If you were to view any of the more than 2 million farms in the United States, some scenes might look familiar: red barns, grazing cows, fields of wheat, rows of vegetables.

But in recent years, new and more unexpected components have entered these images: drones, sensors, and computer systems equipped with powerful machine-learning technology. Using the tools of precision agriculture, or “smart farming,” farmers can optimize irrigation, spot pests and diseases, monitor soil quality, and predict yield.

Integrating new technologies into practices that are as old as human civilization is no easy task, though. One major hurdle facing the field of precision agriculture is convincing farmers to have confidence in the technology. Farmers might worry about whether they can rely on the guidance these systems produce. Or they might worry about their privacy — are the data collected being used responsibly?

Farmers’ technological skepticism is a central concern of Maaz Gardezi’s research. Within his first year as an assistant professor in the Virginia Tech Department of Sociology, Gardezi received a National Science Foundation grant for his project, “Testing a Responsible Innovation Approach for Integrating Precision Agriculture Technologies with Future Farm Workers and Work” — the latest stage of an investigation spanning several years and multiple universities.

The project originated in 2020 as a collaboration between South Dakota State University, where Gardezi was an assistant professor of sociology and rural studies, and the University of Vermont. The two universities established the partnership because of the agricultural differences in their states: South Dakota has a large-scale conventional agricultural system that depends on commodity crops such as corn and wheat, while Vermont sees a larger percentage of small-scale farms and organic growing. The collaboration offered an opportunity to compare two distinct systems of food production.

“Many precision agriculture technologies are being produced with a specific farm in mind,” Gardezi explained. “If engineers continue to collect data coming only from a small segment, these models can be biased. So, you need a diversity of data to train those models, and you cannot rely on just one type of farm.”

Now that he is continuing the project at Virginia Tech, Gardezi sees even more fertile research opportunities. He noted that Virginia is one of the most diverse agricultural commodity producers in the nation, supplying specialty crops — such as peanuts and leaf tobacco — that are not grown as much in other states.

Gardezi and his collaborators not only research how farmers use precision agriculture technologies, but they also facilitate responsible design of the tools themselves, with the help of computer scientists, engineers, and agronomists at the partner universities.

To build tools that will be both useful and used, Gardezi’s team also involves farmers in the process of creating and testing the technologies. “Instead of thinking about farmers only as users,” he said, “we’re thinking about them as co-designers and co-evaluators of these technologies.”

For the primary portion of the study, the researchers are taking a living laboratory approach. With as little interference as possible, their goal is to observe farmers using the new technologies in their natural farm settings.

“It’s important to generate trust with these so-called ‘research subjects,’” Gardezi said of the farmers being observed. “In reality, these are our partners, so we have to think differently about how we do engagement.”

In addition to their observations in the field, Gardezi and his team have developed some creative methods to study how farmers respond to a range of circumstances. For example, the team is developing games in which it will place study participants into various scenarios drawn from the researchers’ observations. This approach allows the researchers to watch the farmers’ decision-making processes in an adjustable environment and to gauge their reactions to changes in the technologies themselves or in other areas, such as public policy.

Gardezi hopes the technologies the project generates will help future farmers better manage their resources with tools they are confident using. The algorithms the researchers build will be openly accessible to farmers.

In addition to providing insights into the farming industry, Gardezi is creating research opportunities for social science students interested in working with precision agriculture technologies. At Virginia Tech, for example, postdoctoral associate Edward Prutzer and junior Megan Schaefer are currently working on the project. Gardezi plans to recruit more students as the project continues, including two PhD students joining the university in the fall. From there, he hopes the project will only grow.

“It’s important for social scientists to go out and learn what’s happening beyond their own disciplines,” he said. “That’s important not just for our field, but also for society, as much of what we learn and understand can have a huge impact in the world.”

Written by Mary Crawford