RoboWeedMaPS combines the latest groundbreaking research in Deep Learning and Big Data in a precision agriculture context. It aims to significantly reduce herbicide usage in Denmark initially, but will later be applicable worldwide. The project will address the issues that have so far limited the uptake of precision weed management such as inefficiencies relating to the recognition and mapping of more than 100 different weeds – all of which can reduce yield potential. It is estimated that savings made as a result of the project will be in the region of 40% for standard farming practices, up to > 80% for emerging high-tech farmers that use the new generation of injection sprayers.
Several commercial products will be developed and matured during the project. The core products are: a high-speed camera that automatically determines the weed populations in the field; a cloud service used for uploading images and weed recognition; a Farm Management Information System for providing the farmer with recommendations based on the weed populations of his fields; and finally, an injection spray boom capable of changing the herbicide mixture and dosage on-the-fly based on the detected weed population at the specific spot in the field. The products can operate individually but generate significant synergies when they all interact in the full RoboWeedMaPS chain. All the products are well fit for export via the separate distribution channels already established especially in Europe by the project partners.
The commercial project partners span wide: from sensor and deep learning systems over cloud based systems, Integrated Weed Management decision support systems, Farm Management Information Systems (FMIS), and finally a herbicide sprayer manufacturer, closely supported by Aarhus University within engineering, market analysis, and agroecology.
The project is financed by Innovation Fund Denmark.