The steno-endemic species Impatiens aquatica, previously known from the Eastern Ghats, is here reported as a range extension to the Nilgiris in the Western Ghats of Tamil Nadu. A detailed description, colour photographs, ecological notes, and conservation status are provided for this potential wild ornamental species.
Power system engineers continually face challenges in coordinating primary-backup relay pairs to protect distributed generation (DGs) in penetrated electric power systems. Accurate directional overcurrent relays (DOCRs) settings provide fault clearance in the minimum operating time, thereby improving system reliability. In this paper, GA, BGA, and Rao 3 algorithms are implemented to optimize DOCRs settings in a DGs-penetrated power system.Under defined constraints, the primary objective is to minimize the operating times of primary relays and optimize coordination between primary-backup (P/B) relay pairs. The penalty method has been used to incorporate these constraints into the fitness function. The employed algorithms are framed to tackle the nonlinear DOCRs coordination optimization problem. The effectiveness of all employed GA, BGA, and Rao-3 optimization methods is evaluated on the IEEE 14-bus test system penetrated with DGs, which is modeled and simulated in ETAP software. The coordination of DOCRs from an optimal perspective is achieved using MATLAB-based optimization algorithms. Nonetheless, Rao-3, a simple, metaphor-free, algorithm-specific, parameter-free optimization algorithm, is an extremely effective and robust solution to the complex coordination problem of DOCRs in DGs-penetrated power systems, outperforming existing genetic and breeder genetic algorithms.
Global demand for natural aggregates has grown as a result of the rapid urbanization and infrastructure development, thus increasing the depletion of the resources. Steel slag, a by-product of industry, is a potential substitute. Unlike many previous studies that investigated partial replacement or relied on supplementary cementitious materials, fibres, or chemical admixtures, the present study evaluates complete replacement of both fine and coarse natural aggregates with steel slag in a plain PPC concrete without any admixtures. M30-grade mix was proportioned and evaluated for workability, mechanical and durability performance. The steel slag-based concrete achieved 32.18 MPa (7 Day) and 41.43 MPa (28 Day), which are above the target mean strength criterion of M30-Grade, and also reduced the material cost by 8.1
The Guryul Ravine at Khonmoh, Srinagar, Jammu and Kashmir, represents one of the most complete and fossiliferous records of the Permian–Triassic (P–Tr) boundary in the Tethyan realm. Recently declared a National Geoheritage Site by the Geological Survey of India, it preserves sedimentary, palaeontological, and geochemical evidence documenting the end-Permian mass extinction and Early Triassic recovery. The succession within the Zewan Formation records marine anoxia and oceanic redox fluctuations, faunal turnover, and isotopic anomalies linked to global volcanism and environmental stress. Protection of this site will ensure continued opportunities for high-resolution stratigraphic, geochemical, and palaeoecological research critical to understanding Earth’s most severe biotic crisis. The Guryul Ravine represents one of the most significant Permian–Triassic boundary archives within the Himalayan Tethyan realm, preserving exceptional records of environmental and biotic change across the end-Permian mass extinction. Its recent designation as a National Geoheritage Site ensures protection of this globally significant archive, fostering research, education, and geoconservation in India’s Himalayan geological heritage framework.
Preparing operators and engineers demands instructional technologies that accelerate proficiency while lowering cost. This article introduces the Augmented-Reality Industrial Robotics Education (AR-IRE) framework, which teaches robot kinematics, safety, and programming using Android tablets without physical cells or immersive VR hardware. AR-IRE projects articulated and Cartesian robot digital twins into learners’ workspaces. We (i) align AR activities with ABET student outcomes, (ii) integrate AR-IRE into a four-week manufacturing module, and (iii) compare it to traditional hardware labs and desktop VR simulators in a randomised controlled study ( n = 48 ). AR-IRE achieved superior knowledge gains ( Δ = 23.2% vs. 17.5% VR vs. 8.1% physical), reduced cognitive load (NASA-TLX: 44.8), and highest usability (SUS: 84.0), indicating that mobile AR can address manufacturing skill gaps at scale.
