Machine Learning and Remote Sensing for Soil Moisture and Nutrient Estimation: A Systematic Review and Future Research Roadmap

Machine Learning and Remote Sensing for Soil Moisture and Nutrient Estimation: A Systematic Review and Future Research Roadmap | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Short Report Machine Learning and Remote Sensing for Soil Moisture and Nutrient Estimation: A Systematic Review and Future Research Roadmap Dhara Prajapati, Abishek Murugesan, Rucha Dave This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8220871/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The need for innovative and effective agricultural practices is now more than higher due to the growing demand for food worldwide and the strain on land and water resources. In order to make farming more data-driven, focused, and sustainable, precision agriculture (PA) provides a potent solution by utilizing technologies such as remote sensing, artificial intelligence (AI), machine learning (ML), and the Internet of Things (IoT). This paper discusses the various ways in which precision agriculture is revolutionizing each step of agricultural production, from supply chain optimization and pest and disease detection to soil health evaluation and smart irrigation. Based on current research and practical uses, particularly in India, we demonstrate how technologies like satellite images, unmanned aerial vehicles (UAVs), artificial intelligence (AI)-powered sensors, and automated equipment assist farmers in improving decision-making, cutting waste, conserving resources, and increasing output. Even while PA technologies are becoming increasingly popular, issues including excessive costs, a lack of regulations, and limited availability for smallholder farmers still exist. This study emphasizes how important precision agriculture is to creating a farming system that is more robust, effective, and prepared for the future. Precision Agriculture (PA) Remote Sensing (RS) Artificial Intelligence (AI) Machine Learning (ML) Internet of Things (IoT) Figures Figure 1 Figure 2 Figure 3 Introduction The current global population of 7.6 billion is estimated to increase to 8.6 billion by 2030 and 9.8 billion by 2050 as per the report from the United Nations (UN) Economic and Social Affairs department (R-The World Population Prospects: The 2017 Revision). It is anticipated that the world's need for cereals both for human consumption and animal feed will surpass 3 billion tonnes by 2050, (FAO, 2009). The critical shortages of arable land and water resources, combined with the growing risks of erratic weather, extreme temperatures, and unpredictable rainfall, emphasize the urgent need for precision farming. Additionally, rising costs of fertilizers, pesticides, seeds, fuel, along with agricultural labor force shortage, makes it essential to adopt precision farming techniques that optimize resource use, minimize waste and improve efficiency through automation and data-driven decision-making. Apart from the production constraints, Farmers also struggles with market fluctuations and lack real-time market insights. Such situation demands for a Precision Agriculture (PA) because PA integrates cutting-edge technologies to boost resource efficiency, sustainability and Agricultural productivity. It enables farmers to make informed decisions and adapt to dynamic climatic conditions by utilizing advanced weather forecasting, remote sensing, and AI ( Artificial Intelligence ) driven analytics. The use of Automation, robotics, and intelligent machinery alleviate labor shortages by streamlining operations and increasing production efficiency. Soil analysis, precision nutrient application, and site-specific fertilization contribute to maintaining soil health and promoting sustainable farming practices. Smart irrigation systems, including drip irrigation and IoT-based moisture sensors, optimize water usage, addressing the challenges of water scarcity. Drones, AI-driven pest and disease detection, and precise pesticide application provide early disease and pest management, minimizing unnecessary chemical usage. Real-time monitoring technologies, enable the early detection of crop stress and potential threats. Additionally, data-driven forecasting, digital farm management, and intelligent logistics solutions improve supply chain efficiency and reduce post-harvest losses. Precision farming further contributes to sustainability by decreasing reliance on chemicals, encouraging environmentally friendly agricultural practices, and ensuring compliance with environmental regulations. This review paper focus on how these advancements collectively enhance agricultural productivity, optimize resource use, and promote a more resilient and sustainable food production system for the future. 2. Implementing Precision Agriculture and Acquiring Data Using Remote Sensing, Artificial Intelligence and Machine Learning In precision agriculture, satellite data plays a transformative role by enabling data-driven, site-specific management of crops, soil, and other farm resources. This approach helps maximize yields, optimize inputs (like water, fertilizer, and pesticides), and promote sustainable farming practices. Earth Observation (EO) satellite data, accessible through portals plays a vital role in transforming modern agriculture. Some of the major popular Portals are mentioned in Table 1 . One of its primary applications is crop monitoring, where multi-temporal satellite imagery is used to assess crop growth, detect early signs of stress, and identify disease or pest outbreaks. Satellite data also supports crop type mapping, helping distinguish between different crops based on their spectral signatures, which is essential for land-use planning, agricultural statistics, and policy development. Another significant application is yield estimation, where vegetation indices like NDVI, combined with machine learning models, can predict crop productivity, aiding food security assessments and market forecasting. Soil monitoring using radar and thermal data allows for the evaluation of moisture levels, texture, and nutrient content, leading to better soil health management and optimized fertilizer application. In terms of irrigation management, satellite data helps assess water stress and schedule irrigation more efficiently, conserving water and reducing operational costs. Furthermore, EO data is critical in disaster impact assessments, such as evaluating crop damage from floods, droughts, or pest invasions, which is essential for relief planning and insurance claims. The integration of such data with AI and IoT allows farmers to make data-driven decisions at the micro-level, thereby increasing productivity while minimizing input costs. Satellite imagery also contributes to land use and land cover (LULC) analysis, providing insights into agricultural land dynamics and supporting sustainable development initiatives. Review methodology The methodology for selecting diverse literature to review the application of Remote Sensing, Precision Agriculture, and AI/ML begun with a comprehensive search across various multidisciplinary online databases, including Springer, Taylor & Francis, Wiley, Google Scholar, MDPI, Annual Reviews, Science Direct, Scopus, Web of Science, IEEE Xplore, and other Scopus-indexed journals relevant to various agricultural domains utilizing those approaches. Several academic publications about surface soil, soil parameters, plant physiology, different irrigation techniques, pest, disease, weed assessment, detection and control strategies, harvesting and yield optimization and supply chain logistics found in these large database research libraries. Many significant studies were mentioned in the Table 2 and the most recent, creative and relevant research papers and reviews were chosen for this study. The following keywords: artificial intelligence, machine learning, supply chain logistics, remote sensing, soil management, irrigation, plant diseases, and pests were used for search. To guarantee that the concepts were consistent, the articles were carefully chosen, reviewed, and summarized. Table 1 Popular Earth Observation (EO) satellites data and information Portals from India and world-wide Sr. No. Earth Observation (EO) satellite data Portals Information provided by portals Indian Portals 1. VEDAS (Visualisation of Earth Observation Data and Archival System) ISRO’s platform for EO data visualization and thematic applications 2. BHUVAN (ISRO) Geo-portal for visualizing and downloading satellite data and thematic layers 3. MOSDAC (Meteorological and Oceanographic Satellite Data Archival Centre) Specializes in meteorological, oceanographic, and climate data from ISRO satellites 4. India WRIS (Water Resources Information System) Jointly developed by ISRO and the Ministry of Jal Shakti for water resource monitoring 5. BHOONIDHI Geo-portal Developed by ISRO to provide streamlined access to Earth Observation (EO) satellite data International Portals 6. Copernicus Open Access Hub (ESA - Europe) Free access to Sentinel satellite data (Sentinel-1, -2, -3, and − 5P) 7. USGS Earth Explorer (USA) Offers access to Landsat, MODIS, ASTER, and other datasets 8. NASA Earth data Search Provides a wide range of NASA satellite missions including MODIS, VIIRS, and SMAP 9. Google Earth Engine Cloud-based platform offering access to petabytes of EO data for analysis and visualization 10. Open Data Cube (ODC) Open-source platform providing standardized access to EO data for analytics Precision Agriculture can be practiced at different crop stage ranging from seed selection to harvesting and logistics. All these applications are briefly given (Table 2 ) in tabular form further discussed here under. Table 2 Applications of Precision Agriculture, AI and ML Technologies in different agricultural domains Sr. No. Study Objective Instrument used ( In-situ data) Optical data used Optical data obtained from Methods Study Area Reference 1. Soil textures and nutrients estimation Ground data Landsat − 8 OLI/TIRS C2- L1, Landsat − 8 OLI/TIRS C2- L2, Sentinel 2A United States Geological Survey (USGS) Gradient Boosting (GB), Extreme Gradient Boosting (XGB), Random Forest Regressing (RFR), Adaptive Boosting (ADA) North India-Punjab Dhiman, et al .,(2023) 2. Surface Soil Moisture (SSM) estimation Gravimetric method LISS-III sensor multispectral imagery with a spatial resolution of 23.5 meters Indian Earth observation satellite Resourcesat-2 LISS-III sensor Modified Dubois Model (MDM) for estimating SSM Bhal Region Gujarat (Wheat crop) Dave et al .,(2019) 3. Soil mapping and soil texture estimation Gravimetric method Sentinel-1 (S-1) and Sentinel-2 (S-2) Copernicus (in the form of Level-1 Ground Range Detected (GRD) products support vector machine (SVM) and random forest (RF) algorithms with a three-fold cross-validation approach semi-arid region in central Tunisia Bousbih et al. ( 2019 ) 4. Surface Soil Moisture (SSM) estimation HydraGo Probe sensor and surface roughness synchronizing with satellite pass dates Sentinel-1 C-band SAR data Copernicus Ground Range Detected (GRD) products Modified Dubois Model (MDM) for estimating SSM Anand District of Gujarat Murugesan, et al. ( 2023 ) 5. Operational 500 m surface soil moisture estimation handheld soil probes C-band SAR data ISRO’s high resolution EOS-04 (RISAT-1A) RT model (tau-omega model) Anand (Gujarat), Ludhiana (Punjab) Hisar (Haryana) Kanpur (Uttar Pradesh) and Berambadi watershed (Karnataka) Pandey et al. ( 2024 ) 6. Effects of soil compaction on canopy reflectance cotton yield, soil EC digital cone penetrometer, conductivity meter, spectro-radiometer (EPP 2000C) Ground-based spectral data - Green NDVI Fayetteville, Arkansas Kulkarni et al. ( 2010 ) 7. Assessing potential of polarimetric decomposition parameters for monitoring rice crop biophysical parameters Ground data C-band SAR data from Sentinel-1 C-band SAR data from Sentinel-1 Field surveys synchronized with Sentinel-1 satellite passes Anand District of Gujarat, India Dave et al. ( 2023 ) 8. To analyze and compare performance of three commercial soil nitrate probes Nutrisens, RIKA, JXCT (conductivity-based probes) - - Gravimetric method University of Barcelona, Spain Bellosta-Diest et al. ( 2022 ) 9. To predict organic carbon (OC), pH, available N, P, K (EC), zinc (Zn), soil texture Portable VIs–NIR spectrometer (350–2,500 nm) and conventional lab instruments - Sentinel-2 satellite imagery (for spatial mapping) Ground data (sample collection) and PLSR,SVMR modelling (with PCA) to predict soil variables Tarekswar region, Hooghly district, West Bengal, India Singha et al. ( 2023 ) 10. To develop and test a solar-powered smart irrigation control system Soil moisture & water level sensors, Microcontroller (for automation), - - Gravimetric method validation Kabanyolo, Uganda Wanyama et al. ( 2023 ) 11. To develop and demonstrate a solar-powered IoT irrigation system In-situ data - - Bangladesh Agricultural University, Mymensingh, Bangladesh Al Mamun et al. ( 2025 ) 12. Crop transpiration estimation Infra-Red Camera With UAV system - - Three-temperature (3T) model, canopy segmentation, scale error analysis, isotope validation Liangzhou District, Wuwei City, Gansu province, of Northwest China Hou et al. ( 2021 ) 13. Assessment of Okra YVMV with ground based Hyperspectral imaging Non-imaging Spectroradiometer (SVC-LC-RPPro 350 to 1050 nm) - - NDVI ,RVI, GI, PRI, MCARI, and RVSI, RENDVI Anand District of Gujarat, India Mishra et al. ( 2023 ) 14. Assessment of aphid infestation in Mustard with Hyperspectral imaging FieldSpec-3 hyperspectral spectroradiometer (350–2500 nm) - - NDVI, RVI, Aphid Index (AI), and SIPI Indian Agricultural Research Institute (IARI), New Delhi, India Kumar et al. ( 2013 ) 15. faba bean leaf disease identification with Deep CNN Digital camera - - CNN Architecture based image processing Republic of Korea Jeong and Na ( 2024 ) 16. ML-IoT Sensor based smart, continuous crop disease monitoring system development IoT sensor nodes - - Ensemble nonlinear SVM (ENSVM), CNN, Naïve Bayes, and K-Nearest Neighbors - Nagasubramanian et al. ( 2021 ) 17. AI-powered spot-sprayer to detect/spray weeds precisely Logitech C920 RGB cameras; NVIDIA TX2 & GTX 1070 Ti GPUs; TOPCON RTK GPS - - Hardware setup (pump, nozzles, manifold), Tiny YOLOv3 detection, GPS-enabled nozzle actuation Florida, USA Partel et al. ( 2019 ) 18. UAV-based weed patches detection within/between wheat rows using OBIA. UAV with RGB + multispectral sensors, flying at 30–60 m. - - VI thresholding, object segmentation (OBIA), connected-component classification Pakistan Mateen and Zhu ( 2019 ) 19. Monitoring of harvesting combine with smart devices GPS receiver model: (Teltonika FMB92) for real-time tracking Grain-level sensor installed in the tank to monitor fill levels - - - Tashkent Region, Uzbekistan Astanakulov et al. ( 2021 ) 20. Monitoring of harvesting combine with smart devices to minimize seed loss in sunflower harvesting GPS receiver (Teltonika FMB920) and grain-level sensor (Escort DB-2) mounted on combine - - - Tashkent Region, Uzbekistan Ochildiev et al. ( 2021 ) 21. IoT-based autonomous mobile robot system for pitaya harvesting integrating SLAM, AI object recognition NVIDIA Jetson Nano development board, 2D LiDAR (for SLAM), AI edge computing module, robot arm/gripper with sensors - - 2D SLAM for mapping/navigation; YOLO-based AI for fruit detection Chen et al. ( 2023 ) 22. real-time detection and fruit load estimation in mango orchards RGB camera with 720 W LED flood lighting mounted on a vehicle at 6 km/h - - Compared 6 deep learning architectures (e.g., Faster R-CNN, YOLO variants Australia Koirala et al. ( 2019 ). 23. IoT-enabled mechatronic module for real-time yield monitoring during harvesting of greenhouse and vineyard produce Load sensor-based IoT weighing system on a robotic arm/gripper (gripper-cum-cutter), connected via MQTT to server - - - National Research Centre for Grapes & greenhouses in Pune district, Maharashtra, India Kolhalkar et al. ( 2022 ) 24. Crop yield prediction algorithm IoT sensors across fields collecting climate (temperature, humidity), meteorological, soil chemical, and yield data - - Machine-learning regressors (DecisionTree, RandomForest, ExtraTree), with an active-learning approach to minimize labeled data needs Kafrelsheikh, Egypt Talaat ( 2023 ) 25. Forecast wholesale brinjal prices using various ML models relied on time-series price data from Agmarknet - - GRNN, SVR, RF, GBM vs ARIMA; evaluated using RMSE, MAE, MAPE, Diebold-Mariano tests Odisha, India Paul et al. ( 2022 ) 26. Deep LSTM for Agricultural Price Forecasting historical market price datasets - - Developed deep LSTM model trained on time series of commodity prices Jaiswal et al. ( 2022 ) 27. Risk assessment and monitoring in fresh produce logistics Utilized RFID logs and IoT sensor data (e.g., temperature/humidity in supply chain) - - SVM to predict logistics risk score based on sensor-instrumented shipping data China Zhang et al. ( 2020 ) 28. IoT-Based Real-Time Monitoring and Notification System of Cold Storage Sensors for temperature, humidity, door motion, connectivity via IoT modules - - Web-dashboard alerts, mobile notifications; threshold-based monitoring Pakistan Afreen and Bajwa ( 2021 ) 29. perishable food traceability in supply chain RFID tags + IoT environmental sensors (temp/humidity) on packaging - - Combined RFID and sensor data; ML models identified anomalies to prevent spoilage Indonesia Alfian et al. ( 2020 ) 2.1 Soil Management: Accurate estimation of soil and plant nutrient levels is critical for enhancing crop growth and optimizing agricultural yield. Traditional soil analysis techniques, while accurate, are often labor-intensive and time-consuming effort. Innovative UAVs (Unmanned Aerial Vehicles) mounted with hyperspectral and multispectral imaging are precise but costlier which is why it is not widely adopted. However, with the use of advanced remote sensing technologies such challenges can be accomplished. Dhiman and their co-workers (2023) investigated such an alternative approach by utilizing open-source multispectral satellite imagery in combination with ground truth data collected from the agriculture-intensive Punjab region of India. They estimated concentrations of nine key soil nutrients and textures, i.e. , Nitrogen (N), Phosphorous (P), Carbon (C), Ammonium (N-NH₄), Nitrate Nitrogen (N-NO₃), Calcium (Ca), Magnesium (Mg) and texture Clay and Silt. Further satellite imagery data quality was improved by applying DOS1 atmospheric correction which was compared against raw (uncorrected) datasets. Soil nutrients and textures estimation tasks performed using four ensemble machine learning models. Their findings indicated that raw imagery consistently outperforms the DOS1-corrected data across most parameters. Moreover, the analysis shows that certain soil properties have a stronger correlation with multispectral data, suggesting that remote sensing has significant potential in nutrient and texture estimation, offering a cost-effective and scalable solution for soil monitoring. Sustainable agriculture depends on efficient soil and land management, which is influenced by topographical features (such as elevation and slope), soil characteristics (including moisture content, organic matter, pH, drainage, and electrical conductivity), along with physiological, biochemical, and morphological and crop traits (Nock et al., 2016 ). Proper soil mapping, erosion control and irrigation management are essential for enhancing crop productivity. Additionally, factors like crop water needs, water and energy balance, runoff, evapotranspiration, and climatic modelling contribute significantly to agricultural efficiency. Among these, Surface Soil Moisture (SSM) plays a crucial role in regulating irrigation practices, ensuring water availability, and improving overall crop performance. In this direction, Dave et al. ( 2019 ) evaluated Modified Dubois Model (MDM) for estimating surface soil moisture (SSM) using dual-polarization C-band SAR data from RISAT-1 over winter wheat crop from initial vegetative to maturity stage within Bhal region of Gujarat. By incorporating field-measured soil moisture and surface roughness across different crop stages (bare soil, sowing, vegetated), the model was calibrated and tested for σ⁰ (VV and VH) backscatter sensitivity. The MDM showed strong correlation with measured soil moisture, especially for bare and sparsely vegetated conditions, outperforming the original Dubois model. Validation with independent datasets confirmed its robustness for operational SSM retrieval in agricultural landscapes using RISAT-1 data. Agrawal et al. ( 2019 ) developed and evaluated smart Soil Sensor for Hydrology and land Applications (SHOOL) which is a low-cost, compact soil moisture sensor developed by (Space Applications Centre-Indian Space Research Organisation) SAC-ISRO for field-based soil moisture monitoring. It measures dielectric constant, electrical conductivity, temperature, and volumetric soil moisture, transmits data via a smartphone app, making it suitable for dense sensor networks aimed at validating satellite-derived soil moisture products. They compared performance of two commercial sensors—MP306 (ICT) and Stevens Hydra Probe with SHOOL and validated the data using gravimetric sampling across agricultural fields in Gujarat. They claim that SHOOL performed better than both commercial sensors, achieving R² = 0.82 and RMSE = 2.94, indicating strong correlation and low error based on statistical analysis. To support effective soil and land analysis, various technologies such as soil moisture sensors with probes (Hydrago-stevans, Shool-Neerx), GIS mapping, and remote sensing data from various satellites and aircrafts can be utilized not only they are free but also providing valuable insights for precision agriculture. Alternatively, integrating remotely sensed optical and radar data for can be used for soil mapping, in this direction, Bousbih et al. ( 2019 ) used and integrated remotely sensed optical and radar data for soil mapping and soil texture estimation by utilizing Sentinel-1 (S-1) and Sentinel-2 (S-2) imagery acquired between July and December 2017 for 3000 km² semi-arid region in central Tunisia. They found clay content ranging from 13% to 60% from multiple agricultural fields alongside satellite-derived indicators, including Short-Wave Infrared (SWIR) bands and soil indices, particularly over bare soils and soil moisture products generated from combined S-1 and S-2 data were also tested as proxies for soil texture. They also applied support vector machine (SVM) and random forest (RF) algorithms for clay content mapping, with a three-fold cross-validation approach for assessment. They found high accuracy classification while using soil moisture indicators derived from combined S-1 and S-2 data, yielding overall accuracies of 63% and 65% for SVM and RF, respectively. Thus, indicating potential of integrating multi-source remote sensing data for improved soil texture mapping in semi-arid environments. The advanced Soil Moisture probes (Fig. 1 , 2 ) can measure soil properties (Moisture, temperature, Bulk EC, Pore water EC, raw real dielectric) along with latitude, longitude for geo referencing. Further it is very easy to transfer data digitally. Murugesan, et al. ( 2023 ) evaluated the Modified Dubois Model (MDM) for estimating surface soil moisture (SSM) in bare agricultural fields using Sentinel-1 C-band SAR data in a semi-arid region of Gujarat, India. They collected field soil moisture data from 102 locations, using HydraGo sensors (Fig. 1 ), gravimetric methods and estimated surface roughness for validated SAR-derived soil moisture values. Further applied MDM to derive soil permittivity from VV-polarized backscatter, which was then converted to volumetric moisture using Topp’s model. They noted strong agreement with ground measurements (R² = 0.81, RMSE = 0.005 m³/m³), especially in the low to moderate moisture range (0.015–0.1 m³/m³) also reduced accuracy observed at higher moisture levels due to surface roughness effects. They confirm that MDM, coupled with Sentinel-1 data, is a reliable tool for SSM monitoring in bare soils and holds potential for improving irrigation planning and water resource management. 