Volume 7 Issue 4 2026

This work presents the design of an ultra-low power Operational Transconductance Amplifier (OTA) in the weak inversion region, desirable for very low voltage applications specifically in the sub-threshold region of operation. The design targets of this work are set according to the circuit characteristics relevant to bio-signal processing, namely ECG, EEG, EMG or EOG signals. Some of the key design techniques include Bulk driven input stages, Current Recycling stages, Current Injection stages, Cascode loads, and a balanced left-to-right circuit symmetry. The simulation results confirm that the OTA achieves a gain of 57dB using only 0.4V voltage supply and a unity gain bandwidth of 3kHz, CMRR of 93dB(at DC), phase margin of 72o, THD < 1% and finally consuming just 6nW of power. These results suggest that the OTA can be utilized as a building block in circuits meant to process bio-signals.

This study examines the influence of blockchain technology adoption, neoliberal economic practices, and the implementation of the Pancasila economic system on inclusive and sustainable economic performance, with digital economic governance, financial inclusion, and institutional trust and social legitimacy serving as mediating mechanisms within the contexts of Indonesia and the Philippines. A quantitative approach was employed, utilizing an explanatory research design and partial least squares structural equation modeling (PLS-SEM). Data were collected through a survey of 400 respondents comprising business actors, digital financial service users, academics, and economic practitioners in both countries who were selected via purposive sampling, yielding 280 usable data samples. The findings reveal that blockchain technology adoption exerts a significant positive effect on digital economic governance, neoliberal economic practices significantly enhance financial inclusion, and the implementation of the Pancasila economic system significantly strengthens institutional trust and social legitimacy. In turn, digital economic governance and institutional trust are shown to significantly influence inclusive and sustainable economic performance, whereas financial inclusion does not demonstrate a significant effect. Among the mediating variables, institutional trust emerges as the most dominant factor in improving inclusive and sustainable economic performance. These findings suggest that inclusive economic development in developing economies is shaped not only by economic liberalization and digital technology but also, more critically, by the quality of governance and the level of public trust in institutions. Theoretically, this study contributes by integrating perspectives from the digital economy, institutional economics, and the Pancasila economic system into a unified conceptual model. From a policy standpoint, the findings underscore the need for governments to strengthen digital governance, enhance institutional quality, and maintain a balance between economic liberalization and social justice to achieve inclusive and sustainable economic development.

The catalytic oxidation kinetics of pentaamminecobalt(III) complexes containing α-hydroxy acid ligands (glycolate, lactate, and mandelate) were investigated using permanganate ion in micellar media. Metal nanoparticles (~20 nm) were synthesized by chemical reduction and characterized by transmission electron microscopy (TEM) and dynamic light scattering (DLS). Oxidation reactions were monitored spectrophotometrically at 525 nm. The reactions followed pseudo-first-order kinetics with respect to permanganate concentration. The catalytic efficiency increased significantly in the presence of nanoparticles and surfactant aggregates such as cetyltrimethylammonium bromide (CTAB) and sodium dodecyl sulfate (SDS). Rate constants were analyzed using Michaelis–Menten kinetics, suggesting saturation-type catalytic behavior. Activation parameters derived from Eyring plots showed negative entropy values indicating formation of an ordered transition state. Furthermore, artificial intelligence models were proposed to predict reaction rates using parameters such as nanoparticle size, surfactant concentration, and ligand environment. The results demonstrate that nanoparticle-mediated micellar catalysis offers an efficient approach to accelerate redox reactions in cobalt coordination systems.

In contemporary organizational environments, managerial decision-making is increasingly characterized by uncertainty, ambiguity, and incomplete information. Traditional decision models based on classical binary logic often fail to capture the complexity of real-world managerial problems, particularly those involving qualitative judgments and linguistic assessments. Fuzzy logic, introduced by Lotfi A. Zadeh, provides an effective mathematical framework for handling vagueness and imprecision by allowing reasoning with degrees of truth rather than absolute values. This study explores the conceptual foundations and managerial applications of fuzzy logic, emphasizing its relevance in decision-making processes across domains such as human resource management, finance, marketing, operations, and strategic management. The paper examines key components including fuzzy sets, membership functions, linguistic variables, and fuzzy inference systems, along with advanced methodologies such as fuzzy multi-criteria decision-making (MCDM) techniques and neuro-fuzzy systems. Furthermore, the research identifies critical gaps in existing literature, particularly in the areas of behavioral integration, adaptive model design, and scalability in data-intensive environments. It proposes a structured framework for incorporating fuzzy logic into managerial decision support systems, highlighting its potential to enhance decision accuracy, flexibility, and realism. The findings suggest that fuzzy logic serves as a vital bridge between human cognitive reasoning and computational intelligence, aligning with the concept of bounded rationality proposed by Herbert A. Simon. By integrating fuzzy logic with emerging technologies such as artificial intelligence and machine learning, organizations can develop more robust, adaptive, and intelligent decisionmaking systems. The study concludes that fuzzy logic is not merely a theoretical construct but a practical and indispensable tool for modern management, offering significant potential for future research in areas such as Industry 4.0, behavioral analytics, and intelligent business systems.

