A team of scientists at the University of Regina created a tool they hope makes farmers’ lives a little easier when it’s time to spray herbicide.
Abdul Bais, an associate professor in the engineering department said the device, known as a SWAT Box, is a sensor box that’s used to collect soil and field data.
He said artificial Intelligence is used to collect information from the soil, water and topography (SWAT) Box along with satellite images and historical field information to determine which kinds of weeds are in fields, crop status and other pertinent information.
Bais said ultimately it leads to more accurate spraying when it’s time to apply herbicides.
“[It] presents that information to the agronomist and then the agronomist will decide which portions, and which areas in the field have which types of weeds, and what is their concentration and what type and what concentration should be applied,” Bais told CBC.
Traditionally, he said an agronomist would explore a farmer’s field, collect data and make their own assessment of the field status.
Bais said a downside to the traditional process is the agronomists’ reports might not highlight in which particular area in a farmer’s field the problems lie, and more herbicide than needed may be used.
Bais said the SWAT Box could help farmers reduce their impacts on the environment and maximize their economic returns.
“If you apply a uniform rate of heavily concentrated herbicide to the entire field, you are wasting the farmer’s money and also, what we’re doing is harming the environment and you’re also harming the crops,” he said.
In the past, Bais said two methods have been used to try and generate more accurate data for farmers about where they should spray: satellite imagery and crop robots.
He said satellite imagery presented a challenge because it was hard to differentiate between what was a crop and what was a weed. Using robots is challenging because the process is hard to scale for the fields of Saskatchewan.
Bais said the SWAT Box as a solution in the middle and essentially an expansion on things that already exist in terms of data collection in, for example, seeding.
“That is scalable and what in this project we want to do is to extend that and apply the herbicides as well,” he said. “This is what makes it profitable for the farmer.”
Bais said the study to create the SWAT Box was sponsored by CropPro Consulting, who he said was testing the method in Saskatchewan and in fields around the world and providing feedback and data to the University of Regina group.