2.2 Sowing and Planting Operations using Precision Agriculture Germination and Seedling emergence are highly influenced by soil moisture and soil temperature. In the same way, fertilizers are usually applied when fields are not excessively wet to reduce nutrient loss ultimately enhancing crop nutrient uptake. Seedling rate for different crops, soil types and variable seasons can be dictated by conducting land mapping studies as it impacts hydrological balance influencing soil moisture and temperature (Khanal et al., 2020 ). Microwave signals are considered more accurate for the soil moisture estimation as they are not get interfered by atmospheric and cloudy conditions unlike the RS data obtained from visible, near-infrared (NIR), and short-wave infrared (SWIR) bands. So, various satellites equipped with such active and passive microwave sensors, i.e. , Soil Moisture Active Passive (SMAP) and Sentinel-1, are currently used for soil moisture monitoring. In order to meet the need for sub-km resolution soil moisture data in agriculture, Pandey et al. ( 2024 ) developed a downscaling framework to disaggregate coarse brightness temperature from the SMAP radiometer using ISRO’s high-resolution EOS-04 (RISAT-1A) C-band SAR data. Another crucial parameter is soil compaction, it impacts hydrolic conductivity, porosity and nutrient availability of soil so it is necessary to assess soil compaction. This information allows end user (farmer) to make a decision about crop selection based on soil properties. In this direction, Kulkarni et al. ( 2010 ) in Arkansas, examined the impact of soil compaction on canopy spectral reflectance and cotton yield using ground-based spectral data (using cone penetrometer) and demonstrated that soil compaction effects can be evaluated by analyzing variations in the Green Normalized Difference Vegetation Index (GNDVI), which is derived from the green and near-infrared (NIR) spectral bands of imagery. The spatial and temporal variations in crop population distribution and density within a field can significantly influence grain and biomass yields, as well as overall efficiency of harvesting. Timely access to this information is critical for replanting and optimizing mid-season management practices, including the precise application of variable-rate fertilizers, herbicides, and pesticides. Very few studies based on the quantification of crop growth stages has been carried out in that direction, Dave et al. ( 2023 ) analysed potential of polarimetric decomposition parameters of Sentinel–1 dual-polarization SAR (Synthetic Aperture Radar) data in monsoon, to estimate biophysical parameters of rice crop i.e. , fresh biomass, dry biomass, vegetation water content (VWC) and plant height and noted that the polarimetric decomposition parameters are sensitive to biophysical parameters and can be used as additional parameters along with the backscatter parameters for rice crop monitoring. Moving on to sensors that can assist in assessing soil fertility and properties prior to sowing and planting operations, it is essential to recognize their significance in optimizing agricultural productivity. Accurate and timely evaluation of soil characteristics plays a critical role in ensuring informed decision-making regarding land preparation, nutrient management, and crop selection. In india, Innovative on-site soil testing and monitoring solutions, such as SoilSens and NutriSens have been developed to facilitate real-time assessment of soil health and support data-driven decision-making in agricultural management. Electrochemical sensors for soil quality assessment offer a cost-effective means of providing real-time, accurate insights into nitrate concentration trends, thereby supporting optimal fertilization practices and minimizing nutrient-related environmental impacts on surface and groundwater, Bellosta-Diest et al. ( 2022 ) conducted nitrate concentration estimation experiment in the soil matrix and evaluated the performance of three electrochemical sensors i.e. , Nutrisens, RIKA and JXCT, across three scenarios: (1) varying electrical conductivity (EC) in the absence of nitrate (0–56 mS/cm), (2) nitrate concentration in aqueous solutions (0–180 mg/L), and (3) nitrate concentration in saturated sand, used as a soil proxy, at 35% volumetric moisture (0–150 mg/L). Among the three, the Nutrisens probe, which utilizes an ion-selective electrode (ISE), demonstrated high sensitivity and consistent, accurate performance, exhibiting a strong inverse relationship with nitrate concentration (R² ≈ 0.99). In contrast, the RIKA and JXCT sensors, which are presumed to rely on conductivity-based measurements, were highly affected by EC variations, showed inconsistent outputs, and displayed poor sensitivity in soil-like conditions likely due to inadequate calibration and a lack of ion specificity for nitrate. Another study of soil nutrients prediction was conducted by Singha et al. ( 2023 ) employing Visible to Near-Infrared (VIs–NIR) spectroscopy in the range of 350–2500 nm, in combination with Partial Least Squares Regression (PLSR) and Support Vector Machine Regression (SVMR) models. The research, carried out in the Tarakeswar region of West Bengal, utilized 200 soil samples along with Sentinel-2 satellite imagery to estimate key soil parameters, including organic carbon (OC), NPK (nitrogen, phosphorus, potassium) pH, and soil texture (sand, silt, and clay). Both PLSR and SVMR models demonstrated high accuracy in predicting OC and P, with moderate performance for other nutrients, while predictions for soil texture were found to be less reliable. Using Sentinel-2 imagery and geostatistical techniques, the study also generated soil suitability maps, achieving up to 89% accuracy demonstrating potential of integrating machine learning and spectroscopy for cost-effective, efficient, and non-destructive soil monitoring. 2.3 Precision Irrigation/ Water Management In Indian agriculture, Rainfall is a significant water source followed by irrigation but in Indian context, rainfall is very inconsistent. Sometimes, excessive amount of it can cause soil erosion and surface runoff while other times, drought conditions can also occur inducing water scarcity. In such situation, irrigation, a regulated technique of supplying crops with water in place of natural rainfall, is frequently used to lessen these impacts, making sure that plants get the moisture as and when they need for healthy growth and development (Oborkhale et al., 2015 ). In conventional irrigation system, water is delivered uniformly, throughout the farm, without accouting for crop-specific water requirements or field variability which causes inefficiencies such as over-irrigation in some areas and under-irrigation in others, inducing water stress. In addition, commercial irrigation controllers are operated by pre-set programs based on empirical data. Furthermore, challenges like dynamic nature of plant water intake and weather patterns, environmental disruptions, and water scarcity driven by drought and climate change must need consideration. (Abioye et al., 2020 ). Under such scenario, precision irrigation reduces the wastage of water by considering spatial and temporal soil variation, soil structure, hydraulic properties, plant responses to water deficit, changing weather variables through efficient Internet of Things (IoT) monitoring, to make better irrigation decisions that could lead to increased production and significant water savings, since most countries in the world are facing water scarcity problems (Bitella et al., 2014 , Capraro et al., 2018 ). According to information reviewed by Abioye et al. ( 2020 ), subsurface irrigation is highly effective in water-saving and gives better yield output when compared with other type of surface irrigation. Different types of soil moisture sensors (Figure: 1,2,4), temperature and humidity sensors are used to monitor environmental conditions and further the data from such sensors can be collected and sent to the cloud using IOT platforms. Then this information is analysed by machine learning algorithms and artificial intelligence. In precision agriculture, IOT platforms are extensively used. Sometimes electricity/power lack is big concern. Harishankar et al. ( 2014 ) have developed power efficient solar based systems wherein the controller was linked to the soil sensor and water supply valve. Moisture sensor measures the water level and decides whether to switch the water valve ON or OFF. Since the solar panel provided the electricity, the system didn't require an external power module. Wanyama et al. ( 2023 ), developed a mobile, solar-powered control system for real-time irrigation scheduling and water delivery, using data from soil moisture sensors. The system included an automated tank water level control mechanism that activates the pump during irrigation. Soil moisture sensors were placed at a rooting depth of 15–30 cm, where two probes allowed electrical current to pass through the soil, measuring resistance to determine moisture levels. Wet soil, being a better conductor, exhibited lower resistance, whereas dry soil, with poor conductivity, showed higher resistance. The study also confirmed that the soil moisture sensors provided readings with no significant difference from the gravimetric method. In same way, Al Mamun et al. ( 2025 ) also proposed an IoT-enabled solar-powered smart irrigation for precision irrigation, included solar panel for power, Raspberry Pi 4 for pump control, and a 12V 7.5Ah battery for energy storage. Microcontroller was integrated with a Capacitive Soil Moisture Sensor and other sensors to monitor air temperature, and humidity data. They also created a website which aids in real-time monitoring and to control irrigation pump remotely. Zhou et al. ( 2021 ) reviewed use of a non-destructive, rapid, and reliable method to assess crop water stress using infrared thermal imagery, from early satellite-based systems to modern uncooled thermal cameras mounted on ground-based platforms and unmanned aerial vehicles (UAVs) to optimize irrigation and maintain crop productivity. They used canopy segmentation and the Crop Water Stress Index (CWSI), which correlate intensely with physiological indicators such as leaf water potential and stomata conductance giving valuable insights of plant health. Further, integration of deep learning enhance the accuracy of stress detection and prediction, enabling more efficient and data-driven irrigation scheduling especially in water-scarce regions. Hou et al. ( 2021 ) conducted a study using thermal infrared (TIR) imaging to estimate crop transpiration in corn and soybean under regulated deficit irrigation. They used a three-temperature (3T) model to analyze the scale effect between ground and UAV thermal images. This allowed them to detect water stress early, improve water use, and avoid excessive and poor watering, enabling precision irrigation. UAV-based thermal imaging captured large field spatial variability, identifying stressed areas for targeted actions like fertilizer application or irrigation modifications. This approach allows for comprehensive crop health monitoring, including illnesses and nutrient deficiencies. Hence, using precise sensors to monitor soil moisture levels can help optimize irrigation by determining how much water to apply and where it is needed. Integrating AI and machine learning-driven systems allows for real-time water level adjustments while incorporating local climate data can enhance irrigation efficiency. 2.4 Crop Health & Disease Management The production of crop plants is hampered by both biotic and abiotic factors, such as weeds, viruses, diseases, arthropods, and other animal pests. The potential loss caused by Fungi, oomycetes, bacteria, and viruses and the actual losses have been predicted by Oerke ( 2020 ) to reach 16% and 11% of the world's achievable crop production, respectively. So, in order to enhance crop productivity, crops must be protected from such large per cent of damage. So as to manage any pest or disease, correct diagnosis is crucial. Fields are sprayed at constant rates because disease management in agricultural crops frequently assumes that pathogens are dispersed uniformly. However, crop heterogeneity, which is influenced by geography, soil conditions, neighbouring fields, microclimate shifts, and infection sources, usually leads to irregular disease patterns that manifest as patches, gradients, or haphazard clusters. In the same field, these patterns might change throughout the course of a single growing season, as well as between places and years. In order to determine whether spraying is required, farmers typically visually evaluate a representative number of samples to determine the occurrence and severity of the disease and take decision whether to spray or not in field. Since, disease diagnosis and quantification currently rely on visual rating of plants which may be inaccurate and when in doubt, special serological and nucleic acid technologies i.e., ELISA (Enzyme-Linked Immunosorbent Assay), LAMP (Loop-Mediated Isothermal Amplification), PCR (Polymerase Chain Reaction), PAGE (Polyacrylamide Gel Electrophoresis) can be performed but those laboratory procedures are destructive and invasive covering only a representative sample of crops of interest. So, here remote sensing approaches can be a great help especially with the diseases which can induce latent or slow infection i.e., viral diseases, continuous monitoring can help in prediction of their occurrence and further spread. Apart from that, remote sensing technologies like Thermography, fluorescence imaging, and spectral techniques all are non-invasive and non-destructive which potentially performs repeated monitoring of all plants of interest. For the early detection of grapes diseases Patil & Thorat ( 2016 ) developed agricultural monitoring system based on machine learning and relative humidity sensors, temperature sensors and also the leaf wetness sensors. The data collected at regular intervals are sent using Zig-bee module to the server.. The server here employs the hidden Markov model algorithm towards training the data sets pertaining to temperature, relative humidity and leaf wetness for analysing the data towards predicting the chance of disease on grapes prior infection. This information is sent as an alert message via SMS to the farmer and expert. The system employs machine learning in early and accurate detection of disease in grapes and suggests pesticide to protect the crop from disease and reduce manual disease detection efforts which can be helpful for farmers providing information for scheduling fertilizers, pesticide spraying, irrigation etc further,improving the quality and quantity of grapes. The main drawback of this system is its inability to handle the system by farmers because of their lack of knowledge in the technology and the literacy problems. Because hyperspectral remote sensing captures comprehensive reflectance data throughout a wide spectral range, it has proven to be an effective technique for early plant disease identification. It was shown to be effective in evaluating Okra Yellow Vein Mosaic Virus (OYVMD) infection by Mishra et al., ( 2023 ), where they used variety of vegetative indices to categorize the severity of the disease. Chlorophyll absorption-related notable spectrum fluctuations around the 650 nm wavelength were identified as important markers of plant stress. Principal Component Analysis (PCA), which was primarily influenced by indices such as NDVI and RVI, stated 75% of the spectral variation in the study. K-means clustering was able to distinguish between healthy and moderately diseased plants, demonstrating its reliability in pattern recognition. The potential of hyperspectral remote sensing to improve disease classification accuracy and advance precision agriculture techniques is demonstrated by this integrated strategy. Kushwaha et al. ( 2022 ) demonstrated a robust approach of non-destructive estimation of chlorophyll-a and -b content across various crops using airborne hyperspectral data from AVIRIS-NG. They employ second derivative reflectance analysis, identifying wavelengths at 672 nm and 587 nm as most sensitive to Chl-a and Chl-b, respectively. They developed linear models within a bootstrapped resampling framework, achieving moderate R² values (0.46 for Chl-a and 0.51 for Chl-b during calibration, dropping to 0.3 in validation). Kumar et al. ( 2013 ) demonstrated the potential of hyper-spectral remote sensing for assessing aphid infestation in mustard crops. Conducted over two growing seasons, their research compares healthy and aphid-infested canopies, revealing significant reductions in leaf area index (67–94%) and chlorophyll concentration (up to 50%) due to infestation. Along with such physiological changes, reductions in oil content (13–16% lower in infested plants) and seed yield, underscore the severity of aphid damage. Spectral analysis of healthy plants showed stronger reflectance peaks around 550–560 nm and higher near-infrared reflectance, while infested plants exhibit diminished signals, particularly between 1950–2450 nm — a range identified for distinguishing infestation levels. They also correlated spectral indices (NDVI, RVI, AI, SIPI) with aphid damage which may serve as non-invasive tool for early detection and monitoring of crop health for on time pest management. Given the essential role of legumes in food security and the rising demand-supply gap, particularly in South Asia, the study by Jeong and Na ( 2024 ) presented an advanced approach to Faba bean leaf disease identification using deep Convolutional Neural Networks (CNNs). They used the model structured with eight convolutional layers, four max-pooling layers, and three dropout layers to prevent overfitting, achieving impressive performance — 99.37% accuracy during training and 89.69% during validation. The CNN demonstrated high precision, recall, and F1 scores across multiple disease categories: Chocolate Spot, Faba Bean Gall, Rust, and Healthy leaves, with an overall accuracy of 91%. Notably, Chocolate Spot and Rust showed the highest classification performance. They further propose future enhancements like incorporating hybrid machine learning models (e.g., Support Vector Machines, decision trees) and expanding datasets to include stem and root infections for more comprehensive disease detection. The application of IoT and machine learning in agriculture, namely for precision farming and crop disease monitoring, has been the subject of many research investigations. Among these, the study by Nagasubramanian et al. ( 2021 ) presents an all-inclusive system called ECPRC (Ensemble Classification and Pattern Recognition for Crops), which uses intelligent classification models and Internet of Things-based sensing to detect plant diseases in real time. The system combines spectral cameras that take high-resolution pictures of leaves with a variety of environmental sensors that track temperature, soil moisture, and light intensity. An Arduino-based hardware module is used to process these data, and convolutional neural networks (CNN) and ensemble nonlinear support vector machines (ENSVM), two sophisticated machine learning approaches, are used for analysis. CNN succeeds at image-based recognition via deep-layer filtering and pattern learning, while ENSVM manages complicated disease patterns using multiple sub-classifiers. Pre-processing techniques such as image segmentation, noise reduction, and enhancement further improve detection accuracy by isolating disease symptoms like spots and discolorations. as stated in the study, CNN performs better in terms of accuracy, recall, and precision than conventional classifiers like Naïve Bayes, K-Nearest Neighbors, and regular SVM. In the meantime, ENSVM is perfect for lightweight IoT devices because it exhibits effective performance with reduced computing demands. The technology offers insights for managing irrigation and nutrients, facilitates real-time environmental monitoring, and detects diseases. With its mobile accessibility and cloud storage, ECPRC provides farmers with a useful tool for decision support. This study promotes precise, scalable, and real-time crop health monitoring solutions, highlighting the growing importance of hybrid IoT-ML/DL systems in present-day agriculture. 2.5 Precision weed control Weeds are undesirable plants that invade farmland, competing with crops for essential resources like nutrients, sunlight, and space. If left unchecked, they can hinder crop development, leading to lower yields and, ultimately, financial losses for farmers by increasing input costs, (Marco et al., 2021 ). Moreover, compared to crop diseases (25%) and insect pests (20%), weeds account for almost 45% of field crop output losses, making them the most expensive category of agricultural risks (Gnanavel, 2015 ). Weeds which are not removed can cause a yield loss of 100%, (Chauhan, 2020 ). Furthermore, weeds interact with other organisms [insects and pathogens (bacteria and fungi)] in the ecosystem and serves as host for spreading various diseases which can seriously harm crop plants (Esposito et al., 2021 ). Additionally, weeds degrade the value of land, particularly perennial and parasitic weeds, and slow down water management by reducing water flow in irrigation ditches and increasing evapotranspiration losses according to Monteiro and Santos ( 2022 ). Hence effective weed control is demand of time which not only promotes healthier plants but also improves resource efficiency by allowing crops to absorb more water, nutrients, and sunlight. Amongst different weed management methods, farmers readily relies on herbicides and weedicide which again can accumulate in food chain contributing to more environment pollution. Other physical, mechanical, cultural and methods includes hand-weeding, mulching, soil solarisation and weeding equipment. On the other side integrating precision weed management practices aids in managing weeds early which prevents them from producing seeds or developing deeper root systems. In, precision weed management, monitoring and detection of weed is first step followed by spot spraying or robotic weed removal. Different methods of monitoring and detection includes, Remote Sensing and UAVs, Robotic Technology and Sensor-Based detection. With this context, Partel et al. ( 2019 ) developed low cost, AI-driven smart sprayer prototype which comprised of a machine vision software (AI-based) that uses deep learning to detect specific target weeds along with hardware setup with 12 individual fast-response nozzles for spraying. The sprayer was evaluated by using both artificial and real plants wherein, two GPU united were tested [(i) NVIDIA TX2 GPU, and (ii) NVIDIA GTX 1070 Ti GPU]. The GTX 1070 Ti GPU outperformed the TX2 GPU. In comparison to the TX2 GPU, the GTX 1070 Ti GPU reduced the number of missed targets (Portulaca weed) by 81% (from 43% to 8%), improved the detection system's precision and recall by 20% and 77%, and increased the spraying system's precision and recall by 10% and 59%, respectively. This improvement is explained by the fact that difficult scenarios, where targets are highly similar to non-targets and require more processing power, necessitate a more powerful neural network. This smart sprayer only sprayed on the target weed, Portulaca, after correctly differentiating it from non-target sedge weeds and pepper plants. For further precision of weed detection on field, a weed map was developed using an RTK GPS and smart sprayer, generated map showed a good visual correlation to the real scenario’s picture. Additionally, they suggested using a sensor fusion strategy (like the Kalman filter) to eliminate noise and improve the accuracy of weed recognition. Another study was conducted by Mateen and Zhu ( 2019 ) where they exploited UAV technology in order to detect weed for its precise identification within the crop rows as well as between the crop rows in wheat field. The main obstacle to weed patch identification in real agriculture fields is the spectral similarity between weed and crop patches, to solve false identification, extract weed properties and identifing weed patterns, they employed object-based image analysis (OBIA) alongside with color-based thresholding and machine learning techniques. Additionally, they used ortho-mapping to improve image standards. RGB-equipped UAVs were utilized for field image capturing, followed by pre-processing (noise reduction, color enhancement, and segmentation), and feature extraction based on color, texture, and shape. In order to categorize the pixels as either crop or weed, machine learning algorithms were also trained and evaluated. The performance of the classification model was evaluated using metrics such as accuracy, precision, recall, and F1-score. The findings demonstrate that UAV-based weed detection provides a scalable and effective precision agriculture solution, facilitating targeted herbicide delivery and encouraging environmentally friendly agricultural methods. 