Black-grass (Alopecurus myosuroides) is one of the most persistent and costly weeds affecting UK cereal farmers. Now, Bosch and its consortium partners have concluded a three-year project that could transform how growers tackle the problem. Using high-tech cameras mounted on a crop sprayer boom, the system identifies Black-grass and its growth stage in real time – enabling spot application of herbicide only where needed.
This precision approach not only reduces chemical use and input costs but also helps suppress Black-grass prevalence over time.
AI at the heart of the innovation
The core of the system is an AI algorithm trained to detect Black-grass in field conditions. Rothamsted Research played a key role in the early stages, photographing the weed in controlled environments to help Bosch develop a robust image recognition model. Once trained, the AI could analyse thousands of images captured by cameras mounted on a Chafer Machinery sprayer, identifying infestation zones without needing to target individual plants.
Muhammad Kassem, AI expert at Bosch, said that around 5,000 images across different seasons and crops including wheat and barley were scanned. “As we progressed through the project the model became much more mature to the point where we could detect Black-grass on unseen images with a high level of accuracy.”
Engineering for the field: From lab to boom
As the project evolved, the sprayer setup was continuously refined. The final configuration featured 28 cameras mounted at varying boom heights to maximise detection accuracy.
Peter Frankland, application engineer at Bosch, highlighted the collaborative nature of the work: “The agronomic part of the process was managed by BASF Digital Farming, using its xarvio platform to generate customised maps for use by the Chafer sprayer and to determine the herbicide dose and type. Rothamsted gave us the understanding of types of cultivation that farmers use and some methods of solving the Black-grass problem without using crop protection. We learnt that farmers increasingly adopt no-till or low-till methods and so weed control becomes a more significant challenge. Unlike ploughing, which buries Black-grass seeds and prevents germination, these practices leave seeds closer to the surface.”
Validating results and measuring impact
To confirm the system’s effectiveness, Rothamsted researchers used precisely localised quadrants to count Black-grass occurrences over time. This method provided consistent data showing a reduction in weed prevalence.
The project also demonstrated the feasibility of large-scale image processing and AI training in agricultural settings. Each sprayer pass generated such a high volume of images that downloading and processing took days, requiring custom code to clean, format, and feed data into the AI model.
Funding and strategic fit
Bosch led the consortium, which included BASF/xarvio, Chafer Machinery, and Rothamsted Research. The project received £1.45 million in funding from DEFRA and Innovate UK under the Farming Innovation Programme and UKRI’s Transforming Food Production challenge.
Bharath Jayakumar, director of global key accounts at Bosch UK, said: “This project demonstrates the opportunities we have in the UK to access government funding to enable key innovations for the local market. It also aligns with Bosch Mobility UK’s strategy to grow in sectors beyond our core business.”
