Machine Learning-Powered Nano-Energy Systems for Sustainable Precision Agriculture and Soil Health Monitoring

Authors

  • Shanthi Palaniappan Author
  • J M Prabhudass Author
  • V Vijayalakshmi Author
  • S. Srividhya Author
  • Manoj Praveen V Author
  • Praveenkumar. C Author
  • M.P. Thiruvenkatasuresh Author
  • S.D. Vijayakumar Author

Keywords:

Precision Agriculture; Nano-Energy Harvesting; PHOENIX-Ag; Machine Learning; Soil Health Monitoring

Abstract

Precision agriculture requires autonomous sensing technologies capable of continuously monitoring soil conditions while reducing energy consumption, maintenance requirements, and dependence on conventional batteries. Nano-energy harvesting provides a promising pathway for powering distributed agricultural sensors through ambient solar, mechanical, thermal, and vibrational resources; existing approaches frequently optimize energy harvesting, sensing, and soil-health prediction independently, limiting adaptability under dynamic field conditions. This paper proposes PHOENIX-Ag (Physics-guided Hybrid Optimization of Energy and Intelligent Nano-Sensing for Agriculture), a machine-learning-powered framework integrating hybrid nano-energy harvesting, physics-guided energy prediction, adaptive sensing, and intelligent soil-health assessment. PHOENIX-Ag combines photovoltaic, triboelectric, and piezoelectric nanogenerators to generate distributed energy, while a physics-guided learning model predicts energy availability from environmental, harvesting, and operational variables. A hybrid optimization mechanism dynamically regulates sensing frequency, energy allocation, storage utilization, and wireless communication according to predicted energy availability and soil state dynamics. Multimodal soil observations, including moisture, temperature, pH, electrical conductivity, and nutrient indicators, are processed for soil-health classification and critical measurement prioritization. Experimental results demonstrate 98.1% energy-availability prediction accuracy, 97.2% soil-health classification accuracy, 95.1% energy-harvesting efficiency, and 94.5% adaptive sensing efficiency, with 286 mJ energy consumption and 329 days of sensor operational lifetime. Compared with conventional adaptive agricultural monitoring, PHOENIX-Ag reduces energy consumption by 30.6% and extends operational lifetime by 34.8%. These results establish PHOENIX-Ag as an energy-intelligent framework for sustainable, autonomous, and reliable soil-health monitoring in precision agriculture.

Downloads

Download data is not yet available.

Downloads

Published

2026-09-22

How to Cite

Machine Learning-Powered Nano-Energy Systems for Sustainable Precision Agriculture and Soil Health Monitoring . (2026). Minimax Theory and Its Applications, 11(2s), 490-507. https://journalmta.com/index.php/jmta/article/view/431