2.6 Harvesting and yield optimization Recent advances in deep learning have significantly increased fruit detection accuracy, enabling trustworthy identification even in complex and difficult environments. However, there have been difficulties in developing 3D fruit localization, which is essential for robotic harvesting (Thakur et al., 2019 ). The efficiency of existing sensors in precisely finding fruits in natural orchard situations has been hampered by issues such complex fruit shapes, various orientations, clustering, variable light levels, and foliage obstacles (Moreno et al. , 2023). However, by combining the use of laser scanning with depth cameras, the Active Laser-Camera Scanning (ALACS) system overcomes such challenges (Jain et al., 2023 ). ALACS has proven to be more accurate at accurately identifying apples in orchards. When ALACS was tested with a harvesting robot that included a vacuum-based end effector, it was able to harvest fruits with excellent efficiency. (Sharma and Shivandu, 2024 ). For modern agriculture, it is essential to understand the spatial distribution of fruits inside tree canopies. Such information allows for more than just fruit counting; it allows for accurate predictions regarding when to harvest, which in turn supports accurate scheduling and efficient resource allocation (Sassu et al. ,2024). Robots designed for activities like picking fruit depend heavily on spatial data (Kolhalkar et al., 2022 ). These robots may mitigate fruit damage and manage orchards more effectively when they have accurate information about where the fruit is located (Sharma et al. , 2024). Conventional yield prediction methods have depended on statistical models that employ basic weather variables and historical yield data. Although these models offer a fundamental comprehension, the intricacy of agricultural ecosystems often leads them to be misleading. While in Morden approach, machine learning techniques have been incorporated in recent developments to improve yield prediction accuracy. Numerous machine learning algorithms, including random forests, support vector machines (SVM), linear regression, and deep learning models, have been used in studies to examine a variety of variables, such as crop management strategies, soil properties, and weather parameters (Sree et al., 2025 ). Predictive models could precisely forecast agricultural yields for the next growing season by utilizing data on soil health, climate patterns, weather forecasts, soil conditions, and past crop yields (Sharma et al. , 2024). The integration of GPS and sensor technologies into combine harvesters is used to enhance efficiency and reduce grain loss during harvesting, with this perspective, Astanakulov et al. ( 2021 ) used a Dominator-130 combine harvester fitted with a GPS receiver (Teltonika FMB920) and a grain level sensor (Escort DB-2). Incorporation of Grain level sensor and GPS receiver indicated how much grain was being uploaded into the grain tank and at what velocity the combine harvester was operating at in real time. They noticed that the work efficiency and grain loss of the combine were greatly impacted by higher grain yield, higher weed density, and lower moisture content. Grain loss increased and harvester efficiency decreased as yield increased. Similarly, efficiency decreased and grain loss increased in fields with significant weed infestation. In order to reduce losses and increase harvesting productivity, the study emphasizes the importance of adjusting combine settings in real-time field conditions, such as crop yield, weed levels, and grain moisture. Similarly, Astanakulov et al. ( 2021 ) they also monitored effects of combine harvester with smart devices in soybean harvesting. One more similar study of grain harvester combine Dominator130 was carried out by Ochildiev et al. ( 2021 ) by focusing on enhancing the performance of the Dominator-130 grain harvester for sunflower harvesting by integrating a GPS receiver and grain level indicator sensor. Their objective was to reduce seed loss and improve operational efficiency through precision-based modifications. A specially adapted header was developed, featuring a segmented cutting unit, auger, and custom-designed divider-guides. The researchers tested various lengths and gaps of the divider-guides and found that improper dimensions (either too short or too long) led to significant seed loss due to poor alignment or interference with sunflower stems. The optimized configuration ensured more accurate cuts and effective collection within basket. Additionally, the study examined how combine speed affects performance, observing that while faster speeds increased field capacity, they also resulted in higher seed losses, especially in the header and thresher. Through real-time data collection enabled by GPS and sensors, the team was able to fine-tune the machine settings under actual field conditions. This study demonstrates how precision agriculture tools and simple mechanical adaptations can significantly improve the harvesting efficiency of existing machinery for non-traditional crops like sunflower. Traditional farming relies heavily on experience and labor. One of the main issues facing agriculture today is a shortage of labor. To address the increasing shortage of agricultural labor, Chen et al. ( 2023 ) developed an AIoT-based Autonomous Mobile Robot (AMR) system specifically for efficient dragon fruit (pitaya) harvesting. This system integrates artificial intelligence with Internet of Things (IoT) technologies, featuring edge computing powered by the NVIDIA Jetson Nano and utilizing 2D Simultaneous Localization and Mapping (SLAM) for autonomous navigation in dynamic orchard environments. The robot is equipped with a camera, LiDAR sensor, and a nine-axis gyroscope, enabling it to collect real-time data for mapping, obstacle detection, and accurate fruit identification.A notable component of the system is its AI object detection module, which employs the YOLOv3-Tiny model to distinguish between ripe and unripe pitayas with an impressive 96.7% accuracy. Despite a relatively small dataset, the system's performance was significantly improved through data augmentation and transfer learning techniques. During experimental testing, the robot achieved a 97% harvest success rate, efficiently harvesting up to 30 fruits within an hour. The integration of the Dijkstra algorithm for route optimization further enhanced navigation efficiency. With its modular architecture, low-cost components, and software based on ROS 2 and Ubuntu 18.04, the system is well-suited for real-world deployment in outdoor orchard environments. A customized deep learning model called MangoYOLO, a mix of YOLOv2 (small) and YOLOv3, was presented by Koirala et al. ( 2019 ). It is designed and created for the speed and accuracy of real-time mango fruit detection and orchard yield estimation. With an F1 score of 0.968 and Average Precision (AP) of 0.983, MangoYOLO outperforms other object detection architectures such as Faster R-CNN, SSD, and YOLO versions (v2, v3, and small), detecting mangoes in just 8 ms per tile. Training and testing involved more than 1,500 photos from five orchards, and it performed consistently across various lighting conditions, camera types, orchards, and cultivars. Using a correction factor for occluded fruits, MangoYOLO was able to estimate orchard fruit load within 4.6% to 15.2% of real packhouse counts, offering a useful, affordable, in-field fruit detection approach that is essential for precision agriculture and yield forecasting. Kolhalkar et al. ( 2022 ) developed an IoT-enabled mechatronic harvesting module for real-time yield monitoring of grape harvests in vineyards. The system was successfully tested at NRCG, Pune, and integrates a robotic arm equipped with a gripper-cutter mechanism and a strain-gauge-based weighing module to measure the weight of each fruit cluster as it's harvested. Data is transmitted and stored using Wi-Fi and cloud servers, making it accessible remotely. It minimize fruit damage by gripping the stem rather than the fruit surface, thereby preserving the natural waxy layer, which contributes to extended shelf life and improved post-harvest quality. It also significantly reduces labour dependence and associated costs while providing real-time yield data through IoT integration. n addition to field testing, the system was evaluated under laboratory conditions, where various greenhouse vegetables such as bell peppers, long chili peppers, bitter melons, guavas, eggplants, and okra were overhung at natural heights and harvested using the module. However, currently module lacks the capability to detect fruit ripeness, and its application is restricted when dealing with large-sized fruits like watermelon or papaya but these limitations can be overcome by improving the module. With the goal to enhance precision agriculture, Talaat ( 2023 ) created a revolutionary Crop Yield Prediction Algorithm (CYPA) that combines machine learning and Internet of Things (IoT) technology. By analyzing a variety of agricultural data, such as yield-specific, meteorological, environmental, and chemical factors, the CYPA system is intended to predict crop yields with high accuracy. The model obtained excellent predicted accuracy—up to 99.33%—by combining real-time data collecting via IoT devices with sophisticated regression models like DecisionTreeRegressor, RandomForestRegressor, and ExtraTreeRegressor. They took advantage of a noteworthy innovation called "active learning," which increases model efficiency by choosing the most informative data samples for training, minimizing the need for massive amounts of labeled data, and allowing the model to dynamically adjust to shifting field conditions like pest outbreaks or weather variations. The algorithm was trained and validated using Global agricultural datasets for important crops like wheat, rice, maize, potatoes, soybeans. Furthermore, the study also used statistical methods like multiple regression analysis and Pearson's correlation coefficient to assess how significantly variables like temperature, rainfall, and pesticide use influence crop productivity. The average temperature and yield were shown to be negatively correlated, but rainfall and pesticide use were found to be weakly positively correlated. As a result, CYPA stands apart in its ability to facilitate data-driven agricultural decision-making, providing farmers and policymakers with insightful information for more effective resource management and enhancing food security in the face of climate change. Furthermore, it performed superior in terms of accuracy and usefulness than current benchmark models, such as RNN-based techniques. 2.7 Supply Chain and logistics The agricultural supply chain is composed of four major components: producers (farmers), distributors, retailers, and consumers. Efficient coordination among these stakeholders is vital for ensuring the smooth flow of agricultural goods from farms to markets. Recent advancements in artificial intelligence (AI), particularly in machine learning (ML), have introduced transformative solutions across various stages of the supply chain, enhancing productivity, reducing waste, and increasing profitability. One of the earliest and most impactful applications of ML in agricultural supply chains is demand forecasting. By leveraging historical sales data, market trends, and external variables such as weather patterns and economic indicators, machine learning algorithms can predict future consumer demand with a high degree of accuracy. Compared to traditional forecasting methods, these algorithms provide superior precision, enabling supply chain actors to optimize inventory management, adjust production schedules, and fine-tune distribution strategies. Such foresight helps to minimize both stockouts and overstock situations, improving operational efficiency and customer satisfaction (Saxena et al., 2023 ). As the supply chain progresses from production to distribution, route optimization becomes a critical factor. ML algorithms can analyze real-time data on traffic conditions, weather disruptions, and road infrastructure to identify the most efficient transportation routes. This dynamic decision-making capability results in reduced fuel consumption, lower transportation costs, and faster delivery times, thereby enhancing service reliability and customer experience (Saxena et al., 2023 ). Risk management is another crucial area of supply chain management where machine learning is being used. The agricultural supply chain is susceptible to a wide range of risks, including geopolitical tensions, climate variability, market volatility, and supplier disruptions. ML models can process and synthesize data from diverse sources to detect potential risks early. These predictive insights allow supply chain managers to implement proactive mitigation strategies and develop contingency plans, ensuring resilience and continuity in operations (Fan et al., 2021 ). At the production level, price prediction and market analysis are crucial for farmers who must navigate volatile agricultural markets. Prices can fluctuate dramatically due to imbalances in supply and demand, weather events, and shifts in global trade policies. By analyzing historical pricing data, current market dynamics, and projected returns, ML models provide actionable insights that empower farmers to make strategic decisions. For example, AI tools can recommend delaying a harvest or temporarily storing produce during periods of anticipated oversupply, thereby helping farmers avoid low prices and potential financial losses (Hassan et al., 2022 ). Beyond individual farm profitability, accurate price forecasting plays a broader socio-economic role. Agricultural price instability can affect national economic performance and trigger public concern, especially in countries with large farming populations. Unpredictable fluctuations in the prices of staple crops and food products can lead to reduced farmer income, consumer distress, and, in severe cases, social unrest. In this context, the ability of AI and ML technologies to forecast price trends contributes not only to market efficiency but also to national food security and economic stability. This is particularly vital in major agricultural economics bearing countries such as China and india where price stability supports both rural livelihoods and macroeconomic balance (Wang, 2023 ; Assimakopoulos et al., 2024 ). In the context of price forecasting, Paul et al. ( 2022 ) explored the comparative performance of machine learning (ML) algorithms in forecasting daily wholesale prices of brinjal across seventeen major markets in Odisha. Given the highly volatile and seasonal nature of vegetable prices, traditional statistical models such as the Autoregressive Integrated Moving Average (ARIMA) often fall short in capturing complex, nonlinear dynamics. In this regard, the study evaluates the predictive capabilities of four advanced ML models—Generalized Regression Neural Network (GRNN), Support Vector Regression (SVR), Random Forest (RF), and Gradient Boosting Machine (GBM)—against the ARIMA model. The analysis employed daily price data from January 2015 to May 2021, obtained from the AGMARKNET portal, and assessed model performance using statistical metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Error (ME), and Mean Absolute Percentage Error (MAPE). Additionally, the Diebold-Mariano test and Model Confidence Set (MCS) approach were utilized to statistically compare forecast accuracies. Results revealed that GRNN consistently outperformed other models in the majority of the markets, demonstrating superior predictive accuracy. In a few markets, RF exhibited comparable performance to GRNN, while SVR, GBM, and ARIMA models showed relatively lower accuracy across most cases.The findings emphasize the efficacy of machine learning techniques, particularly GRNN, in modeling complex time series data in the agricultural domain. These models offer robust tools for farmers, market planners, and policymakers to make informed decisions regarding market timing and location, thus optimizing profit margins and reducing market risks. Another study conducted by Jaiswal et al. ( 2022 ) wherein they presented a hybrid forecasting model combining Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) networks to predict the prices of key agricultural commodities such as wheat, maize, and other crops. The RNN-LSTM architecture leverages the temporal dependencies in time series data, which is particularly important in agriculture where prices are influenced by seasonal patterns, weather events, and cyclical market behaviour. The study highlights the effectiveness of such deep learning models in regions characterized by high seasonal variability, where price fluctuations are more pronounced and unpredictable. By incorporating historical trends, seasonal cycles, and potentially exogenous variables (e.g., rainfall, temperature, policy changes), the hybrid RNN-LSTM model can provide granular insights into future market conditions. These predictive capabilities empower farmers with timely information, enabling them to make more informed marketing and production decisions—such as when to harvest, where to sell, or whether to store produce for a better price. Another logistic related study was carried out by Zhang et al. ( 2020 ) wherein they presented a robust machine learning-based framework for evaluating logistic risks in the transportation of perishable agricultural commodities. Focusing on fresh produce, with strawberries as a case study, they built comprehensive risk assessment system grounded in five principal dimensions namly, technological, biological, sustainability, environmental, and emergency risks. A total of 30 quantitative and qualitative indicators are identified to capture the multidimensional nature of these risk factors. To address the complexities of high-dimensional, small-sample data typically encountered in fresh produce logistics, they employed a Support Vector Machine (SVM) algorithm due to its strong generalization capabilities and resilience against overfitting. The model utilizes historical monitoring data from cold chain logistics, integrating variables such as temperature, humidity, vibration, gas composition, packaging integrity, and emergency response capacity. Following normalization and parameter optimization using a radial basis function (RBF) kernel, the model is trained and validated on 350 sample datasets, achieving a predictive accuracy between 85% and 97.5%.The findings emphasise the efficacy of the SVM-based model in real-time risk classification and early warning. The proposed methodology offers significant practical implications for logistics managers, enabling data-driven decision-making to mitigate spoilage, enhance food safety, reduce economic losses, and support sustainability targets in agri-food supply chains. One significant challenge in supply chain is monitoring and checking of storage facilities. In that matter, Afreen and Bajwa ( 2021 ) proposed a cost-effective, intelligent system RT-IMNS for monitoring and managing cold storage environments where perishable fruits and vegetables are kept using Internet of Things (IoT) and Artificial Neural Networks (ANN). The system overcomes the drawbacks of conventional systems that just monitor temperature and humidity by constantly measuring temperature, humidity, CO₂ concentration, and light intensity. The ANN model classifies commodity status into three categories: good, unsatisfactory, and alarming. It achieves 99% prediction accuracy, outperforming other models like SVM, Naïve Bayes, Random Forest, and Decision Trees. Real-time data is collected through multiple sensor nodes and sent to a cloud-based Firebase database. The system also integrates an Android application for remote monitoring and instant notification, enabling timely actions to prevent spoilage allowing personnel to remotely monitor environmental parameters and receive real-time alerts when any parameter exceeds critical thresholds. This feature reduces the need for constant physical presence in cold storage facilities, and supports better decision-making in managing storage conditions. The system was validated using a real-world experimental setup involving cold storage of potatoes, with over 5,361 environmental data instances collected over 15 days. This low-cost, user-friendly solution can be valuable for small- and medium-sized enterprises, contributing significantly to reducing food losses in the storage and post-harvest lifecycle. Perishable goods such as fruits, vegetables, dairy, and fermented products are highly sensitive to environmental conditions and logistical handling, making them prone to spoilage, contamination, and economic losses. Under this circumstances, IoT-based environmental sensing, and machine learning can aid in real-time monitoring, direction detection, and data-driven decision-making, thereby reducing waste, ensuring food safety, and strengthening consumer confidence across the entire value chain. In this direction, Alfian et al. ( 2020 ) presented a smart, integrated traceability system for the perishable food supply chain (PFSC), combining RFID technology, IoT sensors, and a machine learning-based direction detection model (XGBoost). A key innovation is integrating machine learning classifier (XGBoost) into RFID gates to identify the directional movement of tagged products, whether they are entering (receiving) or leaving (shipping) cold storage, addressing a major limitation in conventional RFID systems, which cannot inherently distinguish movement direction. The model uses features extracted from received signal strength (RSS) and timestamp data, and achieves superior performance compared to traditional models with accuracy. The solution was validated in a real-world kimchi supply chain, demonstrating its capability to monitor and present complete product traceability covering movement, location, and environmental conditions across producers, transporters, and distributors. The system effectively filters out false-positive tag readings (e.g., from static or turn-back tags), improving inventory accurac y and decision-making reliability. To increase agricultural productivity and sustainability, it is becoming more and more important to integrate artificial intelligence (AI), the Internet of Things (IoT), and remote sensing (RS). Araújo et al. ( 2023 ) carried out a thorough investigation of current machine learning (ML) applications in agriculture within this framework. In order to facilitate the transition from traditional to data-driven agricultural systems, their studies show how machine learning (ML) may optimize crop management, water use, soil monitoring, and livestock oversight. They noted that the research has been conducted on wide range of crops from several different agricultural fields while industrial and plantation crops, cereal crops, vegetable crops, perennial trees, orchards, and vineyards, as well as fruit-bearing trees. The research carried out was mainly associated with crop quality, crop mapping, yield, crop disease, pest/weed detection etc. , (Fig. 5). Challenges and Future Prospects Although the Internet of Things (IoT) and artificial intelligence (AI) hold great potential for improving precision agriculture, a number of significant challenges need to be addressed before they can be widely utilized. Such as privacy and security issues with the collecting, exchanging, and archiving of sensitive agricultural data (Farooq et al., 2019 ). Smooth integration of AI platforms and IoT devices may be complicated by a lack of standardization and integration between them. Enabling data exchange and collaboration requires the development of common standards and protocols IoT devices and AI technologies can be expensive, which can hinder adoption, especially for smallholder farmers in developing nations. To guarantee that the advantages of precision farming are widely distributed, efforts must be made to lower expenses and improve accessibility (Van Klompenburg et al., 2020 ). Declarations Conflicts of Interest: The authors declare no conflict of interest. Author Contribution Dhara Prajapati gathered and reviewed the relevant literature, organized the data, and helped shape the initial draft. Abishek Murugesan interpreted the findings, and refined the scientific content. Rucha Dave provided guidance on the research framework, verified the technical accuracy, and reviewed the final version of the paper. References Abioye, E. A., Abidin, M. S. Z., Mahmud, M. S. A., Buyamin, S., Ishak, M. H. I., Abd Rahman, M. K. I., & Ramli, M. S. A. (2020). A review on monitoring and advanced control strategies for precision irrigation. Computers and Electronics in Agriculture , 173 , 105441. Afreen, H., & Bajwa, I. S. (2021). An IoT-Based Real-Time Intelligent Monitoring and Notification System of Cold Storage. Ieee Access : Practical Innovations, Open Solutions , 9 , 38236–38253. Agrawal, H., Dubey, A., Tiwari, N., Pandey, D. K., Putrevu, D., Misra, A., & Kumar, R. (2019). Performance evaluation of a newly in-house developed in-situ soil moisture sensor with standard industrial sensors and gravimetric sampling. Journal of Geomatics , 13 (2), 280–284. Al Mamun, M. R., Ahmed, A. K., Upoma, S. M., Haque, M. M., & Ashik-E-Rabbani, M. (2025). IoT-enabled solar-powered smart irrigation for precision agriculture. Smart Agricultural Technology , 10 , 100773. Alfian, G., Syafrudin, M., Farooq, U., Ma’arif, M. R., Syaekhoni, M. A., Fitriyani, N. L., Lee, J., & Rhee, J. (2020). Improving efficiency of RFID-based traceability system for perishable food by utilizing IoT sensors and machine learning model. Food Control , 110 , 107016. Araújo, S. O., Peres, R. S., Ramalho, J. C., Lidon, F., & Barata, J. (2023). Machine learning applications in agriculture: current trends, challenges, and future perspectives. Agronomy , 13 (12), 2976. Assimakopoulos, F., Vassilakis, C., Margaris, D., Kotis, K., & Spiliotopoulos, D. (2024). Artificial intelligence tools for the agriculture value chain: Status and prospects. Electronics , 13 (22), 4362. Astanakulov, K., Abdillaev, T., Umirov, A., Fozilov, G., & Hatamov, B. (2021). Monitoring of the combine with smart devices in soybean harvesting. In E3S Web of Conferences (Vol. 227, p. 07003). EDP Sciences. Astanakulov, K., Shovazov, K., Borotov, A., Turdibekov, A., & Ibrokhimov, S. (2021). Wheat harvesting by combine with GPS receiver and grain sensor. In E3S Web of Conferences (Vol. 227, p. 07001). EDP Sciences. Bellosta-Diest, A., Campo-Bescós, M. Á., Zapatería-Miranda, J., Casalí, J., & Arregui, L. M. (2022). Evaluation of nitrate soil probes for a more sustainable agriculture. Sensors (Basel, Switzerland) , 22 (23), 9288. Bitella, G., Rossi, R., Bochicchio, R., Perniola, M., & Amato, M. (2014). A novel low-cost open-hardware platform for monitoring soil water content and multiple soil-air-vegetation parameters. Mdpi Sensors J , 14 , 19639–19659. https://doi.org/10.3390/s141019639 (Crossref.). Bousbih, S., Zribi, M., Pelletier, C., Gorrab, A., Lili-Chabaane, Z., Baghdadi, N., Ben Aissa, N., & Mougenot, B. (2019). Soil texture estimation using radar and optical data from Sentinel-1 and Sentinel-2. Remote Sens , 11 , 1520. Capraro, F., Tosetti, S., Rossomando, F., & Mut, V. (2018). Web-based system for the remote monitoring and management of precision irrigation: a case study in. Sensors MDPI . https://doi.org/10.3390/s18113847 (Crossref.). Chauhan, B. S. (2020). Grand Challenges in Weed Management. In Frontiers in Agronomy; Frontiers Media SA: Lausanne, Switzerland, ; Volume 1. Chen, L. B., Huang, X. R., & Chen, W. H. (2023). Design and Implementation of an Artificial Intelligence of Things-Based Autonomous Mobile Robot System for Pitaya Harvesting. Ieee Sensors Journal , 23 , 13220–13235. Dave, R., Kumar, G., Pandey, D. K., Khan, A., & Bhattacharya, B. (2019). : Evaluation of Modified Dubois Model for estimating surface soil moisture using dual polarization RISAT-1 C-band SAR data, Geo-carto International, 10.1080/10106049.2019.1655801 Dave, R. B., Saha, K., Kushwaha, A., Vithalpura, M., Nidhin, P., & Murugesan, A. (2023). Analysing the potential of polarimetric decomposition parameters of Sentinel–1 dual-pol SAR data for estimation of rice crop biophysical parameters. Journal of Agrometeorology , 25 (1), 105–112. Dhiman, G., Bhattacharya, J., & Roy, S. (2023). Soil textures and nutrients estimation using remote sensing data in north india-Punjab region. Procedia computer science , 218 , 2041–2048. Esposito, M., Crimaldi, M., & Cirillo, V. (2021). Drone and sensor technology for sustainable weed management: A review. Chem Biol Technol Agric , 8 , 18. Fan, X., Nan, Z., Ma, Y., Zhang, Y., & Han, F. (2021). Research