Field Intelligence for Regenerative Agriculture in Predictive Sustainable Groundnut Farming in India

Groundnut, a key oilseed crop in India, faces stagnant yields due to environmental and agronomic challenges. Sustaining agriculture amid climate stress and soil degradation remains a major issue in dryland regions. Hence proposed a novel field intelligence framework to enhance groundnut yield forecasting and agroecological planning using regenerative principles. A Regenerative Agriculture Index (RAI) was developed by combining crop diversity, soil health proxy, and yield-based sustainability metrics. Genetic Algorithms (GA) were used for optimal feature selection, and XGBoost was employed for high-accuracy yield prediction. SHapley Additive Explanations(SHAP) was used to explain the model’s decision logic. The k-means clustering was applied to categorize regions into high, medium, and low sustainability zones. The model achieved an RMSE of 0.204 and R² of 0.89, outperforming traditional baselines. Results highlight the strong influence of RAI and previous yield on performance. This integrated approach supports informed decision-making for sustainable agriculture.