Advancing agricultural research using artificial intelligence algorithms
2021
- Ag Stat, Athens Greece
- Department of Plant Pathology, UW Madison
- Department of Agronomy and Horticulture, Univ. of Nebraska
- Department of Agronomy, UW Madison
Project Media
Field trials are commonly used to estimate the effect of different factors on crop yield. To
date, evaluating the effectiveness of management practices to increase yield has been restricted
to specific soil types and weather conditions (i.e., environments) and background management
cropping systems. Thus, results of such experiments cannot be safely generalized to farms with
diverse soil types and background management. Currently, a method that evaluates and predicts
the effectiveness of tens of thousands of possible cropping system interactions to increase yield
in each specific field across the US does not exist. We have developed a novel approach to
perform such evaluation by aggregating data from thousands of experiments across the US by
leveraging the power of artificial intelligence algorithms. Our approach and algorithms can help
accelerate agricultural research by generating accurate yield estimates for thousands of cropping
systems and environments for specific fields. The result of this work can allow individual
farmers to identify the most appropriate cropping system (i.e., practice adoption) for their
specific environment and ultimately increase yield and/or profitability.