on the Spatio-Temporal Impacts of Environmental Factors on the Fresh Agricultural Product Supply Chain and the Spatial Differentiation Issue—An Empirical Research on 31 Chinese Provinces. International Journal Of Environmental Research And Public Health , 18 , 12141. Farooq, M. S., Riaz, S., Abid, A., Abid, K., & Naeem, M. A. (2019). A Survey on the Role of IoT in Agriculture for the Implementation of Smart Farming. Ieee Access , 7 , 156237–156271. Food and Agricultural Organisation (2009). Expert meeting on How to Feed the World in 2050. https://www.fao.org/4/ak542e/ak542e00.htm Gnanavel, I. (2015). Eco-Friendly Weed Control Options for Sustainable Agriculture. Sci Int , 3 , 37–47. Harishankar, S., Kumar, R. S., Sudharsan, K., Vignesh, U., & Viveknath, T. (2014). Solar powered smart irrigation system. Adv Electr Comput Eng , 4 , 341–346. Hassan, M., Malhotra, K., & Firdaus, M. (2022). Application of Artificial Intelligence in IoT Security for Crop Yield Prediction. Res Rev Sci Technol , 2 , 136–157. Hou, M., Tian, F., Ortega-Farias, S., Riveros-Burgos, C., Zhang, T., & Lin, A. (2021). Estimation of crop transpiration and its scale effect based on ground and UAV thermal infrared remote sensing images. European Journal of Agronomy , 131 , 126389. Implementation of Automatic Irrigation Control System IOSR J Comput Eng (IOSR- JCE) 17 (4), 99–111. https://doi.org/10.9790/0661-174299111 . (Crossref.). Jain, S., Bhujel, S., Shrivastava, U., Mohapatra, A., & Mishra, G. (2023). Advancements in drone technology for fruit crop management: a comprehensive review. Int J Envir Clim Change , 13 (11), 4367–4378. Jaiswal, R., Jha, G. K., Kumar, R. R., & Choudhary, K. (2022). Deep Long Short-Term Memory Based Model for Agricultural Price Forecasting. Neural Computing And Applications , 34 , 4661–4676. Jeong, H. Y., & Na, I. S. (2024). Efficient faba bean leaf disease identification through smart detection using deep convolutional neural networks. Legume Research , 47 (8), 1404–1411. Khanal, S., Kc, K., Fulton, J. P., Shearer, S., & Ozkan, E. (2020). Remote sensing in agriculture—accomplishments, limitations, and opportunities. Remote Sensing , 12 (22), 3783. Koirala, A., Walsh, K. B., Wang, Z., & McCarthy, C. (2019). Deep learning for real-time fruit detection and orchard fruit load estimation: benchmarking of ‘MangoYOLO’. Precis Agric , 20 (6), 1107–1135. Kolhalkar, N. R., Krishnan, V. L., Somkuwar, R. G., & Pandit, A. A. (2022). Design of IoT-based mechatronic module for real-time yield monitoring during harvesting of greenhouse produce and vineyards, Int. Rev. Mech. Eng. 16 (2022) 48–55. Kulkarni, S. S., Bajwa, S. G., & Huitink, G. (2010). Investigation of the effects of soil compaction in cotton. Am Soc Agric Biol Eng , 53 , 667–674. Kumar, J., Vashisth, A., Sehgal, V. K., & Gupta, V. K. (2013). Assessment of aphid infestation in mustard by hyperspectral remote sensing. Journal of the Indian Society of Remote Sensing , 41 , 83–90. Kushwaha, A., Das, A., Dave, R., & Bhattacharya, B. K. (2022). A non-destructive estimation of chlorophyll-a and-b over different crops using airborne imaging spectroscopy observations. Advances in Space Research , 73 (2), 1290–1303. Marco, E., Mariano, C., Valerio, C., Fabrizio, S., & Albino, M. (2021). Drone and sensor technology for sustainable weed management: a review. Chem Biol Technol in Agricult , 8 , 18. https://doi.org/10.1186/s40538-021-00217-8 Mateen, A., & Zhu, Q. (2019). Weed detection in wheat crop using UAV for precision agriculture. Pakistan Journal of Agricultural Sciences , 56 (3), 775–784. Mishra, S., Nidhin, P., Dave, R., Pandey, P., Kumari, P., & May (2023). Assessment of Okra Yellow Vein Mosaic Virus using Ground Based Hyperspectral Remote Sensing (Poster presentation). ISMPP 42nd annual Conference & National Symposium on Plant Health Management: A Way Forward for Food Safety, Security and Sustainability. Anand Agricultural University, Anand. Monteiro, A., & Santos, S. (2022). Sustainable approach to weed management: The role of precision weed management. Agronomy , 12 (1), 118. Moreno, H., & Andujar, D. (2023). Proximal sensing for geometric characterization of vines: a review of the latest advances. Computers And Electronics In Agriculture , 210 , 107901. Murugesan, A., Dave, R., Kushwaha, A., Pandey, D. K., & Saha, K. (2023). Surface soil moisture estimation in bare agricultural soil using modified Dubois model for Sentinel-1 C-band SAR data. Journal of Agrometeorology , 25 (4), 517–524. Nagasubramanian, G., Sakthivel, R. K., Patan, R., Sankayya, M., Daneshmand, M., & Gandomi, A. H. (2021). Ensemble classification and IoT-based pattern recognition for crop disease monitoring system. IEEE Internet of Things Journal , 8 (16), 12847–12854. Nock, C. A., Vogt, R. J., & Beisner, B. E. (2016). In eLS (Ed.), Functional Traits . John Wiley & Sons, Ltd. 10.1002/9780470015902.a0026282 Oborkhale, L. I., Abioye, A. E., Egonwa, B. O., & Olalekan, T. A. (2015). Design and. Oborkhale, L. I., Abioye, A. E., Egonwa, B. O., & Olalekan, T. A. (2015). Design and Implementation of Automatic Irrigation Control System. IOSR J Comput Eng (IOSR-JCE) , 17 (4), 99–111. https://doi.org/10.9790/0661-174299111 (Crossref.). Ochildiev, O., Fozilov, G., Achildiev, S., Karimov, M., & Ashurov, N. (2021). Indicators of the combine equipped with GPS receiver in sunflower harvesting. In E3S Web of Conferences (Vol. 227, p. 07002). EDP Sciences. Oerke, E. C. (2020). Remote sensing of diseases. Annual review of phytopathology , 58 (1), 225–252. Pandey, D. K., Srivastava, P. K., Dave, R., Setia, R. K., Sinha, R., Sekhar, M., & Kumar, R. (2024). Operational 500 m surface soil moisture product using EOS-04 C-band SAR over Indian agricultural croplands. Current Science (00113891) , 126 (9). Partel, V., Kakarla, S. C., & Ampatzidis, Y. (2019). Development and Evaluation of a Low-Cost and Smart Technology for Precision Weed Management Utilizing Artificial Intelligence. Computers & Electrical Engineering , 157 , 339–350. researchgate.net + 1ar. Patil, S., & Thorat, S. (2016). Early detection of grapes diseases using machine Learning and 10.1109/CCIP.2016.7802887 Paul, R. K., Yeasin, M., Kumar, P., Kumar, P., Balasubramanian, M., Roy, H. S., Paul, A. K., & Gupta, A. (2022). Machine Learning Techniques for Forecasting Agricultural Prices: A Case of Brinjal in Odisha, India. Plos One , 17 , e0270553. Rucha Dave, G., Kumar, D. K., Pandey, A., & Bimal Bhattacharya. (2019). Evaluation of Modified Dubois Model for estimating surface soil moisture using dual polarization RISAT-1 C-band SAR data. Geocarto International . 10.1080/10106049.2019.1655801 Sassu, A., Psiroukis, V., Bettucci, F., Ghiani, L., Fountas, S., & Gambella, F. (2024). Unmanned aerial system plant protection products spraying performance evaluation on a vineyard. Precis Agric , 1–31. https://doi.org/10.1007/s11119-024-101558 Saxena, K., Jakhete, M. D., Kumari, P. L., Jain, M., Mane, A., & Karthik, H. P. (2023). Optimizing Agricultural Supply Chains with Machine Learning Algorithms. Journal of Advanced Zoology,44, S-2, Pp.3146:3156. Sharma, K., & Shivandu, S. K. (2024). Integrating artificial intelligence and internet of things (IoT) for enhanced crop monitoring and management in precision agriculture. Sensors International , 100292. Singh, A., Kumar, S., Barman, Aparna, P. M., Mohammed Hisham, M., Sabarinathan Babu, N., Satheeshkumar, M., & Ramasamy, and Thiruvengadam K (2025). Integration of AI and IoT for Yield Optimization in Precision Farming. Journal of Experimental Agriculture International , 47 (3), 233–241. Singha, C., Swain, K. C., Sahoo, S., & Govind, A. (2023). Prediction of soil nutrients through PLSR and SVMR models by VIs-NIR reflectance spectroscopy. The Egyptian Journal of Remote Sensing and Space Sciences , 26 (4), 901–918. Sree, B. A., Dasari, D. D. B., Subramanyam, D., & Chethan, P. (2025). Precision Agriculture Optimization: Integrating Machine Learning for Crop Yield Prediction and Resource Management. Precision Agriculture Optimization: Integrating Machine Learning for Crop Yield Prediction and Resource Management (January 18, 2025). Talaat, F. M. (2023). Crop yield prediction algorithm (CYPA) in precision agriculture based on IoT techniques and climate changes. Neural Computing and Applications , 35 (23), 17281–17292. Thakur, D., Kumar, Y., Kumar, A., & Singh, P. K. (2019). Applicability of wireless sensor networks in precision agriculture: a review, Wireless Pers. Commun , 107 , 471–512. Van Klompenburg, T., Kassahun, A., & Catal, C. (2020). Crop yield prediction using machine learning: A systematic literature review. Computers and electronics in agriculture , 177 , 105709. Wang, Y. (2023). Agricultural Products Price Prediction Based on Improved RBF Neural Network Model. Appl Artif Intell , 37 , 2204600. Wanyama, J., Soddo, P., Nakawuka, P., Tumutegyereize, P., Bwambale, E., Oluk, I., & Komakech, A. J. (2023). Development of a solar powered smart irrigation control system Kit. Smart Agricultural Technology , 5 , 100273. Zhang, G., Li, G., & Peng, J. (2020). Risk assessment and monitoring of green logistics for fresh produce based on a support vector machine. Sustainability , 12 (18), 7569. Zhou, Z., Majeed, Y., Naranjo, G. D., & Gambacorta, E. M. (2021). Assessment for crop water stress with infrared thermal imagery in precision agriculture: A review and future prospects for deep learning applications. Computers and Electronics in Agriculture , 182 , 106019. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8220871","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Short Report","associatedPublications":[],"authors":[{"id":555426242,"identity":"5c22c7a4-89c7-4b97-aed7-2eafcae9c089","order_by":0,"name":"Dhara Prajapati","email":"","orcid":"","institution":"Anand Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Dhara","middleName":"","lastName":"Prajapati","suffix":""},{"id":555426243,"identity":"366df4a5-b6e4-4098-b332-215b681314b4","order_by":1,"name":"Abishek Murugesan","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA50lEQVRIiWNgGAWjYFACHobDfyps5NjYG4AcAwvitBzgOZNmzM9zAKRFgjgtDLxthxNnzkgA8YjQwt9+9uABiTOHGTfcfH51w48CCaBIdwJeLRJn8hIOGFSkMxvczim72QN0mMSZsxvwW3Mgx+BAwhlrNqCWtBs8QC0GErn4tciff2Nw4GAbM4/BzTNpN/8Qo8XgRo7BwcY2ZwnJGezHbhNli+GNdwmHGc6kGfDz5LDdljGQ4CHoF7nzuYc/M1TY1LexH392880fGzn+9l4C3kcAHgMwSaxyEGB/QIrqUTAKRsEoGEEAABSSTlwaM9TEAAAAAElFTkSuQmCC","orcid":"","institution":"Anand Agricultural University","correspondingAuthor":true,"prefix":"","firstName":"Abishek","middleName":"","lastName":"Murugesan","suffix":""},{"id":555426244,"identity":"7e8cc4b9-f26e-4721-829e-1a94adabbbde","order_by":2,"name":"Rucha Dave","email":"","orcid":"","institution":"Anand Agricultural University","correspondingAuthor":false,"prefix":"","firstName":"Rucha","middleName":"","lastName":"Dave","suffix":""}],"badges":[],"createdAt":"2025-11-27 10:23:10","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8220871/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8220871/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98778036,"identity":"917a8c04-9a05-4f5f-b3e2-650ecd015db7","added_by":"auto","created_at":"2025-12-22 12:28:50","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1406112,"visible":true,"origin":"","legend":"","description":"","filename":"FINALMANUSCRIPT.docx","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/bd45fb692923a37094388b6c.docx"},{"id":98763657,"identity":"2a2e9609-c4c0-484e-bf9e-e074abd8ae8a","added_by":"auto","created_at":"2025-12-22 10:05:31","extension":"json","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":4882,"visible":true,"origin":"","legend":"","description":"","filename":"06ad20d4646848439f10c401ca5c87fc.json","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/a69434b84cd67e1f8ff7d2a5.json"},{"id":98779119,"identity":"56eb7357-6459-4b71-a733-5d8782529b61","added_by":"auto","created_at":"2025-12-22 12:29:59","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":194012,"visible":true,"origin":"","legend":"","description":"","filename":"06ad20d4646848439f10c401ca5c87fc1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/fd05666a7574c871c375d5a0.xml"},{"id":98778416,"identity":"c403f34c-9083-42fe-b7e3-b7aa55bae5bf","added_by":"auto","created_at":"2025-12-22 12:29:14","extension":"gif","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":43,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.gif","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/21c881ed4f508452706172df.gif"},{"id":98779464,"identity":"9e94caa9-b5e7-48d7-999c-94ba4e35d86e","added_by":"auto","created_at":"2025-12-22 12:30:22","extension":"jpeg","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":328844,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/0c3781449c51f05a4828163c.jpeg"},{"id":98763667,"identity":"68c27c9c-714f-45fc-96b8-efa0f7e825ce","added_by":"auto","created_at":"2025-12-22 10:05:31","extension":"png","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":308,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/9023206fef2d0020b7211f57.png"},{"id":98779189,"identity":"6df25c9d-3b32-4546-b44c-4164ba9a802f","added_by":"auto","created_at":"2025-12-22 12:30:03","extension":"png","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":78582,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/d8bf01b20b50999bb4881845.png"},{"id":98780107,"identity":"9100236d-60d3-40bc-b4e9-74f45d337993","added_by":"auto","created_at":"2025-12-22 12:31:03","extension":"png","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":122051,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/3fb4eaf34fd013ad23dff340.png"},{"id":98780176,"identity":"cebf7066-7028-4cb1-af33-031918d1b0e1","added_by":"auto","created_at":"2025-12-22 12:31:08","extension":"png","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":23338,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/b6bcfc158c626b14a45f9e10.png"},{"id":98780405,"identity":"045eadba-e3b4-4c56-b7b4-ebf903d86b2c","added_by":"auto","created_at":"2025-12-22 12:31:18","extension":"xml","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":191669,"visible":true,"origin":"","legend":"","description":"","filename":"06ad20d4646848439f10c401ca5c87fc1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/542eb463f1821173945b0f1f.xml"},{"id":98763668,"identity":"4d67b8fd-c597-4b03-94e7-681abed5be94","added_by":"auto","created_at":"2025-12-22 10:05:31","extension":"html","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":202761,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/1ba7134e000f6e31405d73f0.html"},{"id":98763665,"identity":"b60c8344-4f21-4763-bc65-3beb5fefefe7","added_by":"auto","created_at":"2025-12-22 10:05:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":221075,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e(a) \u003c/strong\u003eSoil Moisture Probe (Stevens Hydra Go), \u003cstrong\u003e(b)\u003c/strong\u003e Interface of Mobile Application of Stevens HydraGo\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/2592b287ccbd9f9eb3f35964.png"},{"id":98777962,"identity":"3af61b5c-7a6d-439d-b67e-479bf8c9f293","added_by":"auto","created_at":"2025-12-22 12:28:43","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":461733,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003e(a) \u003c/strong\u003eSHOOL moisture probe concept Note, \u003cstrong\u003e(b) \u003c/strong\u003eAndroid app for sensor \u003cstrong\u003e(\u003c/strong\u003eAgrawal \u003cem\u003eet al\u003c/em\u003e., 2019)\u003c/p\u003e","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/4584c6a78fc68957dfae6620.jpeg"},{"id":98780182,"identity":"32dc7f5f-77e9-498f-b8fa-61bbcb5f0a81","added_by":"auto","created_at":"2025-12-22 12:31:08","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":115461,"visible":true,"origin":"","legend":"\u003cp\u003eDistribution of application domains in agriculture: crop, water, soil, and animal management.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/4af7d97e3a5e52a3a9c6e9eb.png"},{"id":99790799,"identity":"7fe518ad-2aa8-4467-b629-fed299c76eb4","added_by":"auto","created_at":"2026-01-08 12:58:43","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1946428,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8220871/v1/b48df990-6be7-44e5-837d-f976e2b6bda0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Machine Learning and Remote Sensing for Soil Moisture and Nutrient Estimation: A Systematic Review and Future Research Roadmap","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe current global population of 7.6\u0026nbsp;billion is estimated to increase to 8.6\u0026nbsp;billion by 2030 and 9.8\u0026nbsp;billion by 2050 as per the report from the United Nations (UN) Economic and Social Affairs department (R-The World Population Prospects: The 2017 Revision). It is anticipated that the world's need for cereals both for human consumption and animal feed will surpass 3\u0026nbsp;billion tonnes by 2050, (FAO, 2009). The critical shortages of arable land and water resources, combined with the growing risks of erratic weather, extreme temperatures, and unpredictable rainfall, emphasize the urgent need for precision farming. Additionally, rising costs of fertilizers, pesticides, seeds, fuel, along with agricultural labor force shortage, makes it essential to adopt precision farming techniques that optimize resource use, minimize waste and improve efficiency through automation and data-driven decision-making. Apart from the production constraints, Farmers also struggles with market fluctuations and lack real-time market insights. Such situation demands for a Precision Agriculture (PA) because PA integrates cutting-edge technologies to boost resource efficiency, sustainability and Agricultural productivity. It enables farmers to make informed decisions and adapt to dynamic climatic conditions by utilizing advanced weather forecasting, remote sensing, and AI \u003cb\u003e(\u003c/b\u003eArtificial Intelligence\u003cb\u003e)\u003c/b\u003e driven analytics. The use of Automation, robotics, and intelligent machinery alleviate labor shortages by streamlining operations and increasing production efficiency. Soil analysis, precision nutrient application, and site-specific fertilization contribute to maintaining soil health and promoting sustainable farming practices. Smart irrigation systems, including drip irrigation and IoT-based moisture sensors, optimize water usage, addressing the challenges of water scarcity. Drones, AI-driven pest and disease detection, and precise pesticide application provide early disease and pest management, minimizing unnecessary chemical usage. Real-time monitoring technologies, enable the early detection of crop stress and potential threats. Additionally, data-driven forecasting, digital farm management, and intelligent logistics solutions improve supply chain efficiency and reduce post-harvest losses. Precision farming further contributes to sustainability by decreasing reliance on chemicals, encouraging environmentally friendly agricultural practices, and ensuring compliance with environmental regulations. This review paper focus on how these advancements collectively enhance agricultural productivity, optimize resource use, and promote a more resilient and sustainable food production system for the future.\u003c/p\u003e\n\u003ch3\u003e2. Implementing Precision Agriculture and Acquiring Data Using Remote Sensing, Artificial Intelligence and Machine Learning\u003c/h3\u003e\n \u003cp\u003eIn precision agriculture, satellite data plays a transformative role by enabling data-driven, site-specific management of crops, soil, and other farm resources. This approach helps maximize yields, optimize inputs (like water, fertilizer, and pesticides), and promote sustainable farming practices. Earth Observation (EO) satellite data, accessible through portals plays a vital role in transforming modern agriculture. Some of the major popular Portals are mentioned in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. One of its primary applications is crop monitoring, where multi-temporal satellite imagery is used to assess crop growth, detect early signs of stress, and identify disease or pest outbreaks. Satellite data also supports crop type mapping, helping distinguish between different crops based on their spectral signatures, which is essential for land-use planning, agricultural statistics, and policy development. Another significant application is yield estimation, where vegetation indices like NDVI, combined with machine learning models, can predict crop productivity, aiding food security assessments and market forecasting. Soil monitoring using radar and thermal data allows for the evaluation of moisture levels, texture, and nutrient content, leading to better soil health management and optimized fertilizer application. In terms of irrigation management, satellite data helps assess water stress and schedule irrigation more efficiently, conserving water and reducing operational costs. Furthermore, EO data is critical in disaster impact assessments, such as evaluating crop damage from floods, droughts, or pest invasions, which is essential for relief planning and insurance claims. The integration of such data with AI and IoT allows farmers to make data-driven decisions at the micro-level, thereby increasing productivity while minimizing input costs. Satellite imagery also contributes to land use and land cover (LULC) analysis, providing insights into agricultural land dynamics and supporting sustainable development initiatives.\u003c/p\u003e"},{"header":"Review methodology","content":"\u003cp\u003eThe methodology for selecting diverse literature to review the application of Remote Sensing, Precision Agriculture, and AI/ML begun with a comprehensive search across various multidisciplinary online databases, including Springer, Taylor \u0026amp; Francis, Wiley, Google Scholar, MDPI, Annual Reviews, Science Direct, Scopus, Web of Science, IEEE Xplore, and other Scopus-indexed journals relevant to various agricultural domains utilizing those approaches. Several academic publications about surface soil, soil parameters, plant physiology, different irrigation techniques, pest, disease, weed assessment, detection and control strategies, harvesting and yield optimization and supply chain logistics found in these large database research libraries. Many significant studies were mentioned in the Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e and the most recent, creative and relevant research papers and reviews were chosen for this study. The following keywords: artificial intelligence, machine learning, supply chain logistics, remote sensing, soil management, irrigation, plant diseases, and pests were used for search. To guarantee that the concepts were consistent, the articles were carefully chosen, reviewed, and summarized.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003ePopular Earth Observation (EO) satellites data and information Portals from India and world-wide\u003c/b\u003e\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"3\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSr. No.\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEarth Observation (EO) satellite data Portals\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eInformation provided by portals\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003ctr\u003e\u003cth align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003eIndian Portals\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eVEDAS (Visualisation of Earth Observation Data and Archival System)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eISRO’s platform for EO data visualization and thematic applications\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBHUVAN (ISRO)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGeo-portal for visualizing and downloading satellite data and thematic layers\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMOSDAC (Meteorological and Oceanographic Satellite Data Archival Centre)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSpecializes in meteorological, oceanographic, and climate data from ISRO satellites\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIndia WRIS (Water Resources Information System)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eJointly developed by ISRO and the Ministry of Jal Shakti for water resource monitoring\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eBHOONIDHI Geo-portal\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDeveloped by ISRO to provide streamlined access to Earth Observation (EO) satellite data\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colspan=\"3\" nameend=\"c3\" namest=\"c1\"\u003e \u003cp\u003e\u003cb\u003eInternational Portals\u003c/b\u003e\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCopernicus Open Access Hub (ESA - Europe)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFree access to Sentinel satellite data (Sentinel-1, -2, -3, and − 5P)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUSGS Earth Explorer (USA)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOffers access to Landsat, MODIS, ASTER, and other datasets\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eNASA Earth data Search\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eProvides a wide range of NASA satellite missions including MODIS, VIIRS, and SMAP\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eGoogle Earth Engine\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCloud-based platform offering access to petabytes of EO data for analysis and visualization\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOpen Data Cube (ODC)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eOpen-source platform providing standardized access to EO data for analytics\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003ePrecision Agriculture can be practiced at different crop stage ranging from seed selection to harvesting and logistics. All these applications are briefly given (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) in tabular form further discussed here under.