We identified remote sensing, AI and ML and GIS as the technological building blocks of the solution.

Multi-temporal satellite data was used to study the crop-growth cycle. Technologies such as Machine Learning and Artificial Intelligence were the backbone of our solutions such as Sowing Intelligence, Crop Identification, Crop Health Monitoring, Crop Loss and Damage Assessment, Irrigation Planning, Crop Acreage Estimation, Crop Yield and Production Estimation and Forecasting, and Precision Farming. The field survey posed its own unique set of challenges for collecting data over 100,00,000 Ha.

On the demographic front, the solution brought better insights to farmers and to the Government for implementing future policies. With minor modifications, the solution is capable enough to accommodate the needs of the agro industries, agri-products industry and seed industries in the near future.

When the solution first commenced, initial figures of 70% data accuracy soon grew to over 80% accuracy. This could now be achieved in a lesser time interval, as the machine learning component had matured over subsequent interactions and historical data availability. The process and domain understanding are continuously developing to offer clearer and more accurate results each time.

Farmers, insurance companies and government agencies were finally on the same page. Farmers were protected from suffering huge financial losses due to risks posed by weather, endemic disease or other uncontrollable factors. Rural societies and the nation as a whole realised the potential that technology in agriculture could serve. Furthermore, the money spent by government agencies for crop cutting experiment to gather crop data was drastically reduced. This project also encouraged other states to implement this solution for the benefit of local farmers.