Agriculture in India is the backbone of millions of livelihoods, yet farmers and policymakers often face a simple but fundamental problem: we don’t always know what is being grown, where, and in what condition. Traditional methods like manual surveys are slow, expensive, and frequently inaccurate. Satellite solutions exist, but in a country with monsoon clouds, fragmented fields, and diverse cropping patterns, they often fall short. This lack of visibility leads to challenges everywhere — farmers miss timely advice, governments struggle with subsidy and insurance planning, and markets fail to anticipate supply and demand.

Our solution bridges this gap by combining the latest in satellite AI with on-ground reality. Using geospatial foundation models trained on Indian datasets (AgriFieldNet + local surveys), we can classify crops with high accuracy at the field level. Unlike regular satellites that get blocked by clouds, we integrate multispectral imagery, Synthetic Aperture Radar (SAR), and crop growth timelines to ensure reliable insights throughout the year, even in the toughest weather.

But data alone is not enough. What makes our approach different is how we deliver these insights: directly to farmers through SMS and WhatsApp in their own languages. A farmer doesn’t see “raw maps”; instead, they get clear, actionable messages like “Your paddy field is water-stressed — irrigate within 2 days.” At the same time, aggregated insights are shared with policymakers, helping them plan subsidies, crop insurance, disaster response, and food supply forecasting with confidence.

Designed to scale across states and crop types, this system brings precision agriculture to the grassroots. By turning cutting-edge satellite models into simple, farmer-friendly advice, we not only boost yields and incomes but also contribute to food security, poverty reduction, and climate resilience. In short, we’re making sure that every farmer — no matter how small their plot — has access to the power of space technology.

Built With

  • flask
  • gdal
  • google-earth-engine
  • postgis
  • rasterio
  • react
  • sentinelhub
Share this project:

Updates