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003cdiv class=\"gridtable\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003eApplications of Precision Agriculture, AI and ML Technologies in different agricultural domains\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003c/colgroup\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSr. No.\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eStudy Objective\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eInstrument used\u003c/p\u003e \u003cp\u003e(\u003cem\u003eIn-situ\u003c/em\u003e data)\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eOptical data used\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eOptical data obtained from\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMethods\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e \u003cp\u003eStudy Area\u003c/p\u003e \u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e \u003cp\u003eReference\u003c/p\u003e \u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e1.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSoil textures and nutrients estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGround data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLandsat − 8 OLI/TIRS C2- L1, Landsat − 8 OLI/TIRS C2- L2, Sentinel 2A\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eUnited States Geological Survey (USGS)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGradient Boosting (GB), Extreme Gradient Boosting (XGB),\u003c/p\u003e \u003cp\u003eRandom Forest Regressing (RFR), Adaptive Boosting (ADA)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eNorth India-Punjab\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eDhiman, \u003cem\u003eet al\u003c/em\u003e.,(2023)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e2.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSurface Soil Moisture (SSM) estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGravimetric method\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eLISS-III sensor multispectral imagery with a spatial resolution of 23.5 meters\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eIndian Earth observation satellite Resourcesat-2 LISS-III sensor\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModified Dubois Model (MDM) for estimating SSM\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBhal Region Gujarat (Wheat crop)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eDave \u003cem\u003eet al\u003c/em\u003e.,(2019)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e3.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSoil mapping and soil texture estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGravimetric method\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSentinel-1 (S-1) and Sentinel-2 (S-2)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCopernicus (in the form of Level-1 Ground Range Detected (GRD) products\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003esupport vector machine (SVM) and random forest (RF) algorithms with a three-fold cross-validation approach\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003esemi-arid region in central Tunisia\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eBousbih et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e4.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eSurface Soil Moisture (SSM) estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eHydraGo Probe sensor and surface roughness synchronizing with satellite pass dates\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eSentinel-1 C-band SAR data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCopernicus Ground Range Detected (GRD) products\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eModified Dubois Model (MDM) for estimating SSM\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAnand District of Gujarat\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMurugesan, et al. (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e5.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eOperational 500 m surface soil moisture estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ehandheld soil probes\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eC-band SAR data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eISRO’s high resolution EOS-04 (RISAT-1A)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eRT model (tau-omega model)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAnand (Gujarat), Ludhiana (Punjab)\u003c/p\u003e \u003cp\u003eHisar (Haryana) Kanpur (Uttar Pradesh) and Berambadi watershed (Karnataka)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePandey et al. (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e6.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eEffects of soil compaction on canopy reflectance cotton yield, soil EC\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003edigital cone penetrometer, conductivity meter, spectro-radiometer (EPP 2000C)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eGround-based spectral data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGreen NDVI\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eFayetteville, Arkansas\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eKulkarni et al. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2010\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e7.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAssessing potential of polarimetric decomposition parameters for monitoring rice crop biophysical parameters\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGround data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003eC-band SAR data from Sentinel-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eC-band SAR data from Sentinel-1\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eField surveys synchronized with Sentinel-1 satellite passes\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAnand District of Gujarat, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eDave et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e8.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo analyze and compare performance of three commercial soil nitrate probes\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNutrisens, RIKA, JXCT (conductivity-based probes)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGravimetric method\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eUniversity of Barcelona, Spain\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eBellosta-Diest et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e9.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo predict organic carbon (OC), pH, available N, P, K (EC), zinc (Zn), soil texture\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ePortable VIs–NIR spectrometer (350–2,500 nm) and conventional lab instruments\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003eSentinel-2 satellite imagery (for spatial mapping)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGround data (sample collection) and PLSR,SVMR modelling (with PCA) to predict soil variables\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTarekswar region, Hooghly district, West Bengal, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eSingha et al. (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e10.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo develop and test a solar-powered smart irrigation control system\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSoil moisture \u0026amp; water level sensors, Microcontroller (for automation),\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGravimetric method validation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eKabanyolo, Uganda\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eWanyama et al. (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e11.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eTo develop and demonstrate a solar-powered IoT irrigation system\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIn-situ data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eBangladesh Agricultural University, Mymensingh, Bangladesh\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAl Mamun et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e12.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCrop transpiration estimation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eInfra-Red Camera With UAV system\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eThree-temperature (3T) model, canopy segmentation, scale error analysis, isotope validation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eLiangzhou District, Wuwei City, Gansu province, of Northwest China\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eHou et al. (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e13.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAssessment of Okra YVMV with ground based Hyperspectral imaging\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNon-imaging Spectroradiometer (SVC-LC-RPPro 350 to 1050 nm)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eNDVI ,RVI, GI, PRI, MCARI, and RVSI, RENDVI\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAnand District of Gujarat, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMishra et al. (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e14.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAssessment of aphid infestation in Mustard with Hyperspectral imaging\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eFieldSpec-3 hyperspectral spectroradiometer (350–2500 nm)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eNDVI, RVI, Aphid Index (AI), and SIPI\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndian Agricultural Research Institute (IARI), New Delhi, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eKumar et al. (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2013\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e15.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003efaba bean leaf disease identification with Deep CNN\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eDigital camera\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCNN Architecture based image processing\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eRepublic of Korea\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eJeong and Na (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e16.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eML-IoT Sensor based smart, continuous crop disease monitoring system development\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIoT sensor nodes\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eEnsemble nonlinear SVM (ENSVM), CNN, Naïve Bayes, and K-Nearest Neighbors\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eNagasubramanian et al. (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e17.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eAI-powered spot-sprayer to detect/spray weeds precisely\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLogitech C920 RGB cameras; NVIDIA TX2 \u0026amp; GTX 1070 Ti GPUs; TOPCON RTK GPS\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eHardware setup (pump, nozzles, manifold), Tiny YOLOv3 detection, GPS-enabled nozzle actuation\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eFlorida, USA\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePartel et al. (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e18.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eUAV-based weed patches detection within/between wheat rows using OBIA.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUAV with RGB + multispectral sensors, flying at 30–60 m.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eVI thresholding, object segmentation (OBIA), connected-component classification\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003ePakistan\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eMateen and Zhu (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e19.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMonitoring of harvesting combine with smart devices\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGPS receiver model: (Teltonika FMB92) for real-time tracking\u003c/p\u003e \u003cp\u003eGrain-level sensor installed in the tank to monitor fill levels\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTashkent Region, Uzbekistan\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAstanakulov et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e20.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eMonitoring of harvesting combine with smart devices to minimize seed loss in sunflower harvesting\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eGPS receiver (Teltonika FMB920) and grain-level sensor (Escort DB-2) mounted on combine\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eTashkent Region, Uzbekistan\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eOchildiev et al. (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e21.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIoT-based autonomous mobile robot system for pitaya harvesting integrating SLAM, AI object recognition\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eNVIDIA Jetson Nano development board, 2D LiDAR (for SLAM), AI edge computing module, robot arm/gripper with sensors\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e2D SLAM for mapping/navigation; YOLO-based AI for fruit detection\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eChen et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e22.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003ereal-time detection and fruit load estimation in mango orchards\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRGB camera with 720 W LED flood lighting mounted on a vehicle at 6 km/h\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCompared 6 deep learning architectures (e.g., Faster R-CNN, YOLO variants\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eAustralia\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eKoirala et al. (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e23.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIoT-enabled mechatronic module for real-time yield monitoring during harvesting of greenhouse and vineyard produce\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eLoad sensor-based IoT weighing system on a robotic arm/gripper (gripper-cum-cutter), connected via MQTT to server\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eNational Research Centre for Grapes \u0026amp; greenhouses in Pune district, Maharashtra, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eKolhalkar et al. (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e24.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eCrop yield prediction algorithm\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eIoT sensors across fields collecting climate (temperature, humidity), meteorological, soil chemical, and yield data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eMachine-learning regressors (DecisionTree, RandomForest, ExtraTree), with an active-learning approach to minimize labeled data needs\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eKafrelsheikh, Egypt\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eTalaat (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e25.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eForecast wholesale brinjal prices using various ML models\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003erelied on time-series price data from Agmarknet\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eGRNN, SVR, RF, GBM vs ARIMA; evaluated using RMSE, MAE, MAPE, Diebold-Mariano tests\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eOdisha, India\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003ePaul et al. (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e26.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eDeep LSTM for Agricultural Price Forecasting\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003ehistorical market price datasets\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eDeveloped deep LSTM model trained on time series of commodity prices\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eJaiswal et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e27.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eRisk assessment and monitoring in fresh produce logistics\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eUtilized RFID logs and IoT sensor data (e.g., temperature/humidity in supply chain)\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eSVM to predict logistics risk score based on sensor-instrumented shipping data\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eChina\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eZhang et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e28.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eIoT-Based Real-Time Monitoring and Notification System of Cold Storage\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eSensors for temperature, humidity, door motion, connectivity via IoT modules\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eWeb-dashboard alerts, mobile notifications; threshold-based monitoring\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003ePakistan\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAfreen and Bajwa (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003e29.\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003eperishable food traceability in supply chain\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003eRFID tags + IoT environmental sensors (temp/humidity) on packaging\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e-\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003eCombined RFID and sensor data; ML models identified anomalies to prevent spoilage\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c7\"\u003e \u003cp\u003eIndonesia\u003c/p\u003e \u003c/td\u003e\u003ctd align=\"left\" colname=\"c8\"\u003e \u003cp\u003eAlfian et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e)\u003c/p\u003e \u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/table\u003e\u003c/div\u003e\u003cp\u003e\u003c/p\u003e\u003ch2\u003e2.1 Soil Management:\u003c/h2\u003e\u003cp\u003eAccurate estimation of soil and plant nutrient levels is critical for enhancing crop growth and optimizing agricultural yield. Traditional soil analysis techniques, while accurate, are often labor-intensive and time-consuming effort. Innovative UAVs (Unmanned Aerial Vehicles) mounted with hyperspectral and multispectral imaging are precise but costlier which is why it is not widely adopted. However, with the use of advanced remote sensing technologies such challenges can be accomplished. Dhiman and their co-workers (2023) investigated such an alternative approach by utilizing open-source multispectral satellite imagery in combination with ground truth data collected from the agriculture-intensive Punjab region of India. They estimated concentrations of nine key soil nutrients and textures, \u003cem\u003ei.e.\u003c/em\u003e, Nitrogen (N), Phosphorous (P), Carbon (C), Ammonium (N-NH₄), Nitrate Nitrogen (N-NO₃), Calcium (Ca), Magnesium (Mg) and texture Clay and Silt. Further satellite imagery data quality was improved by applying DOS1 atmospheric correction which was compared against raw (uncorrected) datasets. Soil nutrients and textures estimation tasks performed using four ensemble machine learning models. Their findings indicated that raw imagery consistently outperforms the DOS1-corrected data across most parameters. Moreover, the analysis shows that certain soil properties have a stronger correlation with multispectral data, suggesting that remote sensing has significant potential in nutrient and texture estimation, offering a cost-effective and scalable solution for soil monitoring. Sustainable agriculture depends on efficient soil and land management, which is influenced by topographical features (such as elevation and slope), soil characteristics (including moisture content, organic matter, pH, drainage, and electrical conductivity), along with physiological, biochemical, and morphological and crop traits (Nock et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Proper soil mapping, erosion control and irrigation management are essential for enhancing crop productivity. Additionally, factors like crop water needs, water and energy balance, runoff, evapotranspiration, and climatic modelling contribute significantly to agricultural efficiency. Among these, Surface Soil Moisture (SSM) plays a crucial role in regulating irrigation practices, ensuring water availability, and improving overall crop performance. In this direction, Dave et al. (\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) evaluated Modified Dubois Model (MDM) for estimating surface soil moisture (SSM) using dual-polarization C-band SAR data from RISAT-1 over winter wheat crop from initial vegetative to maturity stage within Bhal region of Gujarat. By incorporating field-measured soil moisture and surface roughness across different crop stages (bare soil, sowing, vegetated), the model was calibrated and tested for σ⁰ (VV and VH) backscatter sensitivity. The MDM showed strong correlation with measured soil moisture, especially for bare and sparsely vegetated conditions, outperforming the original Dubois model. Validation with independent datasets confirmed its robustness for operational SSM retrieval in agricultural landscapes using RISAT-1 data. Agrawal et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) developed and evaluated smart Soil Sensor for Hydrology and land Applications (SHOOL) which is a low-cost, compact soil moisture sensor developed by (Space Applications Centre-Indian Space Research Organisation) SAC-ISRO for field-based soil moisture monitoring. It measures dielectric constant, electrical conductivity, temperature, and volumetric soil moisture, transmits data via a smartphone app, making it suitable for dense sensor networks aimed at validating satellite-derived soil moisture products. They compared performance of two commercial sensors—MP306 (ICT) and Stevens Hydra Probe with SHOOL and validated the data using gravimetric sampling across agricultural fields in Gujarat. They claim that SHOOL performed better than both commercial sensors, achieving R² = 0.82 and RMSE = 2.94, indicating strong correlation and low error based on statistical analysis. To support effective soil and land analysis, various technologies such as soil moisture sensors with probes (Hydrago-stevans, Shool-Neerx), GIS mapping, and remote sensing data from various satellites and aircrafts can be utilized not only they are free but also providing valuable insights for precision agriculture. Alternatively, integrating remotely sensed optical and radar data for can be used for soil mapping, in this direction, Bousbih et al. (\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) used and integrated remotely sensed optical and radar data for soil mapping and soil texture estimation by utilizing Sentinel-1 (S-1) and Sentinel-2 (S-2) imagery acquired between July and December 2017 for 3000 km² semi-arid region in central Tunisia. They found clay content ranging from 13% to 60% from multiple agricultural fields alongside satellite-derived indicators, including Short-Wave Infrared (SWIR) bands and soil indices, particularly over bare soils and soil moisture products generated from combined S-1 and S-2 data were also tested as proxies for soil texture. They also applied support vector machine (SVM) and random forest (RF) algorithms for clay content mapping, with a three-fold cross-validation approach for assessment. They found high accuracy classification while using soil moisture indicators derived from combined S-1 and S-2 data, yielding overall accuracies of 63% and 65% for SVM and RF, respectively. Thus, indicating potential of integrating multi-source remote sensing data for improved soil texture mapping in semi-arid environments. The advanced Soil Moisture probes (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e, \u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e) can measure soil properties (Moisture, temperature, Bulk EC, Pore water EC, raw real dielectric) along with latitude, longitude for geo referencing. Further it is very easy to transfer data digitally. Murugesan, et al. (\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) evaluated the Modified Dubois Model (MDM) for estimating surface soil moisture (SSM) in bare agricultural fields using Sentinel-1 C-band SAR data in a semi-arid region of Gujarat, India. They collected field soil moisture data from 102 locations, using HydraGo sensors (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), gravimetric methods and estimated surface roughness for validated SAR-derived soil moisture values. Further applied MDM to derive soil permittivity from VV-polarized backscatter, which was then converted to volumetric moisture using Topp’s model. They noted strong agreement with ground measurements (R² = 0.81, RMSE = 0.005 m³/m³), especially in the low to moderate moisture range (0.015–0.1 m³/m³) also reduced accuracy observed at higher moisture levels due to surface roughness effects. They confirm that MDM, coupled with Sentinel-1 data, is a reliable tool for SSM monitoring in bare soils and holds potential for improving irrigation planning and water resource management.\u003c/p\u003e\u003cp\u003e \u003c/p\u003e\u003ch2\u003e2.2 Sowing and Planting Operations using Precision Agriculture\u003c/h2\u003e\u003cp\u003eGermination and Seedling emergence are highly influenced by soil moisture and soil temperature. In the same way, fertilizers are usually applied when fields are not excessively wet to reduce nutrient loss ultimately enhancing crop nutrient uptake. Seedling rate for different crops, soil types and variable seasons can be dictated by conducting land mapping studies as it impacts hydrological balance influencing soil moisture and temperature (Khanal et al., \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Microwave signals are considered more accurate for the soil moisture estimation as they are not get interfered by atmospheric and cloudy conditions unlike the RS data obtained from visible, near-infrared (NIR), and short-wave infrared (SWIR) bands. So, various satellites equipped with such active and passive microwave sensors, \u003cem\u003ei.e.