Wageningen University, IFA, and Agmatix collaborate to analyze crop nutrients big data
SHARE
- 8 minutes read
Crop nutrient optimization across multiple production environments requires large amounts of high-quality data. In an innovative research project, the Consortium for Crop Precision Nutrition (CPCN), Wageningen University (WUR), and the International Fertilizer Association (IFA) collected data from multiple researchers and institutes to build a first-of-its-kind open database of nutrients in crops under an array of environmental conditions. The database will improve our understanding of ongoing trends in crop nutrient uptake and removal. It will also make it easier to create decision support systems to determine how to optimize crop production sustainably under changing environmental conditions.
To date, over 5,000 research trials have been collected from more than 50 researchers who have shared data from 70 countries about four nutritionally and industrially important crops–maize, rice, soybeans, and wheat. Unfortunately, their data files differed considerably in terms of structure and parameter nomenclature. Thus, it was extremely challenging to analyze them in a meaningful way to obtain insights—if at all.
Challenges
Data Fragmentation
The primary challenge was the fragmentation of data, making it difficult to create a cohesive and comprehensive dataset for analysis.
Lack of Standardization
The absence of common naming conventions and protocols significantly hampered our ability to generate a unified dataset, crucial for in-depth analysis and comparison.
Analytical Complexity
Ensuring the accuracy and reliability of big data analytics, given the diverse data sets from various sources, presented a significant challenge in predicting nutrient needs accurately.
Data Integration
Integrating diverse datasets from multiple researchers and institutes into a unified database for analysis was a complex task due to the variability in data structure and parameter nomenclature.
Our Solution
By harnessing Axiom, Agmatix’s state-of-the-art technology and innovative data standardization approaches, they were able to standardize the data and create an open database. The collaboration between Agmatix, CPCN, IFA, WUR, and many others aims to revolutionize our understanding and application of nitrogen in corn production, balancing high yields with environmental sustainability.
Unified Data Standardization
Leveraging Agmatix’s GUARDS protocol to standardize and harmonize diverse agronomic data, enabling insightful analysis across global datasets.
AI-Driven Insights
Utilizing AI technology to transform big data into actionable models, predicting nitrogen requirements with precision for sustainable crop management.
Comprehensive Database Creation
Assembling a first-of-its-kind nutrient database, compiled from data across 70 countries, to support global research and decision-making.
Predictive Modeling for Nitrogen Optimization
Developing an advanced machine learning model to accurately predict grain nitrogen concentration, with a mean absolute deviation of 0.09% and an average prediction error of 7.2%.
Key Factor Analysis
Identifying crucial determinants of nitrogen availability, including cultivar maturity, nitrogen input, and soil organic matter, to guide optimized fertilizer application.
Collaborative Research Platform
Encouraging researchers and organizations worldwide to contribute to and benefit from this collective knowledge base, enhancing agricultural productivity and sustainability.
Results
Unified Agronomic Database
Initially data from 5377 observations collected from three countries – the United States, China, and Nigeria – were standardized and harmonized using the Agmatix platform. Since then the database has expanded with an additional 2000 datasets and are feely available at: cropnutrientdata.net
Data Augmentation
Multiple covariates such as nutrient input, soil texture, organic matter in the soil, and cultivar maturity were augmented with site-specific weekly rainfall and temperature data.
Predictive Modeling for Grain Nitrogen Concentration
Developed a machine learning model generating an ensemble of decision trees to predict grain nitrogen concentration, where 25% of the data were reserved for validation purposes.
Model Performance
The model successfully predicted the percentage of nitrogen in grain from these three countries with a mean absolute deviation (MAD) of 0.09 [% Nitrogen] and an average prediction error of 7.2%.
Our collaboration with Agmatix and the adoption of their Trial Management platform has revolutionized our product trials processes. We’ve cut months off our data processing time, significantly accelerating product development and market readiness. We are excited about the prospect of bringing new products to market faster, enabling higher yields for our customers.
Dr Prashant Puri,
Head of R&D, Deepak
Conclusions
With its innovative data analytics platform, Agmatix has been at the forefront of transforming how agricultural data is interpreted and utilized. By utilizing expertise and technology, Agmatix has not only been able to manage and analyze the vast datasets used in this project but also ensure that the insight derived is actionable and scientifically backed.
Increased Precision in Nutrient Management
Big data analytics can significantly improve the precision of crop nutrient management.
Collaborative Strength
Collaborative efforts are crucial for overcoming technical and operational challenges in agricultural data analysis.
User-Centric Technology
User-centric design is essential for the adoption of technological solutions in agriculture.
Robust Data Privacy Frameworks
Data privacy and security are manageable with clear guidelines and stakeholder engagement.
Looking Ahead
Cultivating a More Efficient, Sustainable Agricultural Landscape
The collaboration continues to expand its reach, incorporating more data sources and refining its models for broader applicability. The overarching goal of the open database is to catalyze broader adoption of these tools and practices, potentially revolutionizing nutrient management practices on a global scale. To learn more about the Crop Nutrient Data project, please visit: cropnutrientdata.net
