You are here
Using artificial intelligence to collect agricultural data
ANR OCOD project combines intelligent sensors, drones and optimization for data collection in constrained natural environments
Led by INRAE and involving FEMTO-ST, LPCA and Inria-FUN, this €580k project (2025-2029) began on January 1, 2025, with an official launch on February 6 at INRAE's Clermont Ferrand site.
It aims to optimize the collection of agricultural data in difficult-to-access natural environments using a combination of intelligent wireless sensors and drones. These data will be used to facilitate early detection of water stress or disease, and limit the use of inputs (fertilizers).
FEMTO-ST's work on the ANR OCOD project will focus mainly on optimizing the flight plans of drones so that they can collect aerial images, temperature and humidity data, and communicate with sensors on the ground, without having to stop directly above each sensor.
The aim will also be to optimize the structure of the sensor network, taking into account constraints such as energy consumption.
Karine Deschinkel, along with Mourad Hakem and Jean-Claude Charr from her team, will address these 2 emerging challenges and supervise a thesis on the subject starting in September 2025. They will propose innovative approaches based on recent optimization heuristics and learning techniques, with a full-scale test planned on INRAE's Montoldre experimental site.
Contact : Karine Deschinkel