Artificial intelligence (AI) has emerged as one of the most transformative technologies of the twenty-first century, offering unprecedented opportunities while raising complex technical, ethical, social, and economic challenges. This work examines AI within the broader historical trajectory of technological revolutions, highlighting both its rapid adoption and its capacity to reshape science, healthcare, education, climate research, governance, and industry. It explains the foundations of modern AI, particularly deep learning, emphasizing the opacity of black-box models and the challenges they pose for accountability, transparency, and human oversight. The work synthesizes major concerns surrounding algorithmic bias, privacy violations, misinformation, deepfakes, surveillance, and cybersecurity, alongside the growing risks of labor displacement, widening socioeconomic inequalities, and concentration of technological power. It also discusses the ongoing philosophical debate regarding machine consciousness and free will, arguing that these speculative questions should not distract from more immediate governance and societal concerns. Despite several risks, AI offers transformative potential through accelerated scientific discovery, personalized healthcare, renewable energy optimization, precision manufacturing, adaptive education, and environmental monitoring. This work reviews emerging governance frameworks, including risk-based regulation, explainable AI, transparency mechanisms, and international cooperation, while highlighting the challenges faced by developing nations in implementing effective oversight. It emphasizes that the greatest benefits will arise from effective human–AI collaboration, where computational efficiency complements human creativity, contextual reasoning, and ethical judgment under robust human oversight. Responsible AI requires equitable governance, transparency, sustained human oversight, and accountable innovation to maximize societal benefits while minimizing risks, ensuring technology advances the collective good without deepening inequalities. Artificial intelligence, exemplified by ChatGPT and advanced language models, has the potential to revolutionize healthcare, climate action, education, and innovation. However, it also introduces risks of bias, misinformation, and economic disruption. Ethical deployment, robust governance, and human oversight are essential to ensure AI serves equitable, sustainable, and societal advancement.
A graph G is k-geodetic edge traceable if every edge of G lies in at least one geodesic of length k. The maximum geodesic edge-covering number gec_max(G) is the minimum number of largest fixed-length geodesics that cover all edges of G. We study some properties of k-geodetic edge traceable graphs and establish relations between gec_max(G) and the domination number of a connected graph. We investigate the k-geodetic edge traceability of the complete bipartite graph K_m,n , and provide an algorithm to compute gec_max(K_n,m) . We also study the k-geodetic edge traceability of trees and product graphs, namely the Cartesian, strong, lexicographic, and Corona products, and obtain bounds for gec_max of these product graphs.
Neoplastic diseases affecting the reproductive system of teleost fishes remain among the most enigmatic and least explored phenomena in aquatic pathology. This communication reports the first documented case of a testicular fibroma in the Indian oil sardine (Sardinella longiceps), collected from the Parangipettai coastal waters of southeastern India in June 2025. The specimen exhibited pronounced abdominal enlargement initially suspected to be a gastric tumour. However, necropsy revealed a pale reddish mass confined to the left testis, affecting its edge and measuring approximately 2.25 × 1.6 × 1.3 cm, while the right testis appeared immature. Detailed histopathological and histochemical analyses confirmed a benign stromal neoplasm composed of spindle-shaped fibroblastic cells intricately arranged in interlacing bundles within a collagen-rich matrix. Special staining techniques confirmed abundant collagen deposition, and glycoprotein presence. The tumour was well circumscribed, devoid of metastatic spread, and demonstrated histological features consistent with a fibroma. This unprecedented case extends the current knowledge of gonadal neoplasia in teleosts and raises important questions about environmental carcinogenesis in polluted marine ecosystems. It emphasises the necessity of systematic histopathological surveillance and integrative eco-toxicological assessments to elucidate the ecological determinants of tumour development in wild fish populations.
Two species of marine dinoflagellates belonging to two different genera are newly recorded from India: Phalacroma hindmarchii Murray Whitting, 1899, belonging to the family Oxyphysaceae Sournia, 1984, and Protoperidinium oviforme (Dangeard, 1927) Balech, 1974, belonging to the family Protoperidiniaceae J.P. Bujak E.H. Davies, 1983. The present study documents the first record of Phalacroma hindmarchii from the shallow waters off the Lakshadweep Islands and the coastal waters of Andhra Pradesh and Odisha, India. Protoperidinium oviforme has been newly recorded from the shallow waters off the Lakshadweep Islands and Great Nicobar Island, and the coastal waters of Andhra Pradesh, India. These two species are heterotrophic, free-living, and based on the currently available literature, neither species has been reported as toxin-producing or associated with harmful algal blooms (HABs). A comprehensive morphological description of each species, supported by LM (Light Microscopy) and SEM (Scanning Electron Microscopy) observations, has been provided. The specimens analysed in this study have been deposited with the National Zoological Collection at the Zoological Survey of India (ZSI) headquarters.
