest and, therefore, to improve decision making. These decisions are mainly related to aspects of logistics, production volume, sales and production quality.

Eurecat is participating in the project through its Robotics and Automation Unit and its Applied Artificial Intelligence Unit, which will work on the development of prediction models.

The project PREDIVÍ is coordinated by the Catalan wine cluster INNOVÍ. Besides Eurecat, the consortium is made up of Covides, ViTalpe, Unió Corporació Alimentària, Incaví and Raw Data.

General details

Project

PREDIVÍ – Model of wine harvest prediction through Big Data

Project reference

56-21017-2018-2A

Programme and call for tender

Project funded through Operation 16.01.01 of Cooperation for Innovation of the Rural Development Program of Catalonia 2014-2020