Paulo Flores, North Dakota State University assistant professor of agricultural and biosystems engineering, has been in his current role at NDSU since 2019. Previously, he worked for the NDSU Carrington Research Extension Center in two roles, including as a precision agricultural specialist from 2016-19.
Technology has always impacted agriculture, but new methods and equipment are arriving rapidly in today's world. Examples are the use of uncrewed aerial systems, sensors and geographic information systems, which are being utilized by researchers, breeders and growers alike.
Paulo Flores, North Dakota State University assistant professor of agricultural and biosystems engineering, sees the direction this technology is going and the potential for its impact.
“If I had to summarize in one word, it probably would be autonomy,” Flores said. “I think we will see the development and deployment of more autonomous solutions to carry out tasks related to different phases of crop production, including tillage, planting, crop protection and harvesting. In addition, those platforms will be able to perform their tasks on a much smaller scale than we currently do. It might take some time, but I think we will get there.”
Technology and research in UAS, sensors and geographic information systems are all parts of the research being done as part of the Food, Energy and Water Security initiative at NDSU. Flores and his team are applying those in their research program. The FEWS initiative addresses many key issues facing food, energy and water resources today, strongly emphasizing technology. Autonomous systems, sensors, high throughput data systems, agriculture data security, precision agriculture and predictive crop performance are among the areas led by NDSU in that research.
“The work that Dr. Flores, his team and collaborators across NDSU Agriculture are doing in automation is driving real-world impact, advancing technologies that support economic development, improve environmental outcomes and deliver prosperity to our farmers,” said North Dakota Agricultural Experiment Station associate director Frank Casey. “With federal support through the USDA-funded FEWS initiative, we’ve generated tremendous returns. For example, drone imagery research has created new efficiencies for our plant breeders, streamlining field data collection and accelerating the development of high-yielding, disease-resistant varieties. This same research has also led to innovative approaches for identifying and managing weed pressures more effectively, even using existing equipment.”
Flores has been in his current role at NDSU since 2019. Previously, he worked for the NDSU Carrington Research Extension Center in two roles, including as a precision agricultural specialist from 2016-19. He was introduced to and learned GIS 15 years ago as a postdoctoral associate at Virginia Tech University. His GIS knowledge was key for getting involved with UAS research as that technology started to be used at the CREC. He continues to apply his GIS knowledge in his research program involving precision agriculture and high throughput phenotyping at NDSU.
“GIS principles are at the core of everything that we do related to field research since we are always working with geospatial data,” Flores said. “From an HTP point of view, we need to make sure that the data pulled from a plot, based on UAS imagery, matches the plot location in the field from which researchers have collected their ground data so those can be correctly matched. Similarly, for site-specific weed control, one needs to ensure that the weed location provided by our maps accurately matches the location of the weed on the ground and the accuracy of the GPS location on the tractor or sprayer.”
Flores got his first glance at precision agriculture while he was a postdoctoral associate at Virginia Tech. He worked on a grazing project that tracked cows and calves with GPS collars and gathered information on cattle behaviors as moms and calves were separated from each other.
Flores joined the Carrington Research Extension Center in 2014. He was part of a group led by an NDSU agricultural machine specialist that started the first UAS flights using that technology for precision agriculture applications.
Flores commented: “Since then, my research program has focused on using UAS for different applications related to breeding programs and agronomy.”
Flores has researched two primary areas in the FEWS realm. The first is the applications of UAS and sensors to implement HTP applications that improve efficiencies for plant breeding programs at NDSU.
“My team has been working with several researchers from the plant sciences department to develop tools to automate the processes to analyze UAS imagery to extract information regarding traits of interest,” Flores said. “We have created various tools in my lab that decrease the time to perform some tasks from hours to a few minutes or even seconds. My team has been collaborating with the NDSU Agricultural Data Analytics team to bring to life a web-based application (SkySight) incorporating many of the tools developed in my lab for UAS imagery analysis. Such a platform will allow many other researchers to benefit from the tools that we have developed, with an easy-to-use interface developed by our collaborators.”
Jithin Mathew, a graduate research assistant for Flores, has been involved in image-based phenotyping for major cash crop breeding programs. He’s combining machine learning with advanced image processing algorithms to improve the efficiency and accuracy of plant trait analysis.
“My research focuses on applying image processing and computational techniques, particularly machine learning, to address key challenges in agricultural problems,” Mathew said. “Through rigorous real-world testing — both in field conditions and greenhouse environments — our team continuously works to refine and deploy innovative solutions that benefit the broader agricultural research community.”
A second research focus is using UAS imagery to map weeds across fields for various crops and integrate that with site-specific weed control – in a nutshell, to spray areas only where there are weeds in the field. Flores said that sensors mounted on UAS could support site-specific weed control.
“I think that the main impact that my research program will have for growers is the use of UAS to map weeds across fields,” Flores said. “Having a weed map from a field, or fields, can aid in several aspects of weed management and/or control, which can apply to organic and conventional growers. An organic farmer might use those maps to prioritize what fields to tackle first, and at a field level, it can be used to direct crews or equipment to make operations more efficient. In case those farmers are using some autonomous solution, like robots, for weed control, such maps can be used to optimize the deployment location of those assets and/or be used to increase the speed at which those platforms move across the field, such as moving faster across areas free of weeds.
“As for conventional growers, our work related to SSWC has the potential impact of decreasing the amount of chemicals being applied to fields. That impact can be even higher if multiple chemical applications are required through the growing season. The application of fewer chemicals in the field can lead to fewer chemicals in the runoff water that usually end up in water bodies,” Flores added.
The research’s focus fits well with the FEWS initiative.
“As for our work related to high-throughput phenotyping, I can see possible impacts on FEWS in two ways. The first is helping plant breeders shorten the time needed to release better genotypes -- higher yielding, more resistant to diseases or abiotic stresses,” Flores said. “The second one is by helping them to identify more candidate genotypes from larger populations with traits of interest that could be developed into new varieties or hybrids simultaneously, increasing the output of their breeding programs.”
The NDSU FEWS initiative receives funding from the United States Department of Agriculture-ARS.
