Field of data: Drones, sensors help horticulturalists know as they grow

When it comes to horticulture, knowing what a plant is experiencing before it starts to show yellow leaves, droopy stems or a low yield can make all the difference in having a productive growing season.

A resilient plant, despite what viewers can see, may be under stress from drought, over-watering, insect infestation, mildew, fungus, or a myriad of other issues. By the time their appearance reflects the problem, valuable time – and produce – may be lost.

For Harmandeep Sharma, Ph.D., what’s hovering way above the ground can tell her a lot about the condition of what’s below.

“Drones are an amazing tool,” she said. “They are non-destructive and can provide real-time, and accurate information regarding crop status and assist in efficient decision-making process. And not all of them are expensive – some are inexpensive, that can be used by farmers for quick crop surveillance to detect stressors.”

As she operates a controller, a drone equipped with a multi-spectral camera flies over a row of plants, capturing leaf surface reflected radiations – a valuable indirect measure of the plant’s health.

“The drone camera is looking at a near-infrared light band that we can’t see,” Sharma said. “The reflected near-infrared radiation varies based on the health of the plant. For example, shades of red indicate good crop health. A yellowish color indicates unhealthy crops that either need more water or a soil amendment.”

Sharma heads the Digital Agriculture Research Lab within the Department of Natural Resources and Environmental Design. Her team employs a trifecta of modern technology – sensors, drones, and machine-learning algorithms– to design data-driven, sustainable crop production systems. It’s another facet of “digital agriculture,” the use of technology to help agricultural practice.

“I’m especially focused on specialty crops that generate a high-value return,” she said. “Hot peppers are one example, and we’re looking at several different varieties to identify the most suitable pepper cultivars for the region. Technology plays a pivotal role in determining the optimal ways to apply water and fertilizer, which not only enhances crop quality but also saves growers valuable time and resources.”

In the field, sensors measure weather data, such as precipitation, solar radiation, relative humidity, and air-temperature fluctuations. Sap flow sensors offer real-time, direct measurement of plants’ water use, while a portable photosynthesis system tells how effectively a plant is photosynthesizing, or ‘making food,” a key factor influencing its yield. It also tracks transpiration, which reveals its rate of water loss from the leaves.

It’s a comprehensive, multidimensional sensor system that lets the team tailor production-management practices to local climatic conditions, to maximize crop production and quality, she said.

Harmandeep Sharma, Ph.D., right, CAES research assistant professor, leads her students, from left, Janyah Hardison (undergrad), Edmond Kwekutsu (grad), and Kenene Lee (undergrad), in the use of their new tech equipment as they do ag research in the Reid Greenhouse.

Sharma teaches the technology through training sessions and workshops for the community, and classes for her students. Her next research project involves a “sensor-fusion” approach, an innovative combination of aerial and terrestrial sensors that promises greater insights into plant physiology and responses under changing climatic conditions.

“The only limitation lies in the human capacity to take measurements,” Sharma said. “Analyzing drone images requires specialized skills and knowledge, a task not all growers are currently equipped to do. My goal is to train the next generation, especially underrepresented and female students, and farmers, in mastering these technologie so that they know how to use tech better for what is rapidly becoming a data-centric field.”