Vinayaka et al. (2025) Harnessing AI and Remote sensing for precision sugarcane farming: tackling water stress, salinity, and nitrogen challenges

Identification

• Journal: Frontiers in Agronomy
• Year: 2025
• Date: 2025-11-05
• Authors: Vinayaka Vinayaka, P. Rama Chandra Prasad, G. Avinash, Amaresh Amaresh, Rajeev Kumar, P. Murali, C. Palaniswami, Govindaraj Perumal
• DOI: 10.3389/fagro.2025.1681294

Research Groups

• Statistics and Economics Section, ICAR-Sugarcane Breeding Institute, Tamil Nadu, India
• Lab for Spatial Informatics, International Institute of Information Technology, Gachibowli, Telangana, India
• Avyagraha Research and Analytics LLP, Ramasagara, Karnataka, India
• Division of Crop Improvement, ICAR-Sugarcane Breeding Institute, Tamil Nadu, India
• Plant Physiology Section, ICAR-Sugarcane Breeding Institute, Tamil Nadu, India
• Soil Science Section, ICAR-Sugarcane Breeding Institute, Tamil Nadu, India

Short Summary

This review synthesizes the application of artificial intelligence (AI) and remote sensing (RS) technologies for precision sugarcane farming, focusing on detecting and managing water stress, salinity, and nitrogen challenges to enhance crop productivity and environmental sustainability.

Objective

• To evaluate the contribution of AI and RS in assessing water stress in sugarcane cultivation.
• To examine methods for mapping salinity stress using RS and AI tools.
• To highlight the relevance of spectral indices in tracking nitrogen status in sugarcane crops.

Study Configuration

• Spatial Scale: Review of studies across field, plot, and regional scales, including large sugarcane plantations.
• Temporal Scale: Review of literature published from 1981 to 2025.

Methodology and Data

• Models used:
  • Artificial Intelligence/Machine Learning: Machine Learning (ML), Deep Learning (DL), Convolutional Neural Networks (CNNs), Random Forest (RF), Support Vector Regression (SVR), Support Vector Machine (SVM), Spectral Angle Mapper (SAM), Minimum Distance (MD), Maximum Likelihood algorithm (MLA), M5-pruned (M5P), Random Forest Regression (RFR), Gradient-Boosted Regression Trees (GBRT), Extreme Gradient Boosting (XGBoost), Long Short-Term Memory (LSTM), DenseResUNet, Inception-ResNet-v2, Artificial Neural Networks (ANNs).
  • Energy Balance Models: Surface Energy Balance Algorithm for Land (SEBAL), Simplified Assimilation of Evapotranspiration (SAFER), Spatial Evapo-Transpiration Modeling Interface (SETMI), Two-Source Energy Balance Model (TSEB).
  • Crop Models: CANEGRO, CANESIM.
  • Other: Climatological Soil-Water Balance (CSWB).
• Data sources:
  • Satellite Imagery: Landsat (TM, 7, 8), MODIS, Sentinel-2, Sentinel-1 (Synthetic Aperture Radar - SAR), Hyperion, CubeSats.
  • Sensor-based Data: RGB, multispectral, hyperspectral imaging, thermal imaging, unmanned aerial vehicles (UAVs), Light Detection and Ranging (LiDAR), soil moisture sensors, weather sensors, infrared (IR) thermometry, chlorophyll meters, Aquacheck probes.
  • Reanalysis Data: ERA5-Land.
  • Ground-based Data: Soil sampling, tissue analysis, eddy covariance (EC) system, meteorological data, Shuttle Radar Topography Mission (SRTM) digital elevation data, ground truth data.

Main Results

• AI-driven remote sensing is affirmed as a highly effective approach for monitoring and managing critical stress factors (water, salinity, nitrogen) in sugarcane, leading to enhanced yield, crop quality, and significant socio-economic and environmental benefits.
• Water Stress Assessment: Thermal and hyperspectral remote sensing, combined with energy balance models (e.g., SEBAL, CWSI) and deep learning models (e.g., DenseResUNet, Inception-ResNet-v2), provide reliable estimates of evapotranspiration and crop water stress, improving irrigation scheduling. The Crop Water Stress Index (CWSI) derived from Earth Observation data is a preferred metric.
• Salinity Stress Monitoring: Hyperspectral imagery integrated with machine learning classifiers (e.g., SVM, RFR) yields high classification accuracy for soil salinity, with optimized indices like OSAVI and VOG1 identified as reliable predictors. Landsat data is effective for categorical salinity mapping, while Hyperion is better for quantitative estimation.
• Nitrogen Level Estimation: Red-edge and Near-Infrared (NIR) vegetation indices (e.g., SAVI, MSAVI, NDVI, OSAVI) integrated with machine learning models (e.g., RF, SVR, ANNs) show strong correlations with leaf nitrogen content, offering scalable and non-destructive nutrient monitoring solutions.
• Data fusion approaches and cloud-based platforms (e.g., Google Earth Engine) enhance the spatiotemporal resolution and analytical efficiency of stress assessments.

Contributions

• Provides a comprehensive review of current knowledge on AI and remote sensing applications for assessing water stress, salinity stress, and leaf nitrogen content in sugarcane cultivation.
• Consolidates detailed information on advanced sensors, UAV technologies, and novel deep learning models that have received limited attention in prior studies.
• Identifies best-practice solutions for stress detection and management, while also highlighting persistent challenges (e.g., sensor limitations, data processing, model generalizability) and outlining future research directions.
• Emphasizes the transformative potential of AI-driven remote sensing for enhancing sugarcane productivity, promoting environmental sustainability, and delivering socio-economic benefits.

Funding

Research was supported by the Indian Council of Agricultural Research, Department of Agricultural Research and Education, Government of India.

Citation

@article{Vinayaka2025Harnessing,
  author = {Vinayaka, Vinayaka and Prasad, P. Rama Chandra and Avinash, G. and Amaresh, Amaresh and Kumar, Rajeev and Murali, P. and Palaniswami, C. and Perumal, Govindaraj},
  title = {Harnessing AI and Remote sensing for precision sugarcane farming: tackling water stress, salinity, and nitrogen challenges},
  journal = {Frontiers in Agronomy},
  year = {2025},
  doi = {10.3389/fagro.2025.1681294},
  url = {https://doi.org/10.3389/fagro.2025.1681294}
}