Aerospace giant Airbus are marketing Verde, an AI-powered crop analytics service, on the Agri-food Data Marketplace recently launched by Agrimetrics, the UK Agri-Tech Centre for Innovation in Data Science.
“Our goal is to provide easy access to cutting-edge data for every field in the UK,” says Dr David Flanders, CEO of Agrimetrics. “Verde will enable users to answer dozens of questions; whether they’re looking at the national picture or individual fields, it will make yield predictions and benchmarking more accurate, improve crop health, irrigation and nitrogen applications.”
Verde uses AI to translate satellite imagery into 15 attributes, including leaf area index and leaf water content. These attributes are available for every field in the UK via Agrimetrics’s Data Marketplace. Furthermore, every attribute is linked to other important metrics, including soil chemistry, historic and forecasted weather, and ecological indicators.
“Satellite data has huge promise for the food and farming sector,” continues Flanders. “Our collaboration with Airbus is making that data easier and faster to access in a format that users want. In some cases, we’ve halved the required processing time.”
Crop Lodging is a major problem for growers. In the worst cases, yields can be decreased by 75% and additional drying costs can reach £7.50/tonne. The cost for the UK farming sector can reach £170m per year.*
“Lodging can be reduced by applying growth regulators, but this is expensive and difficult to manage,” explains Dr Matthew Smith, Chief Product Officer at Agrimetrics and a former Director at Microsoft. “Verde’s attributes could be fed into AHDB’s Lodging risk calculation model, which is freely available.
“Advisors and growers could then manage Lodging on a hectare-by-hectare basis. Growth regulators could be applied only when and where needed – and this could be worked out without even visiting a field. Waste is reduced, whilst yields and profits are improved.”
In this example, just two of Verde’s fifteen available attributes are required. Other combinations of attributes could be used to calculate irrigation and nitrogen requirements, micro- and macro-level yield predictions, and field-to-field benchmarking.
Verde promises significant benefits for service providers and large agribusinesses, for whom remote field scouting promises substantial time and cost savings.
“One UK-based organisation we’re speaking with is responsible for tens-of-thousands of hectares located around the World,” continues Matthew. “The costs of monitoring these crops are prohibitive, so effective management actions can’t always be taken, and yields suffer. On occasion, managers won’t have discovered a crop has failed until harvest.
“Verde will enable this organisation to monitor every hectare from one UK office. They can be automatically alerted to issues in real-time and address these issues before they become problems. Even if they prevent just one field’s crop from failing, this would be a good return-on-investment.”
Verde consists of 15 attributes, including: Leaf Water Content, Green Cover, Leaf Chlorophyll Content, Brown Cover, Leaf Area Index (LAI), NDVI, and Soil Water Saturation. Each attribute has been linked to AI-derived Field Boundaries and more than 1 billion relevant data points, including historic and forecasted weather, soil composition, and a range of ecological indicators.