The GAINS project is part of the themes promoted by the NODES Ecosystem, with particular reference to sustainability and digital transformation in the agro-industrial sector. The main objective is to develop a smart agriculture service based on the application of predictive Artificial Intelligence and Machine Learning models to support wheat cultivation. In this context, GAINS promotes advanced digitization of agronomic processes through methods of automated extraction and processing of environmental and crop data, facilitating early diagnosis and the consequent optimization of production practices in terms of efficiency and sustainability.

The project is being carried out jointly by Abika and Primo Principio, companies that have gained solid experience in industrial research, technological development, and innovation at the national and European level, particularly in the fields of precision agriculture and environmental sustainability.

From a technical point of view, the models developed will link satellite data, mechanistic DSS, field sensors, and auxiliary sources (e.g., UAVs) to create an integrated decision support platform capable of assisting the user throughout the entire crop cycle. The project focuses on improving the interpretation of satellite images, comparing them with data acquired in situ, and progressively reducing the manual input required from the user.

The main expected results include: increased fertilization efficiency in relation to yield and protein content predictions; improved accuracy in estimating water, nutritional, and physiological stress; crop growth simulation and qualitative and quantitative prediction of biomass produced; and automatic identification of phenological macro-phases (start, end, and duration of the cycle) through the analysis of vegetation indices, such as NDVI.

Within the project, Abika plays a central role in research, development, and technological integration activities. In particular, it contributes to the design and implementation of machine learning algorithms for the automatic recognition of the main agronomic parameters of wheat cultivation, their validation on heterogeneous data from sensors and remote observations, and their subsequent integration into a prototype DSS platform. This contribution also includes the development of software components, interoperability logic between models and data sources, and end-user-oriented decision support features, strengthening the digital and applicative impact of the project results.