Intellectual Property India Publishes Patent Application for System And Method For Automated Crop Prediction Using Machine-Learning-Based Soil Nutrient Estimation

MUMBAI, India, Sept. 22 -- Intellectual Property India has published a patent application (202641110333 A) filed by Dr. Anand M.; Mrs. Pooja N. S.; Dr. Pradeep Kumar R.; Ms. Pushpalatha S. S.; Mrs. Suhasini; Mrs. Divya G. S.; Mrs. Shwetha K. R.; Mrs. Vijayalakshmi M. M.; Mrs. Mahalakshmi B.; Dr. Thilagavathy R.; Dr. Kanike Vinod Kumar; and Mrs. Apoorvashree H. L. on September 14, 2026, for System And Method For Automated Crop Prediction Using Machine-Learning-Based Soil Nutrient Estimation.

Inventors include Dr. Anand M.; Mrs. Pooja N. S.; Dr. Pradeep Kumar R.; Ms. Pushpalatha S. S.; Mrs. Suhasini; Mrs. Divya G. S.; Mrs. Shwetha K. R.; Mrs. Vijayalakshmi M. M.; Mrs. Mahalakshmi B.; Dr. Thilagavathy R.; Dr. Kanike Vinod Kumar; and Mrs. Apoorvashree H. L..

The application for the patent was published on September 18, 2026, under issue no. 38/2026.

Abstract: An automated crop prediction system integrates field soil acquisition with machine-learning-based nutrient estimation and crop suitability inference. A soil sensing interface acquires soil-response data, and a calibration module validates and compensates the data for interfering conditions including moisture, temperature, conductivity, pH or sensor drift. A nutrient estimation engine derives plant-available nutrient values together with uncertainty indicators. A feature vector generator forms a soil-condition representation containing nutrient values, balance indicators and confidence information. A trained crop prediction engine generates suitability values for candidate crops, while an agronomic constraint evaluator suppresses or penalizes candidates that violate cultivation ranges. A ranking module combines model output with measurement quality, nutrient-estimation confidence and contextual completeness to produce one or more ranked crop predictions. Low-confidence measurements may trigger targeted reacquisition, and validated cultivation outcomes may be stored with a model version for controlled adaptation. The architecture thereby propagates soil-measurement reliability into crop prediction and provides field-specific, traceable and confidence-aware crop selection. Accompanied Drawing [FIGS. 1-2].

Disclaimer: Curated by HT Syndication.