Mahto et al. (2025) IoT and AI Integration for Climate‑Smart Farming: A Predictive and Adaptive System for Smallholder Farmers

Identification

• Journal: International Journal of Research in Interdisciplinary Studies
• Year: 2025
• Date: 2025-12-17
• Authors: Bindeshwar Mahto, Rohit Kumar Rana, Niraj Kumar, Mithun Kumar, Arun K. Das, K. Mayank, Mithlesh Kumar Mahto, Sanjay Kumar Mahto
• DOI: 10.65138/ijris.2025.v3i12.238

Research Groups

Department of Computer Science & Engineering and Information Technology, Jharkhand Rai University, Ranchi, India

Short Summary

This research presents an integrated Internet of Things (IoT) and Artificial Intelligence (AI) based climate-smart farming system for smallholder farmers, which continuously monitors environmental conditions, predicts crop responses, and generates adaptive management recommendations. Field experiments demonstrated significant improvements in water efficiency and crop productivity.

Objective

• To develop and validate an integrated IoT and AI-based climate-smart farming system that provides real-time monitoring, predictive analytics, and adaptive management recommendations specifically tailored for smallholder farmers.

Study Configuration

• Spatial Scale: Three smallholder farms, covering three major crop varieties (e.g., rice, wheat, maize).
• Temporal Scale: Data collected over 90 days, with sensor sampling frequency every 30 minutes (0.5 hours), accumulating 11,200 sensor-hours.

Methodology and Data

• Models used:
  • Random Forest for irrigation prediction.
  • Long Short-Term Memory (LSTM) networks for yield forecasting.
  • Gradient Boosting for disease-risk estimation.
  • Rule-based adaptive module for generating recommendations.
• Data sources:
  • IoT sensor data: Soil moisture (volumetric water content), temperature, humidity, pH level, rainfall intensity, light intensity, and soil nutrient levels (Nitrogen, Phosphorus, Potassium).
  • Manual field observations: 240 samples.
  • Daily climate reports: 90 reports.

Main Results

• The system achieved an irrigation prediction accuracy of 96.2%.
• Disease-risk detection accuracy was 93.7%.
• Yield prediction demonstrated a Root Mean Square Error (RMSE) of 0.18.
• Field deployment resulted in 27% water savings compared to baseline practices.
• Productivity gains ranged from 12% to 18% in yield improvement.
• Disease detection was 2 to 5 days earlier than manual identification.
• Farmer feedback indicated 89% reported improved decision-making, 76% found the application intuitive, and 82% saved on irrigation costs.

Contributions

• An integrated six-layer IoT-AI architecture unifying real-time sensing, predictive analytics, and adaptive recommendations.
• A multi-model AI framework combining Random Forest, LSTM, and Gradient Boosting for simultaneous prediction of irrigation needs, disease risk, and yield outcomes.
• A low-cost sensing deployment designed for resource-constrained smallholder farms.
• A rule-based adaptive module that translates predictions into actionable advisories.
• Field validation demonstrating significant improvements in water efficiency (27% savings) and productivity (12–18% increase).

Funding

Not explicitly mentioned in the paper.

Citation

@article{Mahto2025IoT,
  author = {Mahto, Bindeshwar and Rana, Rohit Kumar and Kumar, Niraj and Kumar, Mithun and Das, Arun K. and Mayank, K. and Mahto, Mithlesh Kumar and Mahto, Sanjay Kumar},
  title = {IoT and AI Integration for Climate‑Smart Farming: A Predictive and Adaptive System for Smallholder Farmers},
  journal = {International Journal of Research in Interdisciplinary Studies},
  year = {2025},
  doi = {10.65138/ijris.2025.v3i12.238},
  url = {https://doi.org/10.65138/ijris.2025.v3i12.238}
}