Contexte

The meteorological variability generated by climate change causes uncertainty in the development of crops, increasing the difficulty in crop planning, particularly in fruit farming. Greater variability in the volume and quality of peach and cherry production requires higher investment in resources and work from technical teams in planning the harvests, without providing greater precision.
Technical teams use different techniques (sampling, ripening controls, measurements, etc.) to determine the aforementioned variables of volume and optimum harvesting time in advance, but the reliability of the results provided by these systems has much room for improvement.
The large number of variables affecting both quality and quantity of production (such as weather conditions, field characteristics, production areas, etc.) mean that obtaining reliable predictions using traditional approaches is a very complex task.
The FRUITFORECAST operational group made up of the FRUITS DE PONENT cooperative group and CERIMA CHERRIES, a fruit and vegetable company specialising solely in the production, packaging and export of cherries worldwide, in collaboration with the IRTA Research Centre and the company RAW DATA, which specialises in big data technologies, has developed a tool based on predictive models that anticipate information on changes in quality parameters and harvest volumes for the peach and cherry sector in order to improve reliability in planning harvests.