Cloud-Based Precision Agriculture Platform

Cloud-Based Precision Agriculture Platform Service aided by satellite data provides intelligent analytical decision-support tools to easily monitor, understand and optimize field yield and crop performance.

Modern agriculture has a growing demand for increased land productivity while reducing production costs. Farmers need a reliable yet cost-effective way of fetching data to improve their yields. The client reached Quantum to address this challenge.

Our client wanted to help individual farmers and agriculture organizations interested in measuring the performance of agricultural growth at the field level. For that purpose, a company needed to use satellite and drone data to get vegetation indices to analyze field status.

We developed a platform to solve everyday problems in agriculture by analyzing satellite imagery. It gets satellite data suitable for various crop analytics, delivering wide coverage of agricultural areas.

For instance, it processes anomaly situations like nebulosity and overcast. Depending on the location, information on crops and field performance is updated every 3-5 days, allowing farmers to analyze different indexes to increase field productivity.

As a result, the solution successfully helped individual farmers and organizations to address common agrarian challenges and evaluate yield performance using satellite and drone data. This solution also helped our client handle the barriers to entering a more intelligent business market and assisted business models in scaling up.

Our solution consists of two main components:

• Service hosted at Amazon EC2 that automatically finds the newest datasets captured by the Landsat Sentinel satellite, cuts out areas of interest, calculates the vegetative index (NDVI), and creates a raster from this data.
• A web application to display the collected statistics. Our client gets all the needed information on top of Google Maps, with an analytics platform and geospatial measurement tools. Additionally, we’ve connected the service to statistical data from Google Earth Engine to obtain even more statistics.

The back-end part was developed using Python. As the database, we have chosen PostgreSQL. For data computing, we used AWS Lambda and deployed our service with AWS. The system uses GDAL, a translator library for raster and vector geospatial data formats, and is deployed on the AWS infrastructure. Other technologies we worked with included MBTiles, JavaScript, and PyQt.