Volume 2 Issue 1 (2026): In Progress

Climate change is one of the most serious environmental issues and immediate worldwide action is essential to safeguard the earth for future generations. This study examines the use of regenerative biochar in conjunction with machine learning to assess the potential of carbon in soil for climate smart agriculture. Biochar is cost effective, practical and environmentally benign and it may be used to efficiently sequester carbon dioxide, methane and nitrous oxide, all of which are significant Green House Gases. It is reasonably stable form of carbon, produced by pyrolizing biomass at both high and moderate temperatures. Biochar has been found to increase agricultural productivity, enhance nutrient and water efficiency and help the environment, in addition to assisting in carbon sequestration, gives a more productive choice for sustainable agriculture. It includes a vast range of applications, including construction materials like concrete and asphalt, innovative carbon-based composites, bioplastics, and even medical applications. The use of new artificial intelligence and machine learning technology contributed substantially to understanding climate change challenges without wasting time or money. This paper extensively covers all the regenerative biochar strategies for carbon sequestration and role of emerging technology in measuring and modelling soil organic carbon in agricultural lands.

• View PDF

Rapid and accurate soil nutrient assessment is critical for precision agriculture. This study presents a portable and intelligent soil nutrient analyzer based on ion-selective electrodes (ISE) for rapid, on-site estimation of potassium, nitrate, and chloride. Unlike image-based or machine learning approaches that rely on indirect inference, the proposed system directly measures ion activity using electrochemical sensing, ensuring higher reliability under field conditions. The device integrates sensing, signal conditioning, self-calibration using polynomial regression, and wireless data transmission for real-time soil health assessment. A total of 546 soil samples collected from diverse agricultural locations in Dhanbad district, India, were used for validation, with measurements compared against standard laboratory methods including flame photometry and UV-Vis (ultraviolet-visible) spectrophotometry. The developed system achieved high correlation coefficients of 0.994 (potassium), 0.933 (nitrate), and 0.946 (chloride). Statistical evaluation using RMSE (root mean square error), measurement uncertainty, and hypothesis testing confirms the robustness of the calibration model. The study highlights the advantages of direct sensing over image-based prediction methods, particularly in terms of accuracy, environmental robustness, and practical deployment. Limitations related to sensor drift, soil heterogeneity, and field conditions are also discussed. The proposed system provides a scalable and cost-effective solution for precision agriculture and real-time soil monitoring.

• View PDF

Hydroponic fodder production offers a climate-resilient solution to chronic green fodder shortages in water-scarce semi-arid regions. This study evaluated an automated automated irrigation–based hydroponic maize fodder system within a low-cost polyhouse (180 m²) at Ananthapuramu, India during Kharif 2020. Three bedding materials control (no bedding), paddy straw, and sorghum stover were compared under manual and automated irrigation. One kilogram of maize grain (₹16 kg⁻¹) produced 2.65–4.40 kg green fodder within seven days, with sorghum stover achieving the highest yield (4.40 kg tray⁻¹) and a 66.04% increase over control. The water requirement was 2 L tray⁻¹ day⁻¹ (14 L per cycle). Automation reduced irrigation labour by 12 man-days monthly (₹6,000 saving month⁻¹) compared to the manual system requiring 12 man-days month⁻¹ Water productivity ranged from 0.19–0.31 kg L⁻¹ across treatments, with sorghum stover achieving the highest efficiency. Economic analysis revealed benefit–cost ratios of 0.37–0.87 per tray at market prices; however, system-level economics incorporating labour savings demonstrated a net annual return of ₹1,09,592, a payback period of 2.28 years, and a return on investment of 43.8%, indicating improved feasibility at the integrated system level. This research establishes that integrating automated irrigation with appropriate bedding materials improves viability under integrated system-level conditions and enhances resource-use efficiency for smallholder dairy systems.

• View PDF

Smart agriculture, driven by the Internet of Things (IoT), has emerged as a transformative approach to enhancing agricultural productivity, resource efficiency, and environmental sustainability. This study provides a comprehensive and integrative review of IoT-based precision agriculture systems, focusing on their technical architecture, performance outcomes, and practical implementation feasibility. The analysis demonstrates that IoT-enabled smart irrigation systems can reduce water consumption by 25–50%, while maintaining or increasing crop yields by 10–20%, depending on environmental and operational conditions. Furthermore, integration of multi-sensor networks, artificial intelligence (AI), and cloud–edge computing significantly improves decision accuracy, reduces input waste (fertilizers and pesticides), and enhances real-time responsiveness to environmental stress. The findings confirm that IoT-based systems provide measurable agronomic, economic, and environmental benefits rather than purely conceptual advantages. A layered technical framework is synthesized to illustrate how sensing, communication, data processing, intelligence, and actuation layers form a cyber-physical agricultural system. In addition, the study evaluates key challenges related to scalability, interoperability, infrastructure readiness, and economic feasibility, highlighting pathways for large-scale deployment. Overall, the results demonstrate that IoT-enabled precision agriculture represents a viable and data-driven solution for improving water-use efficiency, farm profitability, and sustainable food production under increasing climatic and resource constraints.

• View PDF

The identification of cattle breeds is an important aspect of livestock management, particularly for maintaining genetic resources and supporting breeding programs of local cattle. In Indonesia, Aceh cattle represent one of the important indigenous breeds whose identification is commonly conducted through manual observation based on physical characteristics. However, conventional identification methods often depend on human expertise and may lead to inconsistencies or misclassification. Recent advances in artificial intelligence, especially in computer vision technologies, provide new opportunities to develop automated systems for livestock identification. This study aims to develop an image-based classification model to distinguish Aceh cattle from non-Aceh cattle using computer vision techniques. A dataset of cattle images was collected from field documentation and various online sources and categorized into two classes. After the image collection process was completed, image adjustment and augmentation processes followed, resulting in a final dataset of 2,360 images, which were used for model training and testing. The dataset consisted of 800 original images expanded through augmentation techniques and was automatically divided into training and validation datasets using an 80:20 ratio. The classification model was developed using Teachable Machine and evaluated using performance metrics such as accuracy, precision, and recall. The experimental results show that the model achieved an accuracy of 89.6%, with precision and recall values of 89.1% and 90.3%, respectively. The findings demonstrate the feasibility of applying low-code artificial intelligence platforms for indigenous cattle breed classification in digital livestock management systems.

• View PDF