New alternative weeding techniques

Research & Development & Engineering

Tests

Already available with existing resources

Performance evaluation of Perception-Decision-Action weeding systems for in-row weed control

inrae
tscf

Test operator

Challenge ROSE

The ROSE Challenge (“Robotics and Sensors for Ecophyto”) supported from 2018 to 2021 by the French National Research Agency (ANR) has enabled the four selected consortia to develop innovative autonomous solutions for intra-row weed control in wide-spaced field crops (corn or sunflower) and field vegetable crops in order to reduce the use of phytopharmaceutical products by at least 50%, or even eliminate their use.

The OPEROSE project corresponds to the organizational part of the ROSE Challenge carried out by LNE and INRAE’s TSCF Unit.

(Video by Daniel + Virginie Barbosa from LNE!!!)

exemple parcelles rose
Example of a ROSE experimental plot for planting crops

The evaluations were carried out under real agronomic conditions but under controlled conditions. The proposed characterizations covered the following points:

The challenge also opened up new perspectives in terms of :

pr‚sentation premiers r‚sultats 28 nov
Presentation of the first Challenge results on 28 Nov 2019 at the Montoldre site

ACRE Challenge of the METRICS European Project

Following on from the Challenge ROSE, the AgroTechnoPôle hosted the 1st official competition of the Challenge ACRE (Agri-food) of the European METRICS project on its Montoldre site (03) from June 6 to 10, 2022.

The European ACRE competition is organized by 3 partners (i/ Institut POLIMI (Politecnico di Milano), ii/ UNIMI (Universita di Milano), iii/ UR TSCF (Unité de Recherche Technologies et Systèmes d’Information pour les Agrosystèmes d’INRAE)), while LNE (Laboratoire National d’Essais) is responsible for overall coordination of METRICS.

consortium metrics acre
METRICS ACRE Competition Consortium

The second official campaign will take place from May 22 to 26, 2023 in Italy.

exemples de tツhes
Examples of intra-row weed control evaluation tasks with coloured markers at the first stages of the Challenge indicating weeds to be weeded (yellow) and crops to be protected (blue).
plantes d'interet
Crops of interest: beans (left) and maize (right)
mauvaises herbes naturelles
From left to right: "Natural" weeds also sown (at different densities) during the establishment of the crops of interest (lamb's-quarters and chamomile).
mauvaises herbes modšles
From left to right: "Model" weeds also sown (at different densities) when planting crops of interest (mustard).

6 tasks are proposed to the participating teams:

Objective: decide which plants in a row are weeds and which are crops (intra-row detection).

Objective: estimate the leaf area of plants along a cultivated row.

Objective: produce a map of an entire culture by exploring it independently

Objective: estimate the above-ground biomass of crops

Objective: carry out fully autonomous intra-row weed control (i.e. eliminate weeds located between cultivated plants in a row without damaging the crop).

Objective: move around a crop without damaging it

Available by the end of 2024 with new resources in the pipeline

Performance evaluation of Perception-Decision-Action systems via the future research and test bench for characterizing alternative weeding systems

Scroll to Top