AgriSenseAI

Inspiration

Farmers will often realize crop diseases only when it is too late, causing yield and income losses. Field inspections will remain costly and limited. We will build a scalable, AI-powered solution that will use farmer IDs and free satellite data to provide early warnings and timely advice directly to farmers

What it does

AgriSenseAI will auto-fetch a farmer’s land boundary from government databases using their registration number. It will analyze satellite imagery (Sentinel-2, Landsat, NRSC Bhuvan), apply vegetation indices and AI models to detect crop stress or disease, and generate localized advisories in the farmer’s own language. Alerts will be delivered via app, SMS, or WhatsApp.

How we built it

Land Data Integration The system will integrate government land record APIs to fetch geospatial farm boundaries.
Satellite Data Collection It will pull free satellite data (Sentinel-2, Landsat) covering those parcels.
Preprocessing The data will be preprocessed using cloud masking, NDVI/NDWI indices to generate clean inputs.
AI Detection Pipeline An AI pipeline (Granite LLM + custom models) will identify crop stress and diseases.
Advisory Generation The system will deliver natural language reports, linking farmers to agro-shops and subsidies.

Challenges we ran into

• Accessing accurate and up-to-date farmer land record APIs. • Handling large satellite image files and preprocessing efficiently. • Differentiating between disease stress vs. water/nutrient stress in crops. • Ensuring reports will be clear, localized, and easily understood by farmers.

Accomplishments that we're proud of

We will create an end-to-end AI pipeline from farm ID → satellite → disease detection → advisory. We will enable multilingual outputs, making the solution inclusive.

What we learned

• How to integrate AI vision models with geospatial data. • The importance of farmer-friendly design (language, simplicity, SMS access). • The challenges of data preprocessing in remote sensing. • Collaboration and workflow automation using IBM’s Agent Development Kit.

What's next for AgriSenseAI

• We will add historical crop health tracking to show season-wise trends. • We will integrate with govt subsidy portals for financial recommendations.