VIDALIA — Vidalia onion. Beloved across the country — so much so that it fetched a $200 million farm gate value in 2022 — but grown only in 20 counties around Vidalia in southeast Georgia, where special soil properties produce a sweet onion that is even sweeter.

The conditions that make this region good for growing are also optimal for a wide range of plant pests and diseases. Onion diseases occurring in production fields can cause serious economic losses to onion growers by reducing the yield and quality of marketable onions.

“Georgians are very proud of Vidalia onions. Growers are very successful with current management models and technology, but with more innovative technologies, we can make it big and bigger,” said Luan Oliveira, assistant professor in the Department of Horticulture at the University of Georgia’s College of Agricultural and Environmental Sciences . .

A multidisciplinary team of UGA researchers aims to improve the competitiveness of Vidalia onion growers in Georgia by giving them the ability to confidently detect onion diseases early, allowing them to make management decisions for their crop at a critical time. These skills, the researchers say, should result in increased marketable onion yield and quality, and an overall increase in efficiency and productivity.

Guoyu Lu, an assistant professor in the UGA College of Engineering, is the principal investigator of the three-year project. Lu is joined by Oliveira, who also serves as a precision agriculture specialist for the UGA Cooperative Extension, and Bhabesh Dutta, a professor in the CAES Department of Plant Pathology and a UGA Extension vegetable disease specialist.

“What makes this project particularly exciting is the partnership between CAES and the College of Engineering, which is promoted through the UGA Institute for Integrative Precision Agriculture, the integrated team – research and extension faculty, county agents and farmers – and the use of AI to detect onion diseases,” said George Vellidis, a professor in the CAES Department of Crop and Soil Sciences and institute faculty.

Disease detection and identification create many challenges for Vidalia growers in Georgia. Farmers and their consultants detect and identify diseases by walking the fields. Because many fields are large, scanning each row is a major efficiency challenge. Growers often survey small areas of a field to save time, but this method leaves most of the field unchecked.

When a disease is detected, sprays—either curative or preventive—are applied over the entire field as a precaution, since the extent of disease distribution cannot be easily determined.

“Farming is a very tiring and expensive job — it can be a tiring job with a lot of uncertainties,” Lu said. “I wanted to develop something that would help farmers save labor and time, ultimately making their jobs easier.”

Lu and the research team, together with three leading producers of Vidalia sweet onions, will apply artificial intelligence and machine learning to create a series of disease management support tools for Vidalia sweet onions.

Another advantage of using robots in onion fields is their ability to collect data on plant health and growth. A recent research study used AI tools to count and measure harvested onions. Next, the researchers will measure the onion to test the accuracy of the model. (Photo Submitted)

Combining data collection and analysis with plant pathology, precision agriculture and robotics, the team will build a photographic library of leaf symptoms caused by onion diseases and other physiological disorders, feed them into AI software and use machine learning to identify the diseases based on pattern and color recognition in images.

“Our disease management DSTs will range from smartphone apps to robotic solutions that will allow growers and their advisors to survey entire fields quickly and target prophylactic and curative sprays to areas of the field where disease may occur or appeared,” Lu said.

Growers and their consultants will be able to use the SmartDetect app to identify diseases that can be difficult to identify without specialized knowledge or equipment, including bacterial leaf blight, pink root or onion root.

“Additionally, the SmartDetect app can be used to track disease outbreaks over time and across locations,” Lu said. “By collecting data on the prevalence and severity of various diseases, growers can make informed decisions about how to manage their crops and prevent future outbreaks.”

Another advantage of using robots in onion fields is their ability to collect data on plant health and growth. Using sensors and cameras, the robots can collect information on factors such as soil moisture, nutrient levels and plant health.

The project aims to increase the competitiveness of Georgia’s Vidalia sweet onion growers by increasing efficiency and profitability. Allowing growers to confidently detect onion diseases early and target disease management will increase the ecological sustainability of Vidalia sweet onion production by reducing the volume of pesticides used to grow the crop.

The project, assigned at the end of 2023, has just completed its first year and will be completed in September 2026. The team will now send all the data to the developer of the disease recognition and localization app. Lu and his colleagues hope to expand the technology to other crops in the future.

“As an extension specialist, my job is to take this research back to the farmers,” Oliveira said. “Having something tangible to support our Georgia producers is incredibly exciting.”

Learn more about research projects within the UGA Institute for Integrative Precision Agriculture at iipa.uga.edu.