That’s going to make for interesting decision making according to Bret Sitzmann, global director of product marketing for The Climate Corp.
Sitzmann says that gives added value to Climate’s new predictive seed selection and placement technology, called Seed Advisor, that has shown strong results for farmers testing the tool in Iowa, Illinois and Minnesota. He says the company is set to expand pre-commercial testing to Wisconsin, Indiana and Missouri in 2019.
Seed Advisor is a data science-driven tool that uses historical region and field data performance to rank every available hybrid from best to worst for that particular field. It then provides farmers recommendations for which hybrid to plant on each field along with an optimal seeding rate recommendation.
“The idea is to create an advanced seed prescription for that field, complete with the best hybrid and the best planting rate from zone to zone,” Sitzmann says. “You want to get the right hybrid at the right population on a sub-field basis.”
2018 Harvest results from Seed Advisor field tests demonstrated an average yield advantage of 9.1 bushels per acre versus what the farmer would have planted without Seed Advisor recommendations, with a more than 80% win rate.
Sitzmann says field testing for the latest model of Seed Advisor is underway, which looks at 160 variables in the field to help it calculate hybrid placement.
Climate is also building a model that helps crop advisers determine which fields respond better to management.
“We can train the model to be predictive,” he says. “It looks for farm fields that match the geographic variables of our research and development fields, so we can send an email to farmers telling them which fields will be the most responsive to advanced seed prescriptions.”
The model is currently focused on corn hybrids, but work is being done in soybeans as well, he said.
Sitzmann says Bayer has been doing corn seeding research for the better part of a decade and has data on about 41,000 acres where they consistently get a good response.
“We give the model a list of features, including historic weather data, soil types, textures, topography and elevation, and ask it to identify which features are the most predictive,” he says. “Consistently, that is weather: temperature and precipitation at critical crop growth stages.”
He says the goal is to build the model on research data, then turn it loose on real fields to allow Climate to be really specific about which of its products fit best.
“For years we have been collecting data and we have a lot of digital records,” he says. “We are starting to leverage all that data to become more predictive. If you can look back decades instead of just a few years, it gives you better odds of winning.”
Sitzmann says he enjoys the opportunity to talk to a lot of farmers, and he enjoys using a combination of all the data and the farmers’ local knowledge of their fields.
“When you put those data sets together, we can give great advice on the best way to put the right seeds on the right acres,” he says.