\u003c/em\u003e, Soil Moisture Active Passive (SMAP) and Sentinel-1, are currently used for soil moisture monitoring. In order to meet the need for sub-km resolution soil moisture data in agriculture, Pandey et al. (\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) developed a downscaling framework to disaggregate coarse brightness temperature from the SMAP radiometer using ISRO’s high-resolution EOS-04 (RISAT-1A) C-band SAR data.\u003c/p\u003e\u003cp\u003eAnother crucial parameter is soil compaction, it impacts hydrolic conductivity, porosity and nutrient availability of soil so it is necessary to assess soil compaction. This information allows end user (farmer) to make a decision about crop selection based on soil properties. In this direction, Kulkarni et al. (\u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2010\u003c/span\u003e) in Arkansas, examined the impact of soil compaction on canopy spectral reflectance and cotton yield using ground-based spectral data (using cone penetrometer) and demonstrated that soil compaction effects can be evaluated by analyzing variations in the Green Normalized Difference Vegetation Index (GNDVI), which is derived from the green and near-infrared (NIR) spectral bands of imagery.\u003c/p\u003e\u003cp\u003eThe spatial and temporal variations in crop population distribution and density within a field can significantly influence grain and biomass yields, as well as overall efficiency of harvesting. Timely access to this information is critical for replanting and optimizing mid-season management practices, including the precise application of variable-rate fertilizers, herbicides, and pesticides. Very few studies based on the quantification of crop growth stages has been carried out in that direction, Dave et al. (\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) analysed potential of polarimetric decomposition parameters of Sentinel–1 dual-polarization SAR (Synthetic Aperture Radar) data in monsoon, to estimate biophysical parameters of rice crop \u003cem\u003ei.e.\u003c/em\u003e, fresh biomass, dry biomass, vegetation water content (VWC) and plant height and noted that the polarimetric decomposition parameters are sensitive to biophysical parameters and can be used as additional parameters along with the backscatter parameters for rice crop monitoring.\u003c/p\u003e\u003cp\u003eMoving on to sensors that can assist in assessing soil fertility and properties prior to sowing and planting operations, it is essential to recognize their significance in optimizing agricultural productivity. Accurate and timely evaluation of soil characteristics plays a critical role in ensuring informed decision-making regarding land preparation, nutrient management, and crop selection. In india, Innovative on-site soil testing and monitoring solutions, such as SoilSens and NutriSens have been developed to facilitate real-time assessment of soil health and support data-driven decision-making in agricultural management.\u003c/p\u003e\u003cp\u003eElectrochemical sensors for soil quality assessment offer a cost-effective means of providing real-time, accurate insights into nitrate concentration trends, thereby supporting optimal fertilization practices and minimizing nutrient-related environmental impacts on surface and groundwater, Bellosta-Diest et al. (\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) conducted nitrate concentration estimation experiment in the soil matrix and evaluated the performance of three electrochemical sensors \u003cem\u003ei.e.\u003c/em\u003e, Nutrisens, RIKA and JXCT, across three scenarios: (1) varying electrical conductivity (EC) in the absence of nitrate (0–56 mS/cm), (2) nitrate concentration in aqueous solutions (0–180 mg/L), and (3) nitrate concentration in saturated sand, used as a soil proxy, at 35% volumetric moisture (0–150 mg/L). Among the three, the Nutrisens probe, which utilizes an ion-selective electrode (ISE), demonstrated high sensitivity and consistent, accurate performance, exhibiting a strong inverse relationship with nitrate concentration (R² ≈ 0.99). In contrast, the RIKA and JXCT sensors, which are presumed to rely on conductivity-based measurements, were highly affected by EC variations, showed inconsistent outputs, and displayed poor sensitivity in soil-like conditions likely due to inadequate calibration and a lack of ion specificity for nitrate.\u003c/p\u003e\u003cp\u003eAnother study of soil nutrients prediction was conducted by Singha et al. (\u003cspan citationid=\"CR58\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) employing Visible to Near-Infrared (VIs–NIR) spectroscopy in the range of 350–2500 nm, in combination with Partial Least Squares Regression (PLSR) and Support Vector Machine Regression (SVMR) models. The research, carried out in the Tarakeswar region of West Bengal, utilized 200 soil samples along with Sentinel-2 satellite imagery to estimate key soil parameters, including organic carbon (OC), NPK (nitrogen, phosphorus, potassium) pH, and soil texture (sand, silt, and clay). Both PLSR and SVMR models demonstrated high accuracy in predicting OC and P, with moderate performance for other nutrients, while predictions for soil texture were found to be less reliable. Using Sentinel-2 imagery and geostatistical techniques, the study also generated soil suitability maps, achieving up to 89% accuracy demonstrating potential of integrating machine learning and spectroscopy for cost-effective, efficient, and non-destructive soil monitoring.\u003c/p\u003e\u003ch2\u003e2.3 Precision Irrigation/ Water Management\u003c/h2\u003e\u003cp\u003eIn Indian agriculture, Rainfall is a significant water source followed by irrigation but in Indian context, rainfall is very inconsistent. Sometimes, excessive amount of it can cause soil erosion and surface runoff while other times, drought conditions can also occur inducing water scarcity. In such situation, irrigation, a regulated technique of supplying crops with water in place of natural rainfall, is frequently used to lessen these impacts, making sure that plants get the moisture as and when they need for healthy growth and development (Oborkhale et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn conventional irrigation system, water is delivered uniformly, throughout the farm, without accouting for crop-specific water requirements or field variability which causes inefficiencies such as over-irrigation in some areas and under-irrigation in others, inducing water stress. In addition, commercial irrigation controllers are operated by pre-set programs based on empirical data. Furthermore, challenges like dynamic nature of plant water intake and weather patterns, environmental disruptions, and water scarcity driven by drought and climate change must need consideration. (Abioye et al., \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Under such scenario, precision irrigation reduces the wastage of water by considering spatial and temporal soil variation, soil structure, hydraulic properties, plant responses to water deficit, changing weather variables through efficient Internet of Things (IoT) monitoring, to make better irrigation decisions that could lead to increased production and significant water savings, since most countries in the world are facing water scarcity problems (Bitella et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2014\u003c/span\u003e, Capraro et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). According to information reviewed by Abioye et al. (\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), subsurface irrigation is highly effective in water-saving and gives better yield output when compared with other type of surface irrigation. Different types of soil moisture sensors (Figure: 1,2,4), temperature and humidity sensors are used to monitor environmental conditions and further the data from such sensors can be collected and sent to the cloud using IOT platforms. Then this information is analysed by machine learning algorithms and artificial intelligence. In precision agriculture, IOT platforms are extensively used. Sometimes electricity/power lack is big concern. Harishankar et al. (\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2014\u003c/span\u003e) have developed power efficient solar based systems wherein the controller was linked to the soil sensor and water supply valve. Moisture sensor measures the water level and decides whether to switch the water valve ON or OFF. Since the solar panel provided the electricity, the system didn't require an external power module. Wanyama et al. (\u003cspan citationid=\"CR64\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), developed a mobile, solar-powered control system for real-time irrigation scheduling and water delivery, using data from soil moisture sensors. The system included an automated tank water level control mechanism that activates the pump during irrigation. Soil moisture sensors were placed at a rooting depth of 15–30 cm, where two probes allowed electrical current to pass through the soil, measuring resistance to determine moisture levels. Wet soil, being a better conductor, exhibited lower resistance, whereas dry soil, with poor conductivity, showed higher resistance. The study also confirmed that the soil moisture sensors provided readings with no significant difference from the gravimetric method.\u003c/p\u003e\u003cp\u003eIn same way, Al Mamun et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2025\u003c/span\u003e) also proposed an IoT-enabled solar-powered smart irrigation for precision irrigation, included solar panel for power, Raspberry Pi 4 for pump control, and a 12V 7.5Ah battery for energy storage. Microcontroller was integrated with a Capacitive Soil Moisture Sensor and other sensors to monitor air temperature, and humidity data. They also created a website which aids in real-time monitoring and to control irrigation pump remotely. Zhou et al. (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) reviewed use of a non-destructive, rapid, and reliable method to assess crop water stress using infrared thermal imagery, from early satellite-based systems to modern uncooled thermal cameras mounted on ground-based platforms and unmanned aerial vehicles (UAVs) to optimize irrigation and maintain crop productivity. They used canopy segmentation and the Crop Water Stress Index (CWSI), which correlate intensely with physiological indicators such as leaf water potential and stomata conductance giving valuable insights of plant health. Further, integration of deep learning enhance the accuracy of stress detection and prediction, enabling more efficient and data-driven irrigation scheduling especially in water-scarce regions. Hou et al. (\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) conducted a study using thermal infrared (TIR) imaging to estimate crop transpiration in corn and soybean under regulated deficit irrigation. They used a three-temperature (3T) model to analyze the scale effect between ground and UAV thermal images. This allowed them to detect water stress early, improve water use, and avoid excessive and poor watering, enabling precision irrigation. UAV-based thermal imaging captured large field spatial variability, identifying stressed areas for targeted actions like fertilizer application or irrigation modifications. This approach allows for comprehensive crop health monitoring, including illnesses and nutrient deficiencies.\u003c/p\u003e\u003cp\u003eHence, using precise sensors to monitor soil moisture levels can help optimize irrigation by determining how much water to apply and where it is needed. Integrating AI and machine learning-driven systems allows for real-time water level adjustments while incorporating local climate data can enhance irrigation efficiency.\u003c/p\u003e\u003ch2\u003e2.4 Crop Health \u0026amp; Disease Management\u003c/h2\u003e\u003cp\u003eThe production of crop plants is hampered by both biotic and abiotic factors, such as weeds, viruses, diseases, arthropods, and other animal pests. The potential loss caused by Fungi, oomycetes, bacteria, and viruses and the actual losses have been predicted by Oerke (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) to reach 16% and 11% of the world's achievable crop production, respectively. So, in order to enhance crop productivity, crops must be protected from such large per cent of damage. So as to manage any pest or disease, correct diagnosis is crucial.\u003c/p\u003e\u003cp\u003eFields are sprayed at constant rates because disease management in agricultural crops frequently assumes that pathogens are dispersed uniformly. However, crop heterogeneity, which is influenced by geography, soil conditions, neighbouring fields, microclimate shifts, and infection sources, usually leads to irregular disease patterns that manifest as patches, gradients, or haphazard clusters. In the same field, these patterns might change throughout the course of a single growing season, as well as between places and years. In order to determine whether spraying is required, farmers typically visually evaluate a representative number of samples to determine the occurrence and severity of the disease and take decision whether to spray or not in field. Since, disease diagnosis and quantification currently rely on visual rating of plants which may be inaccurate and when in doubt, special serological and nucleic acid technologies i.e., ELISA (Enzyme-Linked Immunosorbent Assay), LAMP (Loop-Mediated Isothermal Amplification), PCR (Polymerase Chain Reaction), PAGE (Polyacrylamide Gel Electrophoresis) can be performed but those laboratory procedures are destructive and invasive covering only a representative sample of crops of interest. So, here remote sensing approaches can be a great help especially with the diseases which can induce latent or slow infection i.e., viral diseases, continuous monitoring can help in prediction of their occurrence and further spread. Apart from that, remote sensing technologies like Thermography, fluorescence imaging, and spectral techniques all are non-invasive and non-destructive which potentially performs repeated monitoring of all plants of interest.\u003c/p\u003e\u003cp\u003eFor the early detection of grapes diseases Patil \u0026amp; Thorat (\u003cspan citationid=\"CR51\" class=\"CitationRef\"\u003e2016\u003c/span\u003e) developed agricultural monitoring system based on machine learning and relative humidity sensors, temperature sensors and also the leaf wetness sensors. The data collected at regular intervals are sent using Zig-bee module to the server.. The server here employs the hidden Markov model algorithm towards training the data sets pertaining to temperature, relative humidity and leaf wetness for analysing the data towards predicting the chance of disease on grapes prior infection. This information is sent as an alert message via SMS to the farmer and expert. The system employs machine learning in early and accurate detection of disease in grapes and suggests pesticide to protect the crop from disease and reduce manual disease detection efforts which can be helpful for farmers providing information for scheduling fertilizers, pesticide spraying, irrigation \u003cem\u003eetc\u003c/em\u003e further,improving the quality and quantity of grapes. The main drawback of this system is its inability to handle the system by farmers because of their lack of knowledge in the technology and the literacy problems.\u003c/p\u003e\u003cp\u003eBecause hyperspectral remote sensing captures comprehensive reflectance data throughout a wide spectral range, it has proven to be an effective technique for early plant disease identification. It was shown to be effective in evaluating Okra Yellow Vein Mosaic Virus (OYVMD) infection by Mishra et al., (\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), where they used variety of vegetative indices to categorize the severity of the disease. Chlorophyll absorption-related notable spectrum fluctuations around the 650 nm wavelength were identified as important markers of plant stress. Principal Component Analysis (PCA), which was primarily influenced by indices such as NDVI and RVI, stated 75% of the spectral variation in the study. K-means clustering was able to distinguish between healthy and moderately diseased plants, demonstrating its reliability in pattern recognition. The potential of hyperspectral remote sensing to improve disease classification accuracy and advance precision agriculture techniques is demonstrated by this integrated strategy. Kushwaha et al. (\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) demonstrated a robust approach of non-destructive estimation of chlorophyll-a and -b content across various crops using airborne hyperspectral data from AVIRIS-NG. They employ second derivative reflectance analysis, identifying wavelengths at 672 nm and 587 nm as most sensitive to Chl-a and Chl-b, respectively. They developed linear models within a bootstrapped resampling framework, achieving moderate R² values (0.46 for Chl-a and 0.51 for Chl-b during calibration, dropping to 0.3 in validation). Kumar et al. (\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) demonstrated the potential of hyper-spectral remote sensing for assessing aphid infestation in mustard crops. Conducted over two growing seasons, their research compares healthy and aphid-infested canopies, revealing significant reductions in leaf area index (67–94%) and chlorophyll concentration (up to 50%) due to infestation. Along with such physiological changes, reductions in oil content (13–16% lower in infested plants) and seed yield, underscore the severity of aphid damage. Spectral analysis of healthy plants showed stronger reflectance peaks around 550–560 nm and higher near-infrared reflectance, while infested plants exhibit diminished signals, particularly between 1950–2450 nm — a range identified for distinguishing infestation levels. They also correlated spectral indices (NDVI, RVI, AI, SIPI) with aphid damage which may serve as non-invasive tool for early detection and monitoring of crop health for on time pest management. Given the essential role of legumes in food security and the rising demand-supply gap, particularly in South Asia, the study by Jeong and Na (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) presented an advanced approach to Faba bean leaf disease identification using deep Convolutional Neural Networks (CNNs). They used the model structured with eight convolutional layers, four max-pooling layers, and three dropout layers to prevent overfitting, achieving impressive performance — 99.37% accuracy during training and 89.69% during validation. The CNN demonstrated high precision, recall, and F1 scores across multiple disease categories: Chocolate Spot, Faba Bean Gall, Rust, and Healthy leaves, with an overall accuracy of 91%. Notably, Chocolate Spot and Rust showed the highest classification performance. They further propose future enhancements like incorporating hybrid machine learning models (e.g., Support Vector Machines, decision trees) and expanding datasets to include stem and root infections for more comprehensive disease detection. The application of IoT and machine learning in agriculture, namely for precision farming and crop disease monitoring, has been the subject of many research investigations. Among these, the study by Nagasubramanian et al. (\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) presents an all-inclusive system called ECPRC (Ensemble Classification and Pattern Recognition for Crops), which uses intelligent classification models and Internet of Things-based sensing to detect plant diseases in real time. The system combines spectral cameras that take high-resolution pictures of leaves with a variety of environmental sensors that track temperature, soil moisture, and light intensity. An Arduino-based hardware module is used to process these data, and convolutional neural networks (CNN) and ensemble nonlinear support vector machines (ENSVM), two sophisticated machine learning approaches, are used for analysis. CNN succeeds at image-based recognition via deep-layer filtering and pattern learning, while ENSVM manages complicated disease patterns using multiple sub-classifiers.\u003c/p\u003e\u003cp\u003ePre-processing techniques such as image segmentation, noise reduction, and enhancement further improve detection accuracy by isolating disease symptoms like spots and discolorations. as stated in the study, CNN performs better in terms of accuracy, recall, and precision than conventional classifiers like Naïve Bayes, K-Nearest Neighbors, and regular SVM. In the meantime, ENSVM is perfect for lightweight IoT devices because it exhibits effective performance with reduced computing demands. The technology offers insights for managing irrigation and nutrients, facilitates real-time environmental monitoring, and detects diseases. With its mobile accessibility and cloud storage, ECPRC provides farmers with a useful tool for decision support. This study promotes precise, scalable, and real-time crop health monitoring solutions, highlighting the growing importance of hybrid IoT-ML/DL systems in present-day agriculture.\u003c/p\u003e\u003ch2\u003e2.5 Precision weed control\u003c/h2\u003e\u003cp\u003eWeeds are undesirable plants that invade farmland, competing with crops for essential resources like nutrients, sunlight, and space. If left unchecked, they can hinder crop development, leading to lower yields and, ultimately, financial losses for farmers by increasing input costs, (Marco et al., \u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Moreover, compared to crop diseases (25%) and insect pests (20%), weeds account for almost 45% of field crop output losses, making them the most expensive category of agricultural risks (Gnanavel, \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Weeds which are not removed can cause a yield loss of 100%, (Chauhan, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Furthermore, weeds interact with other organisms [insects and pathogens (bacteria and fungi)] in the ecosystem and serves as host for spreading various diseases which can seriously harm crop plants (Esposito et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Additionally, weeds degrade the value of land, particularly perennial and parasitic weeds, and slow down water management by reducing water flow in irrigation ditches and increasing evapotranspiration losses according to Monteiro and Santos (\u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Hence effective weed control is demand of time which not only promotes healthier plants but also improves resource efficiency by allowing crops to absorb more water, nutrients, and sunlight. Amongst different weed management methods, farmers readily relies on herbicides and weedicide which again can accumulate in food chain contributing to more environment pollution. Other physical, mechanical, cultural and methods includes hand-weeding, mulching, soil solarisation and weeding equipment. On the other side integrating precision weed management practices aids in managing weeds early which prevents them from producing seeds or developing deeper root systems.\u003c/p\u003e\u003cp\u003eIn, precision weed management, monitoring and detection of weed is first step followed by spot spraying or robotic weed removal. Different methods of monitoring and detection includes, Remote Sensing and UAVs, Robotic Technology and Sensor-Based detection. With this context, Partel et al. (\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) developed low cost, AI-driven smart sprayer prototype which comprised of a machine vision software (AI-based) that uses deep learning to detect specific target weeds along with hardware setup with 12 individual fast-response nozzles for spraying. The sprayer was evaluated by using both artificial and real plants wherein, two GPU united were tested [(i) NVIDIA TX2 GPU, and (ii) NVIDIA GTX 1070 Ti GPU]. The GTX 1070 Ti GPU outperformed the TX2 GPU. In comparison to the TX2 GPU, the GTX 1070 Ti GPU reduced the number of missed targets (Portulaca weed) by 81% (from 43% to 8%), improved the detection system's precision and recall by 20% and 77%, and increased the spraying system's precision and recall by 10% and 59%, respectively. This improvement is explained by the fact that difficult scenarios, where targets are highly similar to non-targets and require more processing power, necessitate a more powerful neural network. This smart sprayer only sprayed on the target weed, Portulaca, after correctly differentiating it from non-target sedge weeds and pepper plants. For further precision of weed detection on field, a weed map was developed using an RTK GPS and smart sprayer, generated map showed a good visual correlation to the real scenario’s picture. Additionally, they suggested using a sensor fusion strategy (like the Kalman filter) to eliminate noise and improve the accuracy of weed recognition. Another study was conducted by Mateen and Zhu (\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) where they exploited UAV technology in order to detect weed for its precise identification within the crop rows as well as between the crop rows in wheat field. The main obstacle to weed patch identification in real agriculture fields is the spectral similarity between weed and crop patches, to solve false identification, extract weed properties and identifing weed patterns, they employed object-based image analysis (OBIA) alongside with color-based thresholding and machine learning techniques. Additionally, they used ortho-mapping to improve image standards. RGB-equipped UAVs were utilized for field image capturing, followed by pre-processing (noise reduction, color enhancement, and segmentation), and feature extraction based on color, texture, and shape. In order to categorize the pixels as either crop or weed, machine learning algorithms were also trained and evaluated. The performance of the classification model was evaluated using metrics such as accuracy, precision, recall, and F1-score. The findings demonstrate that UAV-based weed detection provides a scalable and effective precision agriculture solution, facilitating targeted herbicide delivery and encouraging environmentally friendly agricultural methods.