Coorg mandarin is an important species of citrus with narrow genetic variability due to the presence of polyembryony. Similarly, avocado is an emerging exotic high value fruit crop exhibits limited variability in India. Mutation breeding is an important technique for inducing wide genetic variations in fruit crops. Hence, present study was formulated to standardize LD50 value in Coorg mandarin and avocado mainly to create wide genetic variability for different traits. In this study, Coorg mandarin seeds were irradiated with different gamma doses viz., 100 Gy, 150 Gy, 200 Gy, 250 Gy and 300 Gy and avocado seeds were treated with 50 Gy, 60 Gy, 75 Gy, 90 Gy and 100 Gy. The results showed significant differences among the treatments for germination percentage, days taken for germination and number of underdeveloped seedlings. The higher doses of gamma radiation showed the high number of underdeveloped seedlings and prolonged germination time. Probit analysis determined LD50 values of 188 Gy for Coorg mandarin and 77 Gy for avocado. These doses suppressed the germination close to the 50
Preserved fish specimens are widely used in ichthyological and fisheries research, yet the extent to which preservatives alter body shape remains underexplored for Himalayan cyprinids. We investigated the influence of formalin fixation followed by ethanol preservation on the truss morphometrics of Schizothorax esocinus, an ecologically and commercially important snow trout from Dal Lake, Kashmir. Thirty truss distances were extracted from digital images of fresh and preserved individuals over 14 weeks. Allometric size correction and principal component analysis revealed consistent shape alterations, particularly in the head and caudal regions, with the first three principal components explaining 66.2
The construction industry’s reliance on primary aggregates depresses the value of recycled materials, leading to their underutilization. This study introduces Steel Fiber Reinforced Recycled Concrete (SFRRC), a novel concrete composite material made from crushed waste concrete aggregates combined with randomly dispersed steel fibers. Laboratory testing demonstrated that steel fibers markedly enhance tensile strength and effectively control crack propagation. Compared to plain recycled aggregate concrete, SFRRC exhibits superior post-cracking behavior and durability potential, owing to improved fiber-matrix bonding and a strengthened interfacial transition zone. The reinforcement mechanism crack bridging, pull-out resistance, and matrix confinement delays crack growth and imparts toughness, ductility, and impact resistance to the concrete matrix.
This research develops an innovative Deep AutoML approach that integrates an expanded Automated Feature Engineering (AutoFE) suite with automated model selection and hyperparameter optimization to improve intrusion detection in SDN environments. The proposed framework sequentially performs data preprocessing with missing-value handling, scaling, and class-imbalance correction via SMOTE, Enhanced AutoFE with Variance Threshold, Mutual Information, Information Gain (LightGBM), Pearson Correlation and SelectKBest, Enhanced AutoML model selection among k-Nearest Neighbors (KNN), Decision Tree, Random Forest, AdaBoost, and XGBoost, each tuned with four optimizers like Grid Search, Hyperopt, Optunity (CASH), and Optuna. Finally, multi-metric evaluation and Model Explainability Score (MES) were calculated to select and rank the best model. The best-performing configurations pair XGBoost with Pearson Correlation, Information Gain (LightGBM), or Mutual Information feature selection, each reaching 99.98
This short communication introduces the Federated Quantum-Inspired Tensor Compression framework (FQTC), a unified architecture coupling adaptive Tucker decomposition with amplitude-encoded variational gating for collaborative learning across bandwidth-constrained IoT swarms. Unlike prior compressed federated schemes that fix ranks a priori, FQTC’s central innovation is a runtime rank-selection policy driven by channel state and energy budget, enabling lossy yet semantically faithful payload transmission. Evaluated on smart-agriculture, structural-health-monitoring, and wearable-biosignal datasets (5 runs each), FQTC delivers a 38
The papaya mealybug, Paracoccus marginatus Williams and Granara de Willink was observed during 2017 for the first time on Jatropha integerrima plantations in Ludhiana, Punjab, India. Its incidence was recorded from 2017 to 2024 on jatropha during October–December. The incidence of this mealybug was also observed in district Hoshiarpur on jatropha during 2023 and 2024. The infestation ranged from 51.40 ± 0.44
