UC Davis | CottonFormatics Big Data Solution
Research, Intellectual Property Development, Commercial Venture Support
- Productionalized, stable platform to support research and isdustry uses
- Enabled statistical analysis indicating reduced use of pesticides for common pests (Lygus), resulting in immediate cost savings for industry
OUR CLIENT’S CHALLENGE
Although big data’s impact is most widely-known in online advertising and recommendation engines (Amazon, NetFlix, Facebook), there are many applications of big data concepts in traditional industries such as agriculture.
Cotton is one of the oldest commercialized plants, yet researchers and agricultural consultants continue to collect data on all aspects of cotton growth and productivity. Data collection tends to be limited to single installations, allowing researchers to use traditional data management tools such as Microsoft Excel.
A recent meta-study conducted by the University of California at Davis on behalf of Cotton Incorporated and the US Department of Agriculture required a much larger statistical population. The university enlisted the assistance of four agricultural consulting companies in aggregating ten years of insect, crop, plant health, cotton yield, and pesticide use data from multiple sites. The data format and quality varied widely across sites, and there was so much data that manual review and correction were not feasible.
To accomplish the goals of the study and anticipated future initiatives, UC Davis engaged Ten2Eleven to produce a big data platform to ingest, cleanse, validate and summarize the cotton study data, as well as provide defined datasets and reports to researchers in an on-demand mode.
Phase 1 – Needs Analysis: The Ten2Eleven development team worked closely with university staff. Through several requirements gathering sessions, the Ten2Eleven team identified a standard data model and data requirements driven by the goals of the study.
Phase 2 – Solution Development: Ten2Eleven’s data engineers created the Cottonformatics application, using relational database products designed to scale up to Terabytes of data. The application included data acquisition, validation, normalization, aggregation, and reporting components. The application included a secure web-based portal where authorized research staff could request and download data sets and could run ad hoc reports. A modular approach was used to facilitate scaling up or revising individual modules without impacting overall application performance or functionality.
Phase 3 – Product Implementation: Once the application was built and thoroughly tested, the Ten2Eleven implementation team partnered with UC Davis Systems Engineering staff to install Cottonformatics in a production environment. Ten2Eleven staff provided Tier 2 support for the initial period of use, to ensure that issues were addressed promptly and that the new analytics platform was a success.