We are pulling together climate related resources.

Spatial Agricultural Crop Yield Prediction

Stage: Prototype

Opportunity: Idaho National Laboratory (INL), managed and operated by Battelle Energy Alliance, LLC (BEA), is offering the opportunity to enter into a license and/or collaborative research agreement to commercialize the Spatial Agricultural Crop Yield Prediction technology. 

Background: Access to spatiotemporal subfield yield data will play an important role in the adoption of emerging precision agriculture technologies. Although access to such yield data is increasing, many stakeholders lack the ability to analyze and act upon this data at industrially relevant scales. Many growers either lack the tools or resources need to aggregate this type of data, which currently hinders the wide-scale deployment of precision agriculture practices.  

Description: Researchers at INL have developed a new modeling technology to predict agricultural crop yields (corn, grain, wheat, and soybeans) prior to harvest at subfield resolution. This technology uses artificial neural networks (ANN) to analyze electromagnetic (EM) energy reflectance measured by satellite sensors to monitor crop phenology. The models are trained with real world spatial harvester data provided by individual farmers, agricultural retailers and other stakeholders.

Applications and Industries

Applications: Varying agricultural stakeholders would have interest in this capability, including equipment manufacturers or multi-tiered agribusiness entities. Although many harvesters are equipped to capture spatiotemporal yield data, many do not because of challenges due to data management/storage and calibration requirements necessary to ensure values are correct.


Advantages: Although the agricultural space is crowded with varying technology providers in the remote sensing field, many are focused on expensive and difficult-to-obtain data. This capability uses globally available remote sensing data captured with scientific-grade sensors. The structure of industrial partnerships would also ensure a continual flow of high-quality training data to further improve model predictions and maintain pace with ever-changing crop varieties/hybrids.

Impact: This capability would allow stakeholders to access accurate spatiotemporal yield predictions at a subfield resolution. This data could be used as a basis for variable-rate application maps to improve nutrient application, water quality, and stakeholder economics. There is also a potential to re-project yield predictions to improve in-season crop-management decisions.


BA-1115 Crop...n.pdf

Jan 6, 2020