India's rapidly expanding urban road infrastructure has made flexible (bituminous) pavement maintenance a critical challenge for municipal and state road agencies. This study develops quadratic regression prediction models for seven flexible pavement distress types: longitudinal cracking, transverse cracking, ravelling, potholes, rutting, edge cracking, and bleeding/flushing as its primary contribution, based on a two-year visual inspection survey of ten road sections in Bhopal, India, at 500-m intervals and three-month observation cycles. Ensemble machine learning models (Random Forest and Gradient Boosting) are deployed exclusively as nonparametric sensitivity validators-not as competing prediction methods-to verify the quadratic functional form adequacy and establish an accuracy ceiling against which regression performance is benchmarked. Quadratic regression achieved R2 values of 0.931–0.974 across all seven distress types, providing closed-form predictive equations directly usable by municipal engineers via standard statistical software. Gradient Boosting attained R2 of 0.951–0.987; the marginal gain of ΔR2 ≤0.02 over regression confirms that the quadratic functional form is correctly specified rather than indicating a genuine machine learning advantage. Feature importance analysis independently confirms time elapsed as the dominant predictor (61–68
The Internet of Things (IoT) is expanding rapidly, raising urgent concerns about secure and efficient communication in resource-constrained environments. This commentary critically examines the integration of chaotic encryption and convolutional neural networks (CNNs) for securing IoT networks, focusing on their respective strengths, limitations, and complementarities. Rather than presenting new experiments, the paper synthesizes recent findings to highlight ongoing challenges, including complex key management, limited robustness against adaptive attacks, and the absence of benchmark datasets tailored for IoT-scale deepfake detection. By identifying these gaps and outlining potential research directions such as lightweight hybrid encryption frameworks, modular system design, and adaptive security models this commentary seeks to guide future innovations in creating robust, responsive, and resource-efficient IoT security solutions.
Fused Deposition Modeling (FDM) is a widely adopted Additive Manufacturing (AM) technique for fabricating flexible polymer components; however, the printed parts often exhibit poor surface finish due to the inherent stair-stepping effect. This study investigates the effectiveness of Liquid Silicone Rubber (LSR) dip coating in improving the surface quality of Thermoplastic Polyurethane (TPU) parts fabricated by FDM. A two-level fractional factorial Design of Experiments (DOE) was employed to evaluate the effects of layer thickness, print orientation, infill density, coating hardness, and curing time on the reduction in average surface roughness (ΔRa). Surface roughness was measured using a contact profilometer before and after coating. The developed regression model demonstrated good predictive capability with an R2 of 93.12
In 5G wireless systems, the predominant feature of device to device (D2D) communication supports data transfer between devices without the need of base stations. To enhance the existing features in AI-enabled systems, we propose an AI-based framework for mobile devices to augment the transmission rate by increasing the bandwidth. The focus of this letter is to employ a strategy which can amplify the potential of 6G to yield effective, robust and pervasive wireless systems resolving the use of AI in the D2D interaction. The case study and simulation result depict that the framework enables high speed data communication effectively between D2D with less time consumption and reasonable time cost. The training time and throughput analysis of the AI framework is performed with the existing methods to justify the effectiveness of the proposed scheme. The framework transforms the UE ultimately into an active self-optimizing nodes and conserves the device energy.
Cloud segmentation is a key preprocessing task in optical remote sensing because cloud cover and shadows obscure land-surface signals and degrade downstream analysis. This paper presents a deep-learning pipeline for pixel-level cloud segmentation in Landsat 8 imagery using a U-Net decoder with a ResNet34 encoder, four-channel input (RGB + near infrared), and a lightweight channel-attention gate inserted at each skip connection. The attention gate selectively amplifies cloud-discriminative feature channels while suppressing background noise, addressing a core challenge in multispectral cloud detection. We evaluate the model across three benchmark datasets: 38-Cloud, 95-Cloud, and SPARCS. On 38-Cloud, the attention-enhanced ResNet-UNet achieves 94.36
