Combining farm data with other data sources via a digital farming solutions product which includes pest and disease models is helping an increasing number of farmers make decisions about the optimal time to spray.
Digital crop production optimisation platform xarvio, developed by BASF Digital Farming, was launched in January 2020 and is in its second season of use in the UK with 2,500 users.
Bjoern Kiepe, global agronomy lead at xarvio, explains the principles behind the system.
He says: “When we think about offerings in digital farming we always start with data. We want to provide growers with agronomic insights. To do so we need to understand what is happening locally in fields and on their farms. To do this we use a lot of data sources, such as satellites, weather information and farmer information from our Scouting app.”
“We bring this together into an agronomic decision engine which can help with decision making. These can be translated into field-specific actions.”
Establishing a data relationship with growers is key to making these strategies successful, he says.
“When we work with farmers a lot is to do with trust – if we have a trustworthy relationship we will get information from farmers and can build on it. Whenever a grower decides not to continue working with our tools all grower-specific data will be deleted,” says Mr Kiepe.
Digital farming requires strong data competencies, he adds. “We are cloud-only, which allows us to bring new features to farmers faster.”
Artificial intelligence is deeply embedded in what we are doing and helps us develop scalable solutions for various geographies and crops in these markets.
“In the digital world it is important that information runs not only in one direction, but that you also get feedback. One application triggers the next and you need to know what has happened throughout the crop cycle.”
Although models to help understand septoria and yellow rust have existed for a decade or more, better data analysis techniques and bringing data together offer new ways to improve models, says Mr Kiepe.
“All the models we are running are based on data collected from trials and farmer observations. We also have key performance indicators to evaluate the performance of models. We put a lot of effort in to ensure superior model performance.” Growers also have the opportunity to do their own analysis, says Mr Kiepe.
“There are strong features to help farmers grow their understanding of data and the opportunity to do this in a lean and simple way.”
“There is no need to create dozens of folders and copy and paste information – uploading features are intuitive to use.”