New “digital all-seeing eye” to make weed spraying far more precise

When farmers drive through their fields, they often lack a precise overview of where weeds are actually located. As a result, spraying is still frequently applied broadly just to be on the safe side.

A new project aims to change that.

With funding of DKK 8.3 million from Innovation Fund Denmark, a research and innovation partnership is developing a solution that can “see” the field in real time and translate observations into concrete actions. The goal is to significantly reduce the use of herbicides without compromising yields.

The project OdinsAEye – Intelligent Vision for Sustainable Farming addresses a key challenge in modern agriculture: the lack of an effective link between what actually happens in the field and the machines that carry out the work.

The solution is a camera-based system, CropCareCam, mounted directly on tractors and implements. As farmers carry out routine field operations, such as sowing, fertilising, or crop care, the camera continuously captures high-resolution images.

Using artificial intelligence, the images are analysed on-site. The system identifies specific weed species and, within seconds, converts the analysis into a concrete spraying plan that integrates with the farmer’s existing digital systems.

According to project leader at Aarhus University, Rasmus Nyholm Jørgensen, the ambition is not to automate farmers’ decisions, but to qualify them:

“We are developing a system that indicates when action is needed and, just as importantly, when it is safe to refrain. If we get it right, we can reduce herbicide use by at least 40 percent without affecting the bottom line. That would be a major step in the green transition,” he says.

The technology behind the system is based on advanced image analysis and so-called agentic AI, which can both identify weeds and learn from previous treatments.

The ambition is not only to detect problems, but also to evaluate the effects of interventions afterwards. This creates a so-called closed-loop system, where field data continuously improves the basis for decision-making.

At the same time, a key part of the project is ensuring that the technology works outside the lab, in mud, dust, and changing weather conditions. The solution is therefore being continuously tested in practice on a Danish farm.

The project is led by Aarhus University and brings together researchers, technology companies, and farmers:

The collaboration is designed to ensure that the technology not only works in theory, but also creates real value in day-to-day farming.