PGD du projet EVAGRAIN "Des outils intelligents pour une utilisation agile du blé"

Cereals are an essential resource for human and animal nutrition. As the leading European exporter of bread wheat, France sells more than 50% of its annual production on international markets and regularly ranks among the world’s top five exporters. However, it faces strong competition from countries whose agronomic practices, often more extensive, result in lower production costs and naturally higher protein contents—still a key criterion of bread wheat technological quality and a major determinant of breadmaking performance.In France, although breadmaking represents only 25–30% of bread wheat end uses (the remainder being devoted mainly to animal feed and starch production), baking performance continues to structure quality classification and pricing, both at farm level and on commodity markets. Quality assessment relies on a limited set of indicators: protein content, test weight, damaged starch content, parameters obtained from the Chopin alveograph or the Brabender farinograph, and, as the ultimate reference, the standardized breadmaking test NF V03-716. This test, carried out under standardized conditions and evaluated by expert bakers using a complex scoring grid, is time-consuming, costly, and dependent on scarce expertise. Nevertheless, none of the conventional indicators currently predicts its outcome with sufficient reliability to replace it.At the same time, climate change, diversification of cropping systems, and evolving societal expectations are increasing the qualitative variability of wheat harvests. This heterogeneity complicates batch homogenization, silo segregation, and optimization of milling and baking processes, highlighting the need for a more detailed characterization of raw materials. Many studies have shown that technological quality cannot be reduced to protein content or composition alone: lipids, arabinoxylans (pentosans), enzymatic activities (α-amylases, peroxidases, polyphenol oxidases, etc.), and interactions among constituents all play important, yet still poorly quantified, roles.Moreover, current evaluation methods struggle to integrate these data in a comprehensive and predictive manner. There is therefore a strong need for new descriptors, integrative models, and faster, more objective, and more robust decision-support tools to maintain the competitiveness of the French wheat sector in a context of agroecological transition, increasing harvest variability, and international competition.The EVAGRAIN project aimed to provide the French wheat sector with a more robust and predictive quality evaluation system, better suited to a wide range of end uses, under conditions of increasing climate-driven variability and international competition. Two complementary objectives guided the work:To develop a Decision Support System (DSS) providing an integrated assessment of wheat quality, primarily for breadmaking but also for other industrial applications such as biscuit production.The initial goal was to predict the results of the standardized breadmaking test by combining all available sources of information: literature data, baker expertise, existing databases, and new data generated within the project. For biscuit production, where no reference test currently exists, the aim was to identify and formalize relevant criteria in order to propose a comparable evaluation framework.To develop innovative analytical tests that integrate additional determinants of quality not routinely considered in standard evaluations. In particular, the work focused on so-called “minor” components such as lipids and pentosans (arabinoxylans), whose effects on dough properties and final product quality are well established but not routinely measured in flour batches. Given the central role of gluten proteins, advanced characterization of their composition and variability was also necessary to better link flour composition, technological test parameters (alveograph, farinograph), and breadmaking performance.Finally, the project sought to take advantage of new data integration approaches, particularly those based on artificial intelligence, to improve the interpretation of technological test results. These approaches open the way to faster, more objective, and more reliable tools to assess wheat quality and support decision-making throughout the value chain.