WHAT’S THE PROBLEM?

Weeds growing close to crop plants are the hardest to deal with. Inter-row weeds (between rows) can be removed with cultivators or broad spray systems, but intra-row weeds (within 0-15 cm of the crop) sit too close to the plant for conventional tools to reach without causing crop damage. This is especially true in organic systems, where non-selective herbicides kill crops and weeds alike, and mechanical options are limited by soil moisture and weather. Existing precision spray technologies lack the spatial resolution to target individual weeds in this narrow zone. They work well enough between rows or on roadsides, but the intra-row space demands sub-centimetre accuracy and micro-litre delivery volumes that current systems cannot achieve.

WHAT’S OUR SOLUTION?

We have developed a precision micro-jet spray system that can target individual weeds with sub-centimetre accuracy, delivering micro-litre quantities of herbicide directly to the weed growing point. The system uses an array of micro-jet nozzles controlled by solenoid valves, paired with a high-speed computer vision module that detects and classifies weeds in real time. This allows the sprayer to hit a single weed without contacting the adjacent crop plant as the spray system moves through a field.

This work is a collaboration with the USDA Agricultural Research Service, the Precision Sustainable Agriculture group, and Aarhus University in Denmark. CEI:AgER's contribution focuses on the hardware development and systems integration: the spray delivery system, nozzle array design, and integration with computer vision models. The international partnership brings together expertise in weed science, computer vision, robotics, and agricultural engineering across three continents.

By bridging the gap between existing broad-area precision sprayers and the fine-scale targeting needed for intra-row weed control, this research expands the toolkit of site-specific weed management options. The modular design means the system can mount on tractors or autonomous robotic platforms, making it adaptable across different farming operations and scales.