\u003c/p\u003e\u003ch2\u003e2.6 Harvesting and yield optimization\u003c/h2\u003e\u003cp\u003eRecent advances in deep learning have significantly increased fruit detection accuracy, enabling trustworthy identification even in complex and difficult environments. However, there have been difficulties in developing 3D fruit localization, which is essential for robotic harvesting (Thakur et al., \u003cspan citationid=\"CR61\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The efficiency of existing sensors in precisely finding fruits in natural orchard situations has been hampered by issues such complex fruit shapes, various orientations, clustering, variable light levels, and foliage obstacles (Moreno \u003cem\u003eet al.\u003c/em\u003e, 2023). However, by combining the use of laser scanning with depth cameras, the Active Laser-Camera Scanning (ALACS) system overcomes such challenges (Jain et al., \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). ALACS has proven to be more accurate at accurately identifying apples in orchards. When ALACS was tested with a harvesting robot that included a vacuum-based end effector, it was able to harvest fruits with excellent efficiency. (Sharma and Shivandu, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). For modern agriculture, it is essential to understand the spatial distribution of fruits inside tree canopies. Such information allows for more than just fruit counting; it allows for accurate predictions regarding when to harvest, which in turn supports accurate scheduling and efficient resource allocation (Sassu \u003cem\u003eet al.\u003c/em\u003e,2024). Robots designed for activities like picking fruit depend heavily on spatial data (Kolhalkar et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). These robots may mitigate fruit damage and manage orchards more effectively when they have accurate information about where the fruit is located (Sharma \u003cem\u003eet al.\u003c/em\u003e, 2024).\u003c/p\u003e\u003cp\u003eConventional yield prediction methods have depended on statistical models that employ basic weather variables and historical yield data. Although these models offer a fundamental comprehension, the intricacy of agricultural ecosystems often leads them to be misleading. While in Morden approach, machine learning techniques have been incorporated in recent developments to improve yield prediction accuracy. Numerous machine learning algorithms, including random forests, support vector machines (SVM), linear regression, and deep learning models, have been used in studies to examine a variety of variables, such as crop management strategies, soil properties, and weather parameters (Sree et al., \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Predictive models could precisely forecast agricultural yields for the next growing season by utilizing data on soil health, climate patterns, weather forecasts, soil conditions, and past crop yields (Sharma \u003cem\u003eet al.\u003c/em\u003e, 2024).\u003c/p\u003e\u003cp\u003eThe integration of GPS and sensor technologies into combine harvesters is used to enhance efficiency and reduce grain loss during harvesting, with this perspective, Astanakulov et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) used a Dominator-130 combine harvester fitted with a GPS receiver (Teltonika FMB920) and a grain level sensor (Escort DB-2). Incorporation of Grain level sensor and GPS receiver indicated how much grain was being uploaded into the grain tank and at what velocity the combine harvester was operating at in real time. They noticed that the work efficiency and grain loss of the combine were greatly impacted by higher grain yield, higher weed density, and lower moisture content. Grain loss increased and harvester efficiency decreased as yield increased. Similarly, efficiency decreased and grain loss increased in fields with significant weed infestation. In order to reduce losses and increase harvesting productivity, the study emphasizes the importance of adjusting combine settings in real-time field conditions, such as crop yield, weed levels, and grain moisture. Similarly, Astanakulov et al. (\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) they also monitored effects of combine harvester with smart devices in soybean harvesting. One more similar study of grain harvester combine Dominator130 was carried out by Ochildiev et al. (\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) by focusing on enhancing the performance of the Dominator-130 grain harvester for sunflower harvesting by integrating a GPS receiver and grain level indicator sensor. Their objective was to reduce seed loss and improve operational efficiency through precision-based modifications. A specially adapted header was developed, featuring a segmented cutting unit, auger, and custom-designed divider-guides. The researchers tested various lengths and gaps of the divider-guides and found that improper dimensions (either too short or too long) led to significant seed loss due to poor alignment or interference with sunflower stems. The optimized configuration ensured more accurate cuts and effective collection within basket. Additionally, the study examined how combine speed affects performance, observing that while faster speeds increased field capacity, they also resulted in higher seed losses, especially in the header and thresher. Through real-time data collection enabled by GPS and sensors, the team was able to fine-tune the machine settings under actual field conditions. This study demonstrates how precision agriculture tools and simple mechanical adaptations can significantly improve the harvesting efficiency of existing machinery for non-traditional crops like sunflower.\u003c/p\u003e\u003cp\u003eTraditional farming relies heavily on experience and labor. One of the main issues facing agriculture today is a shortage of labor. To address the increasing shortage of agricultural labor, Chen et al. (\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) developed an AIoT-based Autonomous Mobile Robot (AMR) system specifically for efficient dragon fruit (pitaya) harvesting. This system integrates artificial intelligence with Internet of Things (IoT) technologies, featuring edge computing powered by the NVIDIA Jetson Nano and utilizing 2D Simultaneous Localization and Mapping (SLAM) for autonomous navigation in dynamic orchard environments. The robot is equipped with a camera, LiDAR sensor, and a nine-axis gyroscope, enabling it to collect real-time data for mapping, obstacle detection, and accurate fruit identification.A notable component of the system is its AI object detection module, which employs the YOLOv3-Tiny model to distinguish between ripe and unripe pitayas with an impressive 96.7% accuracy. Despite a relatively small dataset, the system's performance was significantly improved through data augmentation and transfer learning techniques. During experimental testing, the robot achieved a 97% harvest success rate, efficiently harvesting up to 30 fruits within an hour. The integration of the Dijkstra algorithm for route optimization further enhanced navigation efficiency. With its modular architecture, low-cost components, and software based on ROS 2 and Ubuntu 18.04, the system is well-suited for real-world deployment in outdoor orchard environments.\u003c/p\u003e\u003cp\u003eA customized deep learning model called MangoYOLO, a mix of YOLOv2 (small) and YOLOv3, was presented by Koirala et al. (\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). It is designed and created for the speed and accuracy of real-time mango fruit detection and orchard yield estimation. With an F1 score of 0.968 and Average Precision (AP) of 0.983, MangoYOLO outperforms other object detection architectures such as Faster R-CNN, SSD, and YOLO versions (v2, v3, and small), detecting mangoes in just 8 ms per tile. Training and testing involved more than 1,500 photos from five orchards, and it performed consistently across various lighting conditions, camera types, orchards, and cultivars. Using a correction factor for occluded fruits, MangoYOLO was able to estimate orchard fruit load within 4.6% to 15.2% of real packhouse counts, offering a useful, affordable, in-field fruit detection approach that is essential for precision agriculture and yield forecasting.\u003c/p\u003e\u003cp\u003eKolhalkar et al. (\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) developed an IoT-enabled mechatronic harvesting module for real-time yield monitoring of grape harvests in vineyards. The system was successfully tested at NRCG, Pune, and integrates a robotic arm equipped with a gripper-cutter mechanism and a strain-gauge-based weighing module to measure the weight of each fruit cluster as it's harvested. Data is transmitted and stored using Wi-Fi and cloud servers, making it accessible remotely. It minimize fruit damage by gripping the stem rather than the fruit surface, thereby preserving the natural waxy layer, which contributes to extended shelf life and improved post-harvest quality. It also significantly reduces labour dependence and associated costs while providing real-time yield data through IoT integration. n addition to field testing, the system was evaluated under laboratory conditions, where various greenhouse vegetables such as bell peppers, long chili peppers, bitter melons, guavas, eggplants, and okra were overhung at natural heights and harvested using the module. However, currently module lacks the capability to detect fruit ripeness, and its application is restricted when dealing with large-sized fruits like watermelon or papaya but these limitations can be overcome by improving the module.\u003c/p\u003e\u003cp\u003eWith the goal to enhance precision agriculture, Talaat (\u003cspan citationid=\"CR60\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) created a revolutionary Crop Yield Prediction Algorithm (CYPA) that combines machine learning and Internet of Things (IoT) technology. By analyzing a variety of agricultural data, such as yield-specific, meteorological, environmental, and chemical factors, the CYPA system is intended to predict crop yields with high accuracy. The model obtained excellent predicted accuracy—up to 99.33%—by combining real-time data collecting via IoT devices with sophisticated regression models like DecisionTreeRegressor, RandomForestRegressor, and ExtraTreeRegressor. They took advantage of a noteworthy innovation called \"active learning,\" which increases model efficiency by choosing the most informative data samples for training, minimizing the need for massive amounts of labeled data, and allowing the model to dynamically adjust to shifting field conditions like pest outbreaks or weather variations. The algorithm was trained and validated using Global agricultural datasets for important crops like wheat, rice, maize, potatoes, soybeans. Furthermore, the study also used statistical methods like multiple regression analysis and Pearson's correlation coefficient to assess how significantly variables like temperature, rainfall, and pesticide use influence crop productivity. The average temperature and yield were shown to be negatively correlated, but rainfall and pesticide use were found to be weakly positively correlated. As a result, CYPA stands apart in its ability to facilitate data-driven agricultural decision-making, providing farmers and policymakers with insightful information for more effective resource management and enhancing food security in the face of climate change. Furthermore, it performed superior in terms of accuracy and usefulness than current benchmark models, such as RNN-based techniques.\u003c/p\u003e\u003ch2\u003e2.7 Supply Chain and logistics\u003c/h2\u003e\u003cp\u003eThe agricultural supply chain is composed of four major components: producers (farmers), distributors, retailers, and consumers. Efficient coordination among these stakeholders is vital for ensuring the smooth flow of agricultural goods from farms to markets. Recent advancements in artificial intelligence (AI), particularly in machine learning (ML), have introduced transformative solutions across various stages of the supply chain, enhancing productivity, reducing waste, and increasing profitability.\u003c/p\u003e\u003cp\u003eOne of the earliest and most impactful applications of ML in agricultural supply chains is demand forecasting. By leveraging historical sales data, market trends, and external variables such as weather patterns and economic indicators, machine learning algorithms can predict future consumer demand with a high degree of accuracy. Compared to traditional forecasting methods, these algorithms provide superior precision, enabling supply chain actors to optimize inventory management, adjust production schedules, and fine-tune distribution strategies. Such foresight helps to minimize both stockouts and overstock situations, improving operational efficiency and customer satisfaction (Saxena et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAs the supply chain progresses from production to distribution, route optimization becomes a critical factor. ML algorithms can analyze real-time data on traffic conditions, weather disruptions, and road infrastructure to identify the most efficient transportation routes. This dynamic decision-making capability results in reduced fuel consumption, lower transportation costs, and faster delivery times, thereby enhancing service reliability and customer experience (Saxena et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRisk management is another crucial area of supply chain management where machine learning is being used. The agricultural supply chain is susceptible to a wide range of risks, including geopolitical tensions, climate variability, market volatility, and supplier disruptions. ML models can process and synthesize data from diverse sources to detect potential risks early. These predictive insights allow supply chain managers to implement proactive mitigation strategies and develop contingency plans, ensuring resilience and continuity in operations (Fan et al., \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2021\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eAt the production level, price prediction and market analysis are crucial for farmers who must navigate volatile agricultural markets. Prices can fluctuate dramatically due to imbalances in supply and demand, weather events, and shifts in global trade policies. By analyzing historical pricing data, current market dynamics, and projected returns, ML models provide actionable insights that empower farmers to make strategic decisions. For example, AI tools can recommend delaying a harvest or temporarily storing produce during periods of anticipated oversupply, thereby helping farmers avoid low prices and potential financial losses (Hassan et al., \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Beyond individual farm profitability, accurate price forecasting plays a broader socio-economic role. Agricultural price instability can affect national economic performance and trigger public concern, especially in countries with large farming populations. Unpredictable fluctuations in the prices of staple crops and food products can lead to reduced farmer income, consumer distress, and, in severe cases, social unrest. In this context, the ability of AI and ML technologies to forecast price trends contributes not only to market efficiency but also to national food security and economic stability. This is particularly vital in major agricultural economics bearing countries such as China and india where price stability supports both rural livelihoods and macroeconomic balance (Wang, \u003cspan citationid=\"CR63\" class=\"CitationRef\"\u003e2023\u003c/span\u003e; Assimakopoulos et al., \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eIn the context of price forecasting, Paul et al. (\u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) explored the comparative performance of machine learning (ML) algorithms in forecasting daily wholesale prices of brinjal across seventeen major markets in Odisha. Given the highly volatile and seasonal nature of vegetable prices, traditional statistical models such as the Autoregressive Integrated Moving Average (ARIMA) often fall short in capturing complex, nonlinear dynamics. In this regard, the study evaluates the predictive capabilities of four advanced ML models—Generalized Regression Neural Network (GRNN), Support Vector Regression (SVR), Random Forest (RF), and Gradient Boosting Machine (GBM)—against the ARIMA model. The analysis employed daily price data from January 2015 to May 2021, obtained from the AGMARKNET portal, and assessed model performance using statistical metrics such as Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Error (ME), and Mean Absolute Percentage Error (MAPE). Additionally, the Diebold-Mariano test and Model Confidence Set (MCS) approach were utilized to statistically compare forecast accuracies. Results revealed that GRNN consistently outperformed other models in the majority of the markets, demonstrating superior predictive accuracy. In a few markets, RF exhibited comparable performance to GRNN, while SVR, GBM, and ARIMA models showed relatively lower accuracy across most cases.The findings emphasize the efficacy of machine learning techniques, particularly GRNN, in modeling complex time series data in the agricultural domain. These models offer robust tools for farmers, market planners, and policymakers to make informed decisions regarding market timing and location, thus optimizing profit margins and reducing market risks.\u003c/p\u003e\u003cp\u003eAnother study conducted by Jaiswal et al. (\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) wherein they presented a hybrid forecasting model combining Recurrent Neural Networks (RNN) and Long Short-Term Memory (LSTM) networks to predict the prices of key agricultural commodities such as wheat, maize, and other crops. The RNN-LSTM architecture leverages the temporal dependencies in time series data, which is particularly important in agriculture where prices are influenced by seasonal patterns, weather events, and cyclical market behaviour. The study highlights the effectiveness of such deep learning models in regions characterized by high seasonal variability, where price fluctuations are more pronounced and unpredictable. By incorporating historical trends, seasonal cycles, and potentially exogenous variables (e.g., rainfall, temperature, policy changes), the hybrid RNN-LSTM model can provide granular insights into future market conditions. These predictive capabilities empower farmers with timely information, enabling them to make more informed marketing and production decisions—such as when to harvest, where to sell, or whether to store produce for a better price.\u003c/p\u003e\u003cp\u003eAnother logistic related study was carried out by Zhang et al. (\u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) wherein they presented a robust machine learning-based framework for evaluating logistic risks in the transportation of perishable agricultural commodities. Focusing on fresh produce, with strawberries as a case study, they built comprehensive risk assessment system grounded in five principal dimensions namly, technological, biological, sustainability, environmental, and emergency risks. A total of 30 quantitative and qualitative indicators are identified to capture the multidimensional nature of these risk factors. To address the complexities of high-dimensional, small-sample data typically encountered in fresh produce logistics, they employed a Support Vector Machine (SVM) algorithm due to its strong generalization capabilities and resilience against overfitting. The model utilizes historical monitoring data from cold chain logistics, integrating variables such as temperature, humidity, vibration, gas composition, packaging integrity, and emergency response capacity. Following normalization and parameter optimization using a radial basis function (RBF) kernel, the model is trained and validated on 350 sample datasets, achieving a predictive accuracy between 85% and 97.5%.The findings emphasise the efficacy of the SVM-based model in real-time risk classification and early warning. The proposed methodology offers significant practical implications for logistics managers, enabling data-driven decision-making to mitigate spoilage, enhance food safety, reduce economic losses, and support sustainability targets in agri-food supply chains.\u003c/p\u003e\u003cp\u003eOne significant challenge in supply chain is monitoring and checking of storage facilities. In that matter, Afreen and Bajwa (\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) proposed a cost-effective, intelligent system RT-IMNS for monitoring and managing cold storage environments where perishable fruits and vegetables are kept using Internet of Things (IoT) and Artificial Neural Networks (ANN). The system overcomes the drawbacks of conventional systems that just monitor temperature and humidity by constantly measuring temperature, humidity, CO₂ concentration, and light intensity. The ANN model classifies commodity status into three categories: good, unsatisfactory, and alarming. It achieves 99% prediction accuracy, outperforming other models like SVM, Naïve Bayes, Random Forest, and Decision Trees. Real-time data is collected through multiple sensor nodes and sent to a cloud-based Firebase database. The system also integrates an Android application for remote monitoring and instant notification, enabling timely actions to prevent spoilage allowing personnel to remotely monitor environmental parameters and receive real-time alerts when any parameter exceeds critical thresholds. This feature reduces the need for constant physical presence in cold storage facilities, and supports better decision-making in managing storage conditions. The system was validated using a real-world experimental setup involving cold storage of potatoes, with over 5,361 environmental data instances collected over 15 days. This low-cost, user-friendly solution can be valuable for small- and medium-sized enterprises, contributing significantly to reducing food losses in the storage and post-harvest lifecycle.\u003c/p\u003e\u003cp\u003ePerishable goods such as fruits, vegetables, dairy, and fermented products are highly sensitive to environmental conditions and logistical handling, making them prone to spoilage, contamination, and economic losses. Under this circumstances, IoT-based environmental sensing, and machine learning can aid in real-time monitoring, direction detection, and data-driven decision-making, thereby reducing waste, ensuring food safety, and strengthening consumer confidence across the entire value chain. In this direction, Alfian et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2020\u003c/span\u003e) presented a smart, integrated traceability system for the perishable food supply chain (PFSC), combining RFID technology, IoT sensors, and a machine learning-based direction detection model (XGBoost). A key innovation is integrating machine learning classifier (XGBoost) into RFID gates to identify the directional movement of tagged products, whether they are entering (receiving) or leaving (shipping) cold storage, addressing a major limitation in conventional RFID systems, which cannot inherently distinguish movement direction. The model uses features extracted from received signal strength (RSS) and timestamp data, and achieves superior performance compared to traditional models with accuracy. The solution was validated in a real-world kimchi supply chain, demonstrating its capability to monitor and present complete product traceability covering movement, location, and environmental conditions across producers, transporters, and distributors. The system effectively filters out false-positive tag readings (e.g., from static or turn-back tags), improving inventory accurac\u003cb\u003ey\u003c/b\u003e and decision-making reliability.\u003c/p\u003e\u003cp\u003eTo increase agricultural productivity and sustainability, it is becoming more and more important to integrate artificial intelligence (AI), the Internet of Things (IoT), and remote sensing (RS). Araújo et al. (\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) carried out a thorough investigation of current machine learning (ML) applications in agriculture within this framework. In order to facilitate the transition from traditional to data-driven agricultural systems, their studies show how machine learning (ML) may optimize crop management, water use, soil monitoring, and livestock oversight. They noted that the research has been conducted on wide range of crops from several different agricultural fields while industrial and plantation crops, cereal crops, vegetable crops, perennial trees, orchards, and vineyards, as well as fruit-bearing trees. The research carried out was mainly associated with crop quality, crop mapping, yield, crop disease, pest/weed detection \u003cem\u003eetc.\u003c/em\u003e, (Fig.\u0026nbsp;5).