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

TABLE 1 Indicative roadmap for operationalizing AI-enabled crop planning Outputs at end of step Role of government Other critical stakeholders 1 Develop strategy and aggregate data –Strategic plan –Documented data needs –Aggregated foundational datasets –Expert group –Pilot scope –Lead strategy –Formulate expert group –Aggregate datasets –Finalize scope of pilots –Expert group: advise on best practices –Agricultural research institutions: share existing datasets –Agricorporates: provide data on historic and demand/ procurement prices 2 Develop AI crop- planning model –Onboarded agency for developing AI model –Predictive models on crop recommendations based on viability/feasibility –Onboard AI model developer –Set data privacy standards –Initiate sandbox for validation and governance –AI model-developer: develop models –Agricultural research institutions: support testing of crop recommendation models through sandboxes and real-time data 3 Generate regional crop plans –Actionable crop recommendations at a regional level based on AI model –Review regional crop plans through a federated structure, including local and national experts –Agricultural research institutions: generate package of practices for recommended crops 4 Deliver recommended crop plan –Package of practices for recommended crops –Delivery plan for package of practices to farmers –Train extension staff to deliver services –Design financial incentives for adoption –Agritechs, extension agents and FPOs: disseminate recommendations and package of practices –Agricorporates: ensure availability of inputs 5 Increase adoption and collect feedback –Adoption by farmers –Feedback mechanism for continuous improvement of model –Deliver extension through channels such as SMS, radio, extension staff and government institutes –AI model developers: continuously improve model –Agritechs, extension agents and FPOs: support adoption within their network of farmers The context Soil health testing in India has become critical because of soil degradation17 and the associated decline in yield. But traditionally, soil testing in India is time-consuming, requiring physical sampling and laboratory work. Additionally, India has only about 8,000 soil-testing labs to serve a farming population of approximately 150 million. It can be relatively expensive, too, so farmers often rely on As of 2021, 97.85 million hectares of land in India has been degraded,15 a considerable proportion of which is agricultural land. The United Nations Food and Agriculture Organization warns that by 2050, 90% of the Earth’s topsoil is likely to be at risk.162.1.2 AI-enabled rapid soil-health analysis Future Farming in India 18