Future Farming in India A Playbook for Scaling Artificial Intelligence in Agriculture 2025

Appendix 5: An assessment of critical datasets for AI in agriculture The report team evaluated datasets for availability, useability and quality, taking on board insights from the United Nations Food and Agriculture Organization, India Meteorological Department (IMD), Copernicus and Agri-Stack. This process enabled the team to identify the 15 most critical datasets and the key gaps in granularity, real- time availability and interoperability. Closing these gaps will strengthen India’s agricultural data ecosystem, ensuring AI adoption that is aligned with global agriculture’s best practices. This assessment was guided by an earlier McKinsey– National Association of Software and Services Companies (NASSCOM) report, and targets AI start-ups and government leaders who can unlock AI’s potential in agriculture. Fifteen critical datasets for AI in agriculture Names of dataset Dataset descriptions Digital land records registry* Digital land records* registry that establishes titling of collateral and has legal validity with various departments (revenue, survey, etc.) Crop calendar and yields*** Crop-cutting experiment data, data on actual yields for crop varieties per area harvest, crop weather calendars of major crops, cropping area under the Regulated Farming Initiative* (Agri-Stack*) Soil health** Agronomic details such as soil type and fertility, including nutrient availability (macro, micro, secondary) and moisture content, for each farm, fertilizer subsidy data Satellite imagery** High-resolution images to identify farm boundaries, crop distribution, yield, etc. Real-time mandi data*** Includes real-time statistics on market prices and arrivals for different crop varieties from commodity trades, along with historic price data across various markets Agriculture market network*** Agriculture market network by location, crop type Import, export volume details*** Import and export volumes for crop varieties by month and location Historical purchase prices for crops*** Historical daily purchase prices for crops by location, market type and level (e.g. farmer, intermediary, etc.) Production and consumption data***Production and consumption volumes for crop varieties by month and locationTABLE 8Reliable datasets are essential to develop AI-based tools for yield forecasting, pest control and supply-chain optimization. Future Farming in India 50