.png)
What is the MATAE project?
With the MATAE project, we are modelling crops and pests using differential equations enhanced by AI and data. Using these models, we are developing optimal control strategies based on control theory, reducing the use of phytosanitary products in favour of biosolutions or agroecological levers.
Our work incorporates a global approach, including economic and environmental dimensions, to guarantee viable long-term solutions. These strategies are then tested in the field, on the crops selected for the project, namely vines, beetroot, potatoes and soft wheat. The data collected is used to continually refine our models.
In the case of vines, the best-performing strategies are incorporated directly into our Agronomic Monitoring service, providing the wine industry with practical tools for managing and securing its practices.
The MATAE project has received :
A consortium to promote agro-ecology
The MATAE project brings together a consortium of experts to meet the challenges of the agro-ecological transition.
(Leader) Involved in all scientific fields (modelling, optimal control, differential equations, agro-ecological network).​
Develops advanced host/pathogen models and mathematical tools to optimise control strategies.
Contributes its expertise in viticulture and vine pathologies to the construction of models, while managing field experiments and collecting data.
Contributes its know-how in field crops (beetroot, potatoes, soft wheat) to develop experimental protocols, manage field trials and enrich models.
(Provider) Develops the agro-ecological network, query tools and a participatory platform.
Example of one of the strategies tested on vines: anti-percolation patterns
How can we significantly reduce the use of plant protection products, over and above improvements to equipment, while continuing to protect our plots effectively?
Rather than spraying uniformly, we create small untreated areas in the plot, called patterns, in order to reduce the amount of product used. However, these patterns are far enough apart to prevent the spread of any disease.
The principle of percolation guarantees that no epidemic will emerge if the quantity and layout of the patterns are appropriate. This is in fact the same approach used to combat forest fires, with firebreaks: the fire has a low probability of finding a continuous path through the forest, because the flammable zones are too 'disconnected' from each other.
Our mathematical models of epidemic risk, combined with field data from Agronomy Monitoring, enable us to design spraying maps with anti-percolation patterns in an optimal way, to protect the plot while reducing the total quantity of product used. The result is effective disease management, while significantly reducing your inputs and costs.
How does it work?
-
Before each treatment, we work together to define the level of reduction in phytosanitary products, based on the history of the plot, the state of health, the risk and the weather forecast.
-
Treatment maps are generated so that you can integrate them directly into your sprayer console.
-
Throughout the season, these maps evolve according to data from the Agronomy Monitoring and model predictions.
Already tested on vines and beetroot, this solution is ready to be deployed in your plots for trials starting this year. Switch to anti-percolation pattern treatments that combine innovation, economy and effectiveness!