Skip to Main Content

Creating Intelligent Robots for Farmers

Drones flying over a farm

Things have changed since the “Old MacDonald Had a Farm” days. In agriculture today, robots are bringing precision automation to farming, from planting seeds, to pulling weeds to harvesting crops. As the global demand for food increases, researchers continue to look for new technological ways to help farmers overcome the challenge of producing more food on fewer acres.

A team at the University of North Florida is doing just that. Led by Dr. Ayan Dutta, assistant professor of computing, and funded by a $499,000-plus grant from the National Science Foundation, the team will be developing software that will enable robots to efficiently gather information without any human oversight. For example, as the devices fly over the farm, they would be able to determine areas in need of watering or take pictures of crops to check weed levels, and deliver that data to the farmer, while reducing labor costs.

It’s referred to as a “prototype multi-robot agricultural information collection system.” To add to the challenge, Dutta and his team also will be including data security mechanisms to make the system safe from cyberattacks. “So that part of the problem has never been looked into, and as far as we know we are the first to do that,” Dutta said. “This is a very novel and challenging problem from the computing side of things, and adding the security component adds a great deal to the complexity of the project. I do not know of many other universities doing this.”

Dutta is working in collaboration with UNF faculty Dr. Swapnoneel Roy, associate professor of computing, and Dr. Patrick Kreidl, associate professor of engineering, and three graduate students. The team is also collaborating with a computer science faculty member and one graduate student from the University of Central Florida. Over the three-year timeframe of the grant, there will be future opportunities to include additional students.

To begin, the team is reviewing research about robotic programs used in other applications. “We are trying to discover what other researchers have learned in nonagricultural applications and adapt that knowledge to the system we plan to develop,” Dutta said. “It’s only a beginning because every application brings its own challenges.” 

Technology itself imposes limitations. When considering battery life, for example, Dutta’s team must focus on collecting quality data, not just any data. “Robots don’t have unlimited battery power, so whatever pictures they take have to be the ones that will tell us the best story,” Dutta said. “We are making robots intelligent. Ideally, we want to be able to flip a switch, and the robots will do everything. That’s our long-term goal.”

During development, researchers will conduct feasibility testing in a step-by-step process using computer simulations in the lab. Once the software is complete, most likely in year three, the team will load the software onto purchased aerial robots and test the system at local farms. Dutta sees this grant as an opportunity for all the participants to learn from one another. He said students will be involved on a day- to-day basis, and he’s excited to work with them and learn from them. “It will be a great opportunity for all of us,” he said. “And it’s a great opportunity for UNF as well to win such a large, competitive grant.”