\u003c/p\u003e\u003cp\u003e \u003cb\u003eChallenges and Future Prospects\u003c/b\u003e \u003c/p\u003e\u003cp\u003eAlthough the Internet of Things (IoT) and artificial intelligence (AI) hold great potential for improving precision agriculture, a number of significant challenges need to be addressed before they can be widely utilized. Such as privacy and security issues with the collecting, exchanging, and archiving of sensitive agricultural data (Farooq et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). Smooth integration of AI platforms and IoT devices may be complicated by a lack of standardization and integration between them. Enabling data exchange and collaboration requires the development of common standards and protocols IoT devices and AI technologies can be expensive, which can hinder adoption, especially for smallholder farmers in developing nations. To guarantee that the advantages of precision farming are widely distributed, efforts must be made to lower expenses and improve accessibility (Van Klompenburg et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e \u003ch2\u003eConflicts of Interest:\u003c/h2\u003e \u003cp\u003eThe authors declare no conflict of interest.\u003c/p\u003e \u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eDhara Prajapati gathered and reviewed the relevant literature, organized the data, and helped shape the initial draft. Abishek Murugesan interpreted the findings, and refined the scientific content. Rucha Dave provided guidance on the research framework, verified the technical accuracy, and reviewed the final version of the paper.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAbioye, E. A., Abidin, M. S. Z., Mahmud, M. S. A., Buyamin, S., Ishak, M. H. I., Abd Rahman, M. K. I., \u0026amp; Ramli, M. S. A. (2020). A review on monitoring and advanced control strategies for precision irrigation. \u003cem\u003eComputers and Electronics in Agriculture\u003c/em\u003e, \u003cem\u003e173\u003c/em\u003e, 105441.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAfreen, H., \u0026amp; Bajwa, I. S. (2021). An IoT-Based Real-Time Intelligent Monitoring and Notification System of Cold Storage. \u003cem\u003eIeee Access : Practical Innovations, Open Solutions\u003c/em\u003e, \u003cem\u003e9\u003c/em\u003e, 38236\u0026ndash;38253.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAgrawal, H., Dubey, A., Tiwari, N., Pandey, D. K., Putrevu, D., Misra, A., \u0026amp; Kumar, R. (2019). Performance evaluation of a newly in-house developed in-situ soil moisture sensor with standard industrial sensors and gravimetric sampling. \u003cem\u003eJournal of Geomatics\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(2), 280\u0026ndash;284.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAl Mamun, M. R., Ahmed, A. K., Upoma, S. M., Haque, M. M., \u0026amp; Ashik-E-Rabbani, M. (2025). IoT-enabled solar-powered smart irrigation for precision agriculture. \u003cem\u003eSmart Agricultural Technology\u003c/em\u003e, \u003cem\u003e10\u003c/em\u003e, 100773.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlfian, G., Syafrudin, M., Farooq, U., Ma\u0026rsquo;arif, M. R., Syaekhoni, M. A., Fitriyani, N. L., Lee, J., \u0026amp; Rhee, J. (2020). Improving efficiency of RFID-based traceability system for perishable food by utilizing IoT sensors and machine learning model. \u003cem\u003eFood Control\u003c/em\u003e, \u003cem\u003e110\u003c/em\u003e, 107016.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAra\u0026uacute;jo, S. O., Peres, R. S., Ramalho, J. C., Lidon, F., \u0026amp; Barata, J. (2023). Machine learning applications in agriculture: current trends, challenges, and future perspectives. \u003cem\u003eAgronomy\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(12), 2976.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAssimakopoulos, F., Vassilakis, C., Margaris, D., Kotis, K., \u0026amp; Spiliotopoulos, D. (2024). Artificial intelligence tools for the agriculture value chain: Status and prospects. \u003cem\u003eElectronics\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(22), 4362.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAstanakulov, K., Abdillaev, T., Umirov, A., Fozilov, G., \u0026amp; Hatamov, B. (2021). Monitoring of the combine with smart devices in soybean harvesting. In E3S Web of Conferences (Vol. 227, p. 07003). EDP Sciences.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAstanakulov, K., Shovazov, K., Borotov, A., Turdibekov, A., \u0026amp; Ibrokhimov, S. (2021). Wheat harvesting by combine with GPS receiver and grain sensor. In E3S Web of Conferences (Vol. 227, p. 07001). EDP Sciences.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBellosta-Diest, A., Campo-Besc\u0026oacute;s, M. \u0026Aacute;., Zapater\u0026iacute;a-Miranda, J., Casal\u0026iacute;, J., \u0026amp; Arregui, L. M. (2022). Evaluation of nitrate soil probes for a more sustainable agriculture. \u003cem\u003eSensors (Basel, Switzerland)\u003c/em\u003e, \u003cem\u003e22\u003c/em\u003e(23), 9288.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBitella, G., Rossi, R., Bochicchio, R., Perniola, M., \u0026amp; Amato, M. (2014). A novel low-cost open-hardware platform for monitoring soil water content and multiple soil-air-vegetation parameters. \u003cem\u003eMdpi Sensors J\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e, 19639\u0026ndash;19659. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s141019639\u003c/span\u003e\u003cspan address=\"10.3390/s141019639\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(Crossref.).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBousbih, S., Zribi, M., Pelletier, C., Gorrab, A., Lili-Chabaane, Z., Baghdadi, N., Ben Aissa, N., \u0026amp; Mougenot, B. (2019). Soil texture estimation using radar and optical data from Sentinel-1 and Sentinel-2. \u003cem\u003eRemote Sens\u003c/em\u003e, \u003cem\u003e11\u003c/em\u003e, 1520.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCapraro, F., Tosetti, S., Rossomando, F., \u0026amp; Mut, V. (2018). Web-based system for the remote monitoring and management of precision irrigation: a case study in. \u003cem\u003eSensors MDPI\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/s18113847\u003c/span\u003e\u003cspan address=\"10.3390/s18113847\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(Crossref.).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChauhan, B. S. (2020). Grand Challenges in Weed Management. In Frontiers in Agronomy; Frontiers Media SA: Lausanne, Switzerland, ; Volume 1.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen, L. B., Huang, X. R., \u0026amp; Chen, W. H. (2023). Design and Implementation of an Artificial Intelligence of Things-Based Autonomous Mobile Robot System for Pitaya Harvesting. \u003cem\u003eIeee Sensors Journal\u003c/em\u003e, \u003cem\u003e23\u003c/em\u003e, 13220\u0026ndash;13235.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDave, R., Kumar, G., Pandey, D. K., Khan, A., \u0026amp; Bhattacharya, B. (2019). : Evaluation of Modified Dubois Model for estimating surface soil moisture using dual polarization RISAT-1 C-band SAR data, Geo-carto International, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/10106049.2019.1655801\u003c/span\u003e\u003cspan address=\"10.1080/10106049.2019.1655801\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDave, R. B., Saha, K., Kushwaha, A., Vithalpura, M., Nidhin, P., \u0026amp; Murugesan, A. (2023). Analysing the potential of polarimetric decomposition parameters of Sentinel\u0026ndash;1 dual-pol SAR data for estimation of rice crop biophysical parameters. \u003cem\u003eJournal of Agrometeorology\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(1), 105\u0026ndash;112.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDhiman, G., Bhattacharya, J., \u0026amp; Roy, S. (2023). Soil textures and nutrients estimation using remote sensing data in north india-Punjab region. \u003cem\u003eProcedia computer science\u003c/em\u003e, \u003cem\u003e218\u003c/em\u003e, 2041\u0026ndash;2048.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eEsposito, M., Crimaldi, M., \u0026amp; Cirillo, V. (2021). Drone and sensor technology for sustainable weed management: A review. \u003cem\u003eChem Biol Technol Agric\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e, 18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFan, X., Nan, Z., Ma, Y., Zhang, Y., \u0026amp; Han, F. (2021). Research on the Spatio-Temporal Impacts of Environmental Factors on the Fresh Agricultural Product Supply Chain and the Spatial Differentiation Issue\u0026mdash;An Empirical Research on 31 Chinese Provinces. \u003cem\u003eInternational Journal Of Environmental Research And Public Health\u003c/em\u003e, \u003cem\u003e18\u003c/em\u003e, 12141.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFarooq, M. S., Riaz, S., Abid, A., Abid, K., \u0026amp; Naeem, M. A. (2019). A Survey on the Role of IoT in Agriculture for the Implementation of Smart Farming. \u003cem\u003eIeee Access\u003c/em\u003e, \u003cem\u003e7\u003c/em\u003e, 156237\u0026ndash;156271.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFood and Agricultural Organisation (2009). Expert meeting on How to Feed the World in 2050. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.fao.org/4/ak542e/ak542e00.htm\u003c/span\u003e\u003cspan address=\"https://www.fao.org/4/ak542e/ak542e00.htm\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGnanavel, I. (2015). Eco-Friendly Weed Control Options for Sustainable Agriculture. \u003cem\u003eSci Int\u003c/em\u003e, \u003cem\u003e3\u003c/em\u003e, 37\u0026ndash;47.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHarishankar, S., Kumar, R. S., Sudharsan, K., Vignesh, U., \u0026amp; Viveknath, T. (2014). Solar powered smart irrigation system. \u003cem\u003eAdv Electr Comput Eng\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e, 341\u0026ndash;346.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHassan, M., Malhotra, K., \u0026amp; Firdaus, M. (2022). Application of Artificial Intelligence in IoT Security for Crop Yield Prediction. \u003cem\u003eRes Rev Sci Technol\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e, 136\u0026ndash;157.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHou, M., Tian, F., Ortega-Farias, S., Riveros-Burgos, C., Zhang, T., \u0026amp; Lin, A. (2021). Estimation of crop transpiration and its scale effect based on ground and UAV thermal infrared remote sensing images. \u003cem\u003eEuropean Journal of Agronomy\u003c/em\u003e, \u003cem\u003e131\u003c/em\u003e, 126389.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eImplementation of Automatic Irrigation Control System \u003cem\u003eIOSR J Comput Eng (IOSR- JCE)\u003c/em\u003e 17 (4), 99\u0026ndash;111. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.9790/0661-174299111\u003c/span\u003e\u003cspan address=\"10.9790/0661-174299111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (Crossref.).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJain, S., Bhujel, S., Shrivastava, U., Mohapatra, A., \u0026amp; Mishra, G. (2023). Advancements in drone technology for fruit crop management: a comprehensive review. \u003cem\u003eInt J Envir Clim Change\u003c/em\u003e, \u003cem\u003e13\u003c/em\u003e(11), 4367\u0026ndash;4378.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJaiswal, R., Jha, G. K., Kumar, R. R., \u0026amp; Choudhary, K. (2022). Deep Long Short-Term Memory Based Model for Agricultural Price Forecasting. \u003cem\u003eNeural Computing And Applications\u003c/em\u003e, \u003cem\u003e34\u003c/em\u003e, 4661\u0026ndash;4676.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJeong, H. Y., \u0026amp; Na, I. S. (2024). Efficient faba bean leaf disease identification through smart detection using deep convolutional neural networks. \u003cem\u003eLegume Research\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(8), 1404\u0026ndash;1411.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhanal, S., Kc, K., Fulton, J. P., Shearer, S., \u0026amp; Ozkan, E. (2020). Remote sensing in agriculture\u0026mdash;accomplishments, limitations, and opportunities. \u003cem\u003eRemote Sensing\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(22), 3783.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKoirala, A., Walsh, K. B., Wang, Z., \u0026amp; McCarthy, C. (2019). Deep learning for real-time fruit detection and orchard fruit load estimation: benchmarking of \u0026lsquo;MangoYOLO\u0026rsquo;. \u003cem\u003ePrecis Agric\u003c/em\u003e, \u003cem\u003e20\u003c/em\u003e(6), 1107\u0026ndash;1135.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKolhalkar, N. R., Krishnan, V. L., Somkuwar, R. G., \u0026amp; Pandit, A. A. (2022). Design of IoT-based mechatronic module for real-time yield monitoring during harvesting of greenhouse produce and vineyards, Int. Rev. Mech. Eng. 16 (2022) 48\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKulkarni, S. S., Bajwa, S. G., \u0026amp; Huitink, G. (2010). Investigation of the effects of soil compaction in cotton. \u003cem\u003eAm Soc Agric Biol Eng\u003c/em\u003e, \u003cem\u003e53\u003c/em\u003e, 667\u0026ndash;674.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKumar, J., Vashisth, A., Sehgal, V. K., \u0026amp; Gupta, V. K. (2013). Assessment of aphid infestation in mustard by hyperspectral remote sensing. \u003cem\u003eJournal of the Indian Society of Remote Sensing\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e, 83\u0026ndash;90.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKushwaha, A., Das, A., Dave, R., \u0026amp; Bhattacharya, B. K. (2022). A non-destructive estimation of chlorophyll-a and-b over different crops using airborne imaging spectroscopy observations. \u003cem\u003eAdvances in Space Research\u003c/em\u003e, \u003cem\u003e73\u003c/em\u003e(2), 1290\u0026ndash;1303.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMarco, E., Mariano, C., Valerio, C., Fabrizio, S., \u0026amp; Albino, M. (2021). Drone and sensor technology for sustainable weed management: a review. \u003cem\u003eChem Biol Technol in Agricult\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e, 18. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40538-021-00217-8\u003c/span\u003e\u003cspan address=\"10.1186/s40538-021-00217-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMateen, A., \u0026amp; Zhu, Q. (2019). Weed detection in wheat crop using UAV for precision agriculture. \u003cem\u003ePakistan Journal of Agricultural Sciences\u003c/em\u003e, \u003cem\u003e56\u003c/em\u003e(3), 775\u0026ndash;784.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMishra, S., Nidhin, P., Dave, R., Pandey, P., Kumari, P., \u0026amp; May (2023). Assessment of Okra Yellow Vein Mosaic Virus using Ground Based Hyperspectral Remote Sensing (Poster presentation). ISMPP 42nd annual Conference \u0026amp; National Symposium on Plant Health Management: A Way Forward for Food Safety, Security and Sustainability. Anand Agricultural University, Anand.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMonteiro, A., \u0026amp; Santos, S. (2022). Sustainable approach to weed management: The role of precision weed management. \u003cem\u003eAgronomy\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 118.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoreno, H., \u0026amp; Andujar, D. (2023). Proximal sensing for geometric characterization of vines: a review of the latest advances. \u003cem\u003eComputers And Electronics In Agriculture\u003c/em\u003e, \u003cem\u003e210\u003c/em\u003e, 107901.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMurugesan, A., Dave, R., Kushwaha, A., Pandey, D. K., \u0026amp; Saha, K. (2023). Surface soil moisture estimation in bare agricultural soil using modified Dubois model for Sentinel-1 C-band SAR data. \u003cem\u003eJournal of Agrometeorology\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(4), 517\u0026ndash;524.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNagasubramanian, G., Sakthivel, R. K., Patan, R., Sankayya, M., Daneshmand, M., \u0026amp; Gandomi, A. H. (2021). Ensemble classification and IoT-based pattern recognition for crop disease monitoring system. \u003cem\u003eIEEE Internet of Things Journal\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e(16), 12847\u0026ndash;12854.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eNock, C. A., Vogt, R. J., \u0026amp; Beisner, B. E. (2016). In eLS (Ed.), \u003cem\u003eFunctional Traits\u003c/em\u003e. John Wiley \u0026amp; Sons, Ltd. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1002/9780470015902.a0026282\u003c/span\u003e\u003cspan address=\"10.1002/9780470015902.a0026282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOborkhale, L. I., Abioye, A. E., Egonwa, B. O., \u0026amp; Olalekan, T. A. (2015). Design and.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOborkhale, L. I., Abioye, A. E., Egonwa, B. O., \u0026amp; Olalekan, T. A. (2015). Design and Implementation of Automatic Irrigation Control System. \u003cem\u003eIOSR J Comput Eng (IOSR-JCE)\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(4), 99\u0026ndash;111. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.9790/0661-174299111\u003c/span\u003e\u003cspan address=\"10.9790/0661-174299111\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e(Crossref.).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOchildiev, O., Fozilov, G., Achildiev, S., Karimov, M., \u0026amp; Ashurov, N. (2021). Indicators of the combine equipped with GPS receiver in sunflower harvesting. In E3S Web of Conferences (Vol. 227, p. 07002). EDP Sciences.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eOerke, E. C. (2020). Remote sensing of diseases. \u003cem\u003eAnnual review of phytopathology\u003c/em\u003e, \u003cem\u003e58\u003c/em\u003e(1), 225\u0026ndash;252.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePandey, D. K., Srivastava, P. K., Dave, R., Setia, R. K., Sinha, R., Sekhar, M., \u0026amp; Kumar, R. (2024). Operational 500 m surface soil moisture product using EOS-04 C-band SAR over Indian agricultural croplands. \u003cem\u003eCurrent Science (00113891)\u003c/em\u003e, \u003cem\u003e126\u003c/em\u003e(9).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePartel, V., Kakarla, S. C., \u0026amp; Ampatzidis, Y. (2019). Development and Evaluation of a Low-Cost and Smart Technology for Precision Weed Management Utilizing Artificial Intelligence. \u003cem\u003eComputers \u0026amp; Electrical Engineering\u003c/em\u003e, \u003cem\u003e157\u003c/em\u003e, 339\u0026ndash;350. researchgate.net\u0026thinsp;+\u0026thinsp;1ar.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePatil, S., \u0026amp; Thorat, S. (2016). Early detection of grapes diseases using machine Learning and \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1109/CCIP.2016.7802887\u003c/span\u003e\u003cspan address=\"10.1109/CCIP.2016.7802887\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePaul, R. K., Yeasin, M., Kumar, P., Kumar, P., Balasubramanian, M., Roy, H. S., Paul, A. K., \u0026amp; Gupta, A. (2022). Machine Learning Techniques for Forecasting Agricultural Prices: A Case of Brinjal in Odisha, India. \u003cem\u003ePlos One\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e, e0270553.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRucha Dave, G., Kumar, D. K., Pandey, A., \u0026amp; Bimal Bhattacharya. (2019). Evaluation of Modified Dubois Model for estimating surface soil moisture using dual polarization RISAT-1 C-band SAR data. \u003cem\u003eGeocarto International\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1080/10106049.2019.1655801\u003c/span\u003e\u003cspan address=\"10.1080/10106049.2019.1655801\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSassu, A., Psiroukis, V., Bettucci, F., Ghiani, L., Fountas, S., \u0026amp; Gambella, F. (2024). Unmanned aerial system plant protection products spraying performance evaluation on a vineyard. \u003cem\u003ePrecis Agric\u003c/em\u003e, 1\u0026ndash;31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11119-024-101558\u003c/span\u003e\u003cspan address=\"10.1007/s11119-024-101558\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaxena, K., Jakhete, M. D., Kumari, P. L., Jain, M., Mane, A., \u0026amp; Karthik, H. P. (2023). Optimizing Agricultural Supply Chains with Machine Learning Algorithms. Journal of Advanced Zoology,44, S-2, Pp.3146:3156.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSharma, K., \u0026amp; Shivandu, S. K. (2024). Integrating artificial intelligence and internet of things (IoT) for enhanced crop monitoring and management in precision agriculture. \u003cem\u003eSensors International\u003c/em\u003e, 100292.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingh, A., Kumar, S., Barman, Aparna, P. M., Mohammed Hisham, M., Sabarinathan Babu, N., Satheeshkumar, M., \u0026amp; Ramasamy, and Thiruvengadam K (2025). Integration of AI and IoT for Yield Optimization in Precision Farming. \u003cem\u003eJournal of Experimental Agriculture International\u003c/em\u003e, \u003cem\u003e47\u003c/em\u003e(3), 233\u0026ndash;241.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSingha, C., Swain, K. C., Sahoo, S., \u0026amp; Govind, A. (2023). Prediction of soil nutrients through PLSR and SVMR models by VIs-NIR reflectance spectroscopy. \u003cem\u003eThe Egyptian Journal of Remote Sensing and Space Sciences\u003c/em\u003e, \u003cem\u003e26\u003c/em\u003e(4), 901\u0026ndash;918.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSree, B. A., Dasari, D. D. B., Subramanyam, D., \u0026amp; Chethan, P. (2025). Precision Agriculture Optimization: Integrating Machine Learning for Crop Yield Prediction and Resource Management. Precision Agriculture Optimization: Integrating Machine Learning for Crop Yield Prediction and Resource Management (January 18, 2025).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eTalaat, F. M. (2023). Crop yield prediction algorithm (CYPA) in precision agriculture based on IoT techniques and climate changes. \u003cem\u003eNeural Computing and Applications\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(23), 17281\u0026ndash;17292.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThakur, D., Kumar, Y., Kumar, A., \u0026amp; Singh, P. K. (2019). Applicability of wireless sensor networks in precision agriculture: a review, Wireless Pers. \u003cem\u003eCommun\u003c/em\u003e, \u003cem\u003e107\u003c/em\u003e, 471\u0026ndash;512.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eVan Klompenburg, T., Kassahun, A., \u0026amp; Catal, C. (2020). Crop yield prediction using machine learning: A systematic literature review. \u003cem\u003eComputers and electronics in agriculture\u003c/em\u003e, \u003cem\u003e177\u003c/em\u003e, 105709.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWang, Y. (2023). Agricultural Products Price Prediction Based on Improved RBF Neural Network Model. \u003cem\u003eAppl Artif Intell\u003c/em\u003e, \u003cem\u003e37\u003c/em\u003e, 2204600.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eWanyama, J., Soddo, P., Nakawuka, P., Tumutegyereize, P., Bwambale, E., Oluk, I., \u0026amp; Komakech, A. J. (2023). Development of a solar powered smart irrigation control system Kit. \u003cem\u003eSmart Agricultural Technology\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e, 100273.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhang, G., Li, G., \u0026amp; Peng, J. (2020). Risk assessment and monitoring of green logistics for fresh produce based on a support vector machine. \u003cem\u003eSustainability\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(18), 7569.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eZhou, Z., Majeed, Y., Naranjo, G. D., \u0026amp; Gambacorta, E. M. (2021). Assessment for crop water stress with infrared thermal imagery in precision agriculture: A review and future prospects for deep learning applications. \u003cem\u003eComputers and Electronics in Agriculture\u003c/em\u003e, \u003cem\u003e182\u003c/em\u003e, 106019.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Precision Agriculture (PA), Remote Sensing (RS), Artificial Intelligence (AI), Machine Learning (ML), Internet of Things (IoT)","lastPublishedDoi":"10.21203/rs.3.rs-8220871/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8220871/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe need for innovative and effective agricultural practices is now more than higher due to the growing demand for food worldwide and the strain on land and water resources. In order to make farming more data-driven, focused, and sustainable, precision agriculture (PA) provides a potent solution by utilizing technologies such as remote sensing, artificial intelligence (AI), machine learning (ML), and the Internet of Things (IoT). This paper discusses the various ways in which precision agriculture is revolutionizing each step of agricultural production, from supply chain optimization and pest and disease detection to soil health evaluation and smart irrigation. Based on current research and practical uses, particularly in India, we demonstrate how technologies like satellite images, unmanned aerial vehicles (UAVs), artificial intelligence (AI)-powered sensors, and automated equipment assist farmers in improving decision-making, cutting waste, conserving resources, and increasing output. Even while PA technologies are becoming increasingly popular, issues including excessive costs, a lack of regulations, and limited availability for smallholder farmers still exist. This study emphasizes how important precision agriculture is to creating a farming system that is more robust, effective, and prepared for the future.\u003c/p\u003e","manuscriptTitle":"Machine Learning and Remote Sensing for Soil Moisture and Nutrient Estimation: A Systematic Review and Future Research Roadmap","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-22 10:05:24","doi":"10.21203/rs.3.rs-8220871/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"5231e62a-ecfa-498f-9567-d05bf301217a","owner":[],"postedDate":"December 22nd, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-05T01:08:46+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-22 10:05:24","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8220871","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8220871","identity":"rs-8220871","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}