Precision and Digital Agriculture: A promise that can revolutionize farming practices

A man crouches down in a field while holding a drone controller in front of a drone parked in the dirt.

Dulis Duron, LSU College of agriculture graduate student, is working with sugarcane farmer Keith Dugas to test digital ag tools that will help predict sugarcane yields on Dugas’ farm.

It’s the promise of less time spent walking the field. It’s the guarantee of pinpointing water or nutrients needs in a field. It’s the hope of saving money, time and resources.

Precision and digital ag tools and techniques offer a lot of promise to farmers, and the LSU AgCenter is aiming to make that promise a reality through new hires, innovative projects and strategic investments.

Precision ag and digital agriculture encompasses alternative uses of GPS, drone technology or autonomous robots combined with the application of artificial intelligence (AI) to analyze comprehensive data sets for improved farm-level decision making including projects like the ones detailed here that the LSU AgCenter is prioritizing.


A man sits in front of a computer monitor with a keyboard on the desk.

Congliang Zhou developed an app that can help strawberry farmers estimate the number of certain insects affecting their crop. Photo by Kyle Peveto

Seeing more clearly what the eye cannot

Two-spotted spider mites are a top concern for strawberry farmers. These tiny insects attack plants, causing leaf cells to collapse and die.

The bugs grow less than a millimeter in length, but accuratel counting the pests can help farmers decide the optimal method for controlling them.

“Most people — at least me — can’t identify them without a magnifying lens,” said Congliang Zhou, who joined the LSU AgCenter this summer as an assistant professor for research and extension.

As a doctoral student at the University of Florida, Zhou used his experience with artificial intelligence and geographic information systems to develop a smartphone application that counts the mites on strawberries.

Zhou’s work in the AgCenter School of Plant, Environmental and Soil Sciences focuses on precision agriculture to make agricultural operations more efficient. He will work with agricultural producers statewide to improve their use of these technologies and develop new tools based on their needs.

Precision agriculture has increased the efficiency of farms worldwide over the past 40 years, and Zhou sees a greater need for labor-saving technologies in the future.

“The labor shortage is a major issue for the modern farming system,” he said. “We cannot find enough people, especially in the United States.”

In his work with the AgCenter, he envisions many additional uses for the smartphone program and plans to collaborate with entomologists and pathologists to broaden its applications.

“It can also be adopted for other things in our department, for nutrient estimation, water stress detection, monitoring, these kinds of things,” he said.

While Zhou has experience with drones using remote sensing technology and has developed smartphone applications, he also sees new frontiers for technology in agriculture. Over the past few years, he has begun developing a robot that can crawl through fields and take multiple measurements that are important to producers.

At the AgCenter, Zhou will work in an extension role along with his research. He plans to meet agricultural producers and learn what needs they may have and then develop tools to make their operations more efficient.

“I like to listen to people, talk to people and then bring the precision agriculture technology to the people,” he said. “I really like to listen to them, talk to them, communicate with them and then help.”


A man stands with his arms folded across his chest.

Above: Thanos Gentimis. Photo by Eddy Perez. Below: Precision replanting map based on YOLOv8 result for the LSU AgCenter Doyle Chambers Central Research Station in Baton Rouge in 2023. Image made in QGIS

Standing out with soybean stand counts

Stand counts are an important part of ensuring optimum soybean yields. Farmers often repair or replant areas where stand counts are below a certain threshold. Getting an accurate stand count can be labor intensive.

A field is mapped with red and green squares representing plot areas.

“How do people do stand counts? Basically, they look around to see if there are any patches in the field,” said Thanos Gentimis, LSU AgCenter assistant professor and statistician. “They walk the fields, maybe for six hours.”

Instead of walking a field for hours, farmers can fly a drone overhead and get a more accurate — and literal — picture of the situation.

LSU graduate student Bhawana Acharya worked under the direction of Tri Setiyono, LSU AgCenter assistant professor in precision agriculture, and Gentimis on a project to improve the ability to get stand counts using precision and digital agriculture tools. Acharya trained the YOLOv8, a computer vision model, to detect the number of soybean plants emerging at certain vegetative stages with high accuracy.

“Just throwing out the hula hoop and counting the plants by hand in a large field can be very time consuming and labor intensive, so we needed an alternative,” Acharya said.

Her work is aimed at helping farmers create replanting or repairing maps in their field following damaging rainstorms. She uses images taken with handheld cameras and drones.

“In my project, I created some categories like full, low and empty plant density in the field and in the images, and based on the result, I made a map for replanting purposes.”

The color-coded maps will show areas that have no damage, some damage or heavy damage and allow farmers to make decisions on whether a replant or a repair is necessary to maintain desired yields.

After heavy rainfalls that damage plants, farmers need to make replanting decisions. With AI imaging analysis, maps can be created quickly and accurately. This saves farmers time and can save them money in the long run.

Acharya said the research will continue over multiple locations in Louisiana to fine-tune the intelligence. “Now you fly the drone for 30 minutes and you come back, and you know exactly where to plant,” Gentimis said.


Lowering the learning curve

Helping to create a better understanding of precision and digital agriculture is an aim of the AgCenter. In Louisiana, adoption rates of precision and digital agriculture tools and techniques are low. However, Gentimis sees a high desire among farmers, extension agents and students to use and learn more about these tools.

He estimates that around 45% of the farmers have smart equipment on their machines. “They just don't use it yet,” Gentimis said.

