Objectives
Based on previous research on non chemical weed removal with electric current, the project WeedElec takes advantage of advances in image recognition and robotics to up-grade and automate weeding tasks. WeedElec developed a combination of a drone and a robot with an arm fitted with an electric weeder. The system allows a better detection of emerging weeds so that the robot can eliminate them more efficiently.
Objectives
NA
Additional information
The WeedElect robotic system uses the drone (UAV) to make a rough mapping of emerging weeds by classifying the signals provided by the UAV hyperspectral sensors. These maps are sent to the robot which uses them to plan the weeding tasks. The robot uses RVB camera sensors which send information to an Artificial Intellingence system that identifies the weeds and discriminates them from the crop. Once the weeds are identified they are eliminated with the electric weeder mounted on the arm of the robot. The robot is powered with solar panels allowing weeding operation on day time but also during night, owing to the energy stored in on-board batteries. First lab and field tests show that the system is able to identify and eliminate up to 80% of the weeds in a crop (depending on weeds age and species), saving farmers time and avoiding the impact of using agrochemicals for weed control.
Project details
- Main funding source
- Other public (national, regional) research funds
- Project acronym
- WeedElec
- Agricultural sectors
- Crops (generic)
Project keyword
Resources
Contacts
Project coordinator
-
Irstea – UMR ITAP
Project coordinator
Project partners
-
CIRAD
Project partner
-
INRIA - EPI ZENITH - LIRMM
Project partner
-
INRA - UMR EMMAH
Project partner
-
AGRIAL
Project partner