Computational intelligence for Harvest Forecasting

Brief description

“Granada La Palma” is a Spanish agriculture cooperative operating in Granada, Southern Spain. It is composed of a group of farmers —several hundred—, who are organized through the cooperative for a most efficient production and commercialization. Their main activity is vegetable and fruit production and sale. A key point for their strategic and operational planning is having a forecast as accurate as possible of their harvest. The complexity of the problem lies in the number of affecting factors as well as the variability of them, some out of the farmer controls, e.g. weather.

What service(s) are provided?

After a thorough analysis of the factors —crop, variety, greenhouse specification, weather, …—, DasCI group built different forecasting models for each of the most relevant products. These models include specific date decomposition procedures as well as a combination of Statistical and Computational Intelligence based techniques. Most of them —over 95%— are able to consistently compute forecasts within the error bounds set by the cooperative.

Relationship to digitization

“Granada La Palma “is a company with a permanent interest in boosting is production processes. They have made important efforts in digitization through greenhouses sensoring, data gathering through cloud computing technologies, processing plant sensoring and automation, … The development of the forecasting models, —conveniently integrated through the company’s information system— is one step ahead towards a comprehensive digitization of the company’s activities, by computing key information for their decision making processes at different levels.

Customer and details

Cooperativa Granada La Palma, S.C.A.
Carchuna (Granada) Spain