He said new farmers want (or are forced) to buy equipment such as tractors that are fully outfitted with digital ag tools.

Gentimis joined the AgCenter in 2018. He said then many farmers and extension agents weren’t talking about drones. Now they are putting together workshops to teach farmers about drone usage including the legal and ethical issues and eventually how to use data collected from them.

A recent cluster hire at the AgCenter, which included Congliang Zhou, is aimed at improving efforts to advance precision and digital agriculture in Louisiana.

Part of that effort is to beef up broadband in rural areas of the state. Programs are in the works to improve connectivity infrastructure which would allow producers to leverage precision and digital ag tools. Once connectivity is in place, the AgCenter would be the educational force behind farmers understanding these tools.

LaCADIAN, the Louisiana Climate and Digital Ag Network, is another program that would harness the powers of precision agriculture to benefit farmers. Vinit Sehgal, another recent hire who is an assistant professor in the School of Plant, Environmental and Soil Sciences, is spearheading this effort, which is funded through the efforts of U.S. Rep. Julia Letlow and the National Resources Conservation Service. LaCADIAN will be a statewide distributed network of automated field observatories and sensors to record critical zone water, energy and carbon fluxes for monitoring, modeling and management of Louisiana’s agriculture, energy, biodiversity, soil health and water resources.

“I'm expecting to see a lot more projects in this area,” Gentimis said.


Dulis precision ag.jpg thumbnail

Tri Setiyono, left, works with graduate students in a test plot of corn at Central Research Station. Students, from left, include Dulis Duron, Rejina Adhikari, Bhawana Acharya and Farner Rontani. Photo by Olivia McClure

Embracing the future

On Keith Dugas’ sugarcane farm in Napoleonville, drones fly over fields every few weeks. The drones are capturing images — images that provide data that can drive decisions on Dugas’ farm.

Dugas doesn’t shy away from change. He signed up to work with the LSU AgCenter to test out GreenSeeker crop sensor technology nearly 15 years ago to pinpoint fertilizer applications in his fields.

“If there is something out there that works better, I want to try it,” Dugas said. “Someone has to be the first one to try it.”

His work with GreenSeeker determined he could use less fertilizer on some areas of the field and see his sugar yields go up.

Now Dugas is collaborating with LSU College of Agriculture graduate student Dulis Duron and his adviser Setiyono, to use drone and satellite imagery to predict sugar yields on his farm. Accurate yield predictions can allow farmers to plan their harvest dates — they don’t want to start too early or too late — and help sugarcane mills approximate the amount of cane they will process that year to determine when their grinding season officially starts.

“Before it was kind of a shot in the dark and mainly using historical averages to predict yields,” Dugas said.

Duron’s research at Dugas’ farm involves creating computer models that can take the images and look at sugarcane stalk width, how the crop covers the soil, the color of the leaves and other data points.

In addition to the satellite data, Duron also incorporated weather data such as precipitation and cumulative solar radiation through growing degree days which measures heat accumulation in the crop.

These are run through advanced AI-based modeling programs run on Python and R — two programming languages — which can identify nonlinear patterns and increase the prediction accuracy.

Researchers have been training the model for three years and have seen accuracy improve over time.

“We can make final yield predictions, and the results are really good,” Duron said. “There is enough data to apply the model to the 2024 harvest.”

Duron has an undergraduate degree in agriculture but didn’t know much about precision and digital agriculture before coming to LSU. He is part of a cohort of agriculture students on the cutting edge of these new technologies.

He said he is able to apply tools such as R, Python and artificial intelligence that he learned in the School of Plant, Environmental and Soil Sciences and in Gentimis’ digital ag class to the work he is doing with Dugas.

Duron is just beginning his journey into precision ag, but Dugas has seen its benefits for the past decade and said he continues to be amazed by the advancements in agriculture.

“I never thought drones could be as useful as they are now — flying chemical and ripeners over the field. These images get better and better, and the predictions become more reliable,” Dugas said. “I don’t know where it is going to be next, but it is amazing.”

Two drones fly in a field.

Drones fly over a sugarcane field at the LSU AgCenter Sugarcane Research Station. Photos by Olivia McClure

The U.S. Department of Agriculture Economic Research Service conducted a study on the adoption of precision agricultural tools on U.S. farms. The study found:

  • A majority of row crop acreage is managed using auto-steer and guidance systems: Auto-steer guidance systems were used on only 5.3 % of planted corn acres in 2001, growing to 58% in 2016.
  • Adoption rates vary by farm size: At least half of relatively large row crop farms rely on yield maps, soil maps, variable rate technology and/or guidance systems. Meanwhile (except for cotton), less than 25% of smaller farms use any of these four technologies.
  • Digital ag (DA) technology adopters use data, acquire crop management recommendations and employ technical/consultant services at higher rates than DA technology nonadopters: DA technology adopters are more likely than nonadopters to download public data for use in decision-making, though overall adoption remains uncommon.
  • Farmers are likely to use precision agriculture technologies for a variety of reasons: As technological capabilities continue to evolve, so have farmers’ rationales for their use. For example, corn and winter wheat farmers tend to rely on yield monitors to track crop moisture content. By contrast, yield monitors are primarily used to help determine chemical input use in cotton, soybean and sorghum production. Many precision agriculture technologies are used in combination with other precision agriculture technologies.

11/21/2024 6:09:12 PM
Rate This Article:

Have a question or comment about the information on this page?

Innovate . Educate . Improve Lives

The LSU AgCenter and the LSU College of Agriculture

Top