Farm406

farm406 48 e line between farming and science is beginning to blur, but the outcome is clear: e more data, the more precise and lucrative agriculture can be. At Montana State University, the College of Agriculture has been diving deep into research for years to improve the business of farming and ranching. Bruce Maxwell, a Department of Land Resources and Environmental Sciences professor, collaborates with John Sheppard, professor in the Gianforte School of Computing, and their graduate students to study the use of precision agriculture for field management. Precision agriculture is a detailed approach to farm management using modern technology. It lets farmers increase the efficiency of their output — products like wheat or meat — by decreasing or maintaining the same input, such as fertilizer, seed, water, and herbicides, while increasing the output. Maxwell works directly with farmers to implement technological solutions to streamline the inputs for wheat. He says his team primarily focuses on nitrogen fertilizer, and no two fields are alike. Instead of using a fixed amount of fertilizer spread evenly over a whole field, Maxwell and his team collects data points using GPS technology aached to farm equipment — to map out fields and apply precise amounts of fertilizer where needed. To verify the precision of fertilizer rates, farmers use devices during harvest to capture thousands of crop yield data points — yield is measured every three seconds and Sasha Loewen, an MSU doctoral student in ecology and environmental sciences, points out maps captured with satellite data of a wheat farm to help predict yields as part of his ongoing research with precision agriculture. wheat protein every 10 seconds — providing field-specific information vital to the crop's market value. e devices instantly generate informational maps that allow farmers to adjust fertilizer rates for future harvests. e agronomic work crosses over into the computer science arena, allowing Maxwell and Sheppard to collaborate and develop artificial intelligence-based tools to analyze the data. "With recent technological advances in remote sensing and precision variable application of things like nitrogen fertilizer, seeds, and herbicides, it has become critically important to develop and apply the latest in intelligent computation and data analysis techniques to help farmers make the best decisions based on their valuable data," Sheppard said. Sheppard and the computer science graduate students focus on developing soware systems that employ artificial intelligence, enabling farmers and crop consultants to optimize management decisions. ose decisions include determining how to design on-farm experiments to collect data about the performance of the field under different input conditions, using machine learning to understand how to predict crop response from input conditions, and optimizing the trade-offs between objectives such as net return, equipment wear and tear, and environmental impact. "All of this work is being done with an eye toward determining economically optimal decisions for the farmers," Sheppard said. "Ultimately, the goal of this research is to advance farm management decision making, as well as state

• Cover