Discover how Colorado State University’s Agricultural Research, Development and Education Center is using technology to innovate the future of agriculture. Video by John Cline/Office of the Vice President for Research
NSF-backed research seeks to provide new, precise irrigation tools to producers
video by John Cline
story by Josh Rhoten
published Sept. 28, 2023
An interdisciplinary team of researchers at Colorado State University is working with farmers to proactively schedule and better monitor irrigation use in their fields using sensors, satellite imagery and deep learning technology.
The work is led by Professor Sangmi Pallickara in the Department of Computer Science and was funded by a $1.2 million award from the National Science Foundation in August. The project continues research into improving the center-pivot irrigation systems that are common across the western U.S. by the team, which includes faculty from the College of Natural Sciences, the Walter Scott, Jr. College of Engineering and the College of Agricultural Sciences.
Nationally, irrigation accounted for 42 percent of total freshwater withdrawals in 2015, according to the United States Geological Survey. That same report showed that Colorado was among the top states by percentage of total withdrawals for irrigation as well – making this a key issue for producers locally and nationally going forward.
In center-pivot irrigation systems, an arm carrying water moves in a circle to water crops evenly. Pallickara said the CSU team is working to improve that now decades-old process by utilizing sensors in the field and satellite imagery to inform models that seek to ensure precise water use across large fields and down to different root zone depths.
Eventually this work could lead to a dashboard agricultural producers could use to make better informed decisions and fine-tune use in their fields.
“The goal is to forecast crop water needs using novel deep learning algorithms and game theoretic methods,” Pallickara said. “Better water use and placement would prevent erosion and nutrient runoff from overwatering while also improving yields. The growers also still retain control in this system to make decisions and consider how variables such as water rights and usage rates may come into play in a given year.”
Deep learning is part of a set of machine learning algorithms that are now frequently applied to fields such as medicine and climate science for improved prediction, discovery and optimization activities.
These models are based on the biological networks seen in human brains, and they can “learn” to better perform tasks like irrigating crops after rapidly reviewing and evaluating available data such as soil types and growing phases from examples to make better decisions overall.
Next generation drone training
The next generation of agricultural producers in Colorado will need to be fluent in a variety of latest technologies and tools that will aid in their daily work. That includes using dashboards to access data on the condition of their crops, based on sensors buried in their fields, or piloting unmanned aerial drones to find lost animals, for example. While researchers at CSU are working to make these systems as easy and accessible as possible, parallel workforce development training will be a key in building awareness, adoption and use of those technologies over the next decade.
With that need in mind, the CSU Drone Center recently launched a training program for high school students and teachers that seeks to expose them to unmanned aerial vehicle technology for use in agriculture and beyond. The effort is funded through a workforce development grant from the Federal Aviation Administration and is done in partnership with the AIMS Community College Flight Training Center in Northern Colorado.
Through the program, students get hands-on flight training to prepare for remote pilot certification, while teachers are given curricular materials and their own drone for future training back in their local classrooms. Participants also met with professional pilots from a variety of backgrounds to get career advice and insight. To date, 87 students and teachers from Rangely, Grand Junction, Mancos and beyond have participated in the program, and many are now fully credentialed remote pilots, said Adam Smith, the assistant director of the CSU Drone Center.
“Our team is grateful to be able to provide these programs to Colorado high school students, and we are very passionate about training the next generation of aviators and instructors, especially when we can reach such a diverse group from so many different backgrounds and areas of the state,” he said.
The training program is open to all students – no matter their background or interest in using drones – and is completed with ongoing support from the center. Interested groups can learn more on the center’s website.
Data for this project comes from growers as well as information collected by Simplot – a precision agricultural company that is working with the team. Other information comes from testing fields at the University of Nebraska-Lincoln.
“Part of the challenge is that the data we get back on precipitation or root health can be ‘noisy,’ meaning it is hard to understand what is really going on in the field. However, these sensors are cheap to use, and we will be doing field work through the Agricultural Research, Development and Education Center on campus in 2024 to help address that issue,” she said. This will also help the team calibrate methods to determine where to place sensors for the best readings.
A key part of the research is ensuring this information is available and accessible to producers without the need for extensive training. The team is partnering with Simplot and Extension experts on campus to address that goal, Pallickara explained.
“We have started developing tutorials and are planning to do outreach activity in the near future as well,” she said. “Working with a company like Simplot has really helped to address those concerns early on, and I would say the farming community has been open to using and learning about the technology, which is great.”
Researchers on the project include Sangmi Pallickara, Shrideep Pallickara, Jay Breidt, Jeffrey Niemann and Allan Andales.
Read more of The Future of Ag is Now
This special report from SOURCE explores the breadth of multidisciplinary, agricultural work happening at CSU — a place where researchers, students and food producers are all collaborating to shape the future of agriculture.