Yield Predictions At The Center Of Partnership Between LSU AgCenter, Ag-Analytics
BATON ROUGE, LA.
The LSU AgCenter is partnering with Ag-Analytics to help farmers maintain and improve yields through the use of machine learning and big data.
Ag-Analytics, a secure farm management platform, works with farmers, industry and universities with the goal of helping farmers make better management decisions and mitigate risk.
The data partnership with the AgCenter is a first of its kind with any university.
“The Ag-Analytics-LSU AgCenter partnership will allow farmers to opt into confidentially and securely share data with their land-grant university researcher in a seamless manner with protocols in place to protect privacy,” said Josh Woodard, CEO of Ag-Analytics and an associate professor of agribusiness at Cornell University. “This research and development will directly result in new tools and models that will allow us to better serve our farmers by leveraging Ag-Analytics and the university extension system.”
The system allows growers with a John Deere Operations Center to sign up for free and easily link their agricultural equipment to the Ag-Analytics platform. Farmers who opt into the LSU AgCenter study can choose to share their data securely and confidentially with LSU AgCenter researchers.
Thanos Gentimis, LSU AgCenter mathematician, plans to take the massive volume of data Ag-Analytics collects, apply machine learning techniques to it – teaching a computer program to recognize patterns in the data – and develop models that can predict yield.
“We can turn a tangled set of data into a valuable asset, but we can’t do this with data from just one field,” Gentimis said. “If we have a million acres, then we can train the model to make more accurate predictions and then apply it to a farmer’s field.”
The benefit to the farmers is a driver of the partnership. Improving yield-predictor models also would allow farmers to account for volatile weather, fluctuations in temperatures or precipitation, and other events that have affect yields.
Gentimis said that yield predictions are just the start. He sees these models expanding to offer farmers other management advice such as optimal planting dates and pesticide applications.
Luciano Shiratsuchi, precision agriculture specialist and coordinator of the digital/precision agriculture program at the AgCenter, plans to use data to generate management parameters for farmers who are equipped for variable rate applications.
“With access to as-applied and yield maps from farmers, several research questions can be solved regarding spatial variability, and customized algorithms can be created by the AgCenter to enable farmers to implement variable rate applications without an extra cost with research experiments,” he said.
Farmers can sign up with Ag-Analytics and opt in to share their data with the AgCenter at https://ag-analytics.org/LSUAgCenter The information gathered on individual farms will remain strictly confidential, and personally identifiable information will be stripped from the data sets. Farmers who sign up can get access to resources such as a real-time feed with information about their fields, weather, and insurance estimates. ∆
A screenshot of a precision profit map from Ag-Analytics uses many precision data layers to determine the profit range on sections of a farmer’s field.
Image provided by Josh Woodard