Women in Ag: Lirong Xiang

Some day in the not-too-distant future, we might see swarms of robots in a field of soybeans, killing pesky weeds with tiny amounts of herbicides. There won’t be a human in sight because these robots will be fully autonomous, able to maneuver a farm field and identify a weed using artificial intelligence.

When that day comes, it will be because of Lirong Xiang, assistant professor in the Department of Biological and Agricultural Engineering at NC State University. Xiang is this year’s winner of the Foundation for Food and Agriculture Research (FFAR) New Innovator Award, which is granted to junior faculty whose research shows immense promise toward addressing serious challenges in food and agriculture. To develop this high-tech robotic weeder, Xiang is collaborating with colleagues Rachel Vann, Chris Reberg-Horton and Steven Mirsky alongside many industry partners. 

In the U.S. alone, weeds cost farmers billions of dollars in crop loss every year, making them one of the biggest threats to our food supply. They are mainly dealt with in two ways: chemicals and human labor. The heavy use of herbicides, while sometimes necessary, has led to weeds that are developing herbicide resistance quicker than scientists can develop new versions. Herbicide use can also harm the environment and human health. Xiang is on a mission to find a better way to deal with the problem.

Xiang grew up in a rural area of China where agriculture was very labor-intensive. Inspired to engineer agricultural solutions throughout her studies, Xiang, one of the few female engineers in her classes, competed in agricultural robotics competitions all over the world. She became the first person at NC State to lead a winning agriculture robotics team and develop an agricultural robotics summer camp for high school students. With support from the department, Xiangl serves as an inspiring mentor, attracting and motivating students to the area of food and agriculture, and educating next-generation agricultural scientists.

In her Automation and Robotics Lab, Xiang isn’t just identifying problems that can be solved with AI: she is actually solving them. Micro-applications of herbicides have been proven in previous scientific studies, but no one has come up with a robotic solution – until now. Xiang’s robotic weeder will go many steps beyond the current application method of broadcasting chemicals with large agricultural machinery.

This robotic weeder of the future is a lightweight ground robot retrofitted with various sensors and motors that give it the dexterity and finesse to identify weeds and apply small amounts of herbicides directly on an individual weed. The prototype is currently manually controlled with a joystick, which is not practical for a farmer with hundreds of acres of crops. The prototype is also best suited for more controlled environments like greenhouses. The new and improved version, which she will research using the FFAR New Innovator Award funds, will be autonomous and better able to handle the complexities of an agricultural field, like wet soil and wind. It will also, most notably, reduce the need for herbicides by 99.99%.

“We will train AI models to take sensor measurements, analyze the data and determine the right control commands for each motor of the robot,” Xiang says.

Xiang plans to commercialize the technology through patents and partnerships with agricultural companies. Farmers, Extension specialists and commodity board leaders have also shown interest in bringing her robots onto North Carolina farms. As such, these robots need to be affordable. Xiang estimates that one robot costs around $5,000, but her lab is trying to reduce costs even further. Other robotic weeding solutions on the market can reach six figures. Xiang anticipates farmers will save money with her technology by reducing labor and chemical use. Part of her research will also involve developing “swarm intelligence” so farmers can release a multi-robot fleet fit to handle hundreds of acres with energy efficiency.

“I just love the feeling of figuring out an issue when I build a robot and then seeing the robot running in the field,” Xiang says.

We’ll be seeing a lot more robots running through the fields of North Carolina in no time.

This post was originally published in College of Agriculture and Life Sciences News.