It uses advanced machine learning to help farmers and agronomists pinpoint what their crops and soils need to boost yield in a sustainable way.
The scientific teams of Bayer Crop Science and Biome Makers tested and disclosed the first application of this ground-breaking technology on bioRxiv, the preprint repository operated by Cold Spring Harbor Laboratory. The study and resulting scientific paper details the analysis of the soil microbiome to assess the effectiveness of Bayer´s biological fungicide Minuet®. Specifically, machine learning software allowed Bayer CS to predict potato yield improvement before the application of the input. The predicted result was a yield bump of up to 40 per cent in one of the fields tested in Idaho.
“It’s a unique approach to utilise soil biology and optimize the use of crop inputs moving forward towards sustainable and economically favourable solutions to improve crop productivity,” said Varghese Thomas the project leader at Bayer CS.
Agronomists should be happy. This technology is a giant leap forward for them who, up until now, have lacked the data required to accurately determine biological solutions for their seasonal soil and crop decisions. Soil is a valuable asset to increase crop yield and quality, but as it currently stands, agronomic recommendations are based on little knowledge of the biological processes occurring in it. But today, the availability of an AI virtual assistant to help predict the effect of different solutions is game-changing, and progress towards a more productive and sustainable agriculture system.
AI is an ever-evolving resource and, as such, is currently being “trained” to resolve other farming concerns as well, including questions about produce shelf-life, nutrient quality of the produce, and projected carbon credits based on the use of different products or management practices. Input manufacturers can add their own, custom solution to the AI recommendation system by testing it under the strict Gheom® field trials protocol, a service Biome Makers offers.
Currently, select growers in the US and EU can test the recommendation system with complimentary functional soil analytics for their fields.
