Led by a consortium comprising CIRAD, CNRS, INRAE, Inria and Institut Agro, the AI-Herbage project (Artificial Intelligence for the holistic efficiency of grassland areas in terms of livestock farming and grassland management) aims to develop a comprehensive system for monitoring, data analysis and decision support to improve the management of grasslands, pastures and herds.

Ultimately, this work will contribute to more efficient and sustainable management of grassland areas, a reduction in inputs and the monitoring of biodiversity.

Beyond the project’s launch, these two days highlighted a wide variety of research projects at the interface between agroecology and digital technology, designed to address the challenges facing grassland systems: the optimisation of local resources, animal welfare, adaptation to climatic constraints and the preservation of ecosystem services.

Among the topics presented were:

• Monitoring animals in pastoral conditions using sensors and GPS collars
• Artificial intelligence and robotics: aerial robotics, automated analysis of soil biodiversity, detection of behavioural anomalies and detection of oestrus
• Remote sensing to produce vegetation maps and better characterise pastoral resources
• Modelling the impact of herbivores on agroecosystems
• Decision-support tools to guide and optimise grazing rotations

The project’s challenges

Sensors and AI offer new opportunities to better characterise and make the most of grassland areas. However, their use faces numerous challenges: sensor battery life, data quality, equipment costs, the heterogeneity of environments, and the continuous monitoring of animals.

Bertrand Cloez, research director at INRAE and scientific co-leader of the AI-Herbage project, discusses these challenges.