Ambulance delays at traffic junctions represent a critical and preventable threat to emergency response efficiency, with each minute of delay reducing cardiac arrest survival probability by approximately 10%. Existing traffic management systems lack automated priority mechanisms for emergency vehicles, and prior IoT-based approaches suffer from single-point-of-failure risks: GPS blackouts in urban canyons, network outages, and absence of government-accepted manual override. This paper presents ClearWay v2, a resilient, multi-layer smart ambulance alert system that addresses all three gaps through: (i) GPS–IMU sensor fusion using a NEO-6M GPS module and MPU-6050 inertial measurement unit for continuous positioning accuracy even during GPS blackouts of up to 45 seconds; (ii) a five-tier offline fallback architecture (Online → Hybrid → Local-LoRa → Dead-Reckoning → SMS) ensuring uninterrupted operation without cloud or cellular connectivity; (iii) HMAC-SHA256 message authentication with TLS 1.3 encryption to prevent rogue preemption attacks; (iv) a hardwarelevel manual override at every junction controller for unconditional police control; and (v) a structured 90-day field deployment protocol at Namakkal Government Hospital, Tamil Nadu, covering 3 ambulances and 6 signalised junctions on NH-544. SUMO simulation across 300 runs (low/medium/high traffic density) demonstrates an 85.8% reduction in per-junction delay at high traffic density, a 63.1% reduction in end-to-end ambulance travel time, and V2I alert broadcast latency under 2.0 seconds. The enhanced architecture scores 9.1/10 on a seven-dimension real-world viability rubric, an improvement from 7.7 for the baseline design, and is aligned with UN SDG 3, SDG 11, and SDG 13.

A new, broadly applicable exergy formulation is presented, built directly on real gas thermodynamics theory and explicitly incorporating both thermal and caloric imperfections. It overcomes the inherent restrictions of classical perfect gas models and offers a single framework suitable for describing energy exchanges in realistic real gas conditions. The robustness of the formulation is examined numerically over several thermodynamic regimes. For open systems, it is tested and investigated on a supersonic isentropic nozzle, whereas for closed systems, it is applied to an isochoric air tank process and to an isobaric process represented by a piston cylinder setup. Comparisons are made between real gas and both the high temperature HT and the conventional perfect gas PG models show that the new expression predicts real gas behavior and energetics potential with higher accuracy. Results from previous models can differ from the current model by up to 30%.The results demonstrate that it can resolve the combined influence of the thermal and caloric imperfections and temperature-pressure effects on energy transfer and conversion. Together, these findings advance real gas exergy theory and support more realistic, predictive and efficient modeling of

Anita Nair’s Ladies Coupe explores the inner lives of Indian women as they navigate patriarchal constraints and reclaim personal support. Set in a women-only train compartment, the novel presents six interwoven narratives that confront issues of gender roles, familial expectations, sexuality, and selfhood. This article analyzes the coupe as a metaphorical feminist space, examining how personal stories function as acts of resistance. Drawing on feminist theories by Simone de Beauvoir, Virginia Woolf, Gayatri Spivak, and others, the paper argues that Nair’s novel is a powerful meditation on the transformative power of storytelling and solidarity. The article also emphasizes the relevance of the novel’s themes in both Indian and global contexts.

Automated Theorem Proving (ATP) in mathematics leverages computer algorithms to prove mathematical theorems. Python, with its extensive libraries and frameworks, has emerged as a powerful tool in this domain. This paper explores the application of Python in ATP, focusing on two case studies: proving the Pythagorean theorem and verifying properties of prime numbers. The results demonstrate Python's effectiveness and potential in facilitating mathematical proofs.

History: Parkinson Disease is a slowly progressive neurodegenerative condition that is mainly marked by the loss of dopaminergic neurons and aggregation of α-synuclein. Historically viewed as a central nervous system condition, there is an increasing body of evidence to support the role of peripheral systems, especially gut–liver-brain axis, in its pathogenesis. Purpose: The purpose of this review is to clarify the mechanistic action of the gut-liver-brain axis in PD focusing on gut microbiota dysbiosis, hepatotoxicity, and neuroinflammatory mechanisms. Methods: An extensive review of the recent literature was performed in order to combine the results on microbial changes, liver dysfunction and their interplay on neurodegeneration. Findings: Gut microbiota dysbiosis causes augmented intestinal permeability and endotoxin translocation (lipopolysaccharides), triggering systemic and hepatic inflammation. Hepatic dysfunction deteriorates the detoxification mechanisms leading to accumulation of neurotoxic metabolites such as ammonia and distorted bile acids. These factors disrupt blood–brain barrier integrity and activate microglia, promoting neuroinflammation, oxidative stress, and mitochondrial dysfunction. Moreover, mechanistic overlap of hepatic encephalopathy and PD illuminates common pathological processes comprising of neurotransmitter imbalance, astrocyte dysfunction, and dopaminergic deficits. The gut-liver-brain axis is therefore an important integrative pathway in the development of PD. Conclusion: The gut-liver-brain axis is an important pathogenesis contributor of PD outside the brain. Microbiota modulation, hepatoprotective, and anti-inflammatory interventions have shown promising opportunities to target this tri-organ axis to achieve early diagnosis and disease modification.

Diabetes is a rapidly growing chronic disease that poses serious health risks, including heart disease, kidney failure, and vision impairment. Early and accurate prediction is essential for effective prevention and treatment. However, medical datasets often suffer from class imbalance, which reduces the performance of traditional machine learning models, and raise concerns regarding patient data privacy. This paper proposes a privacy-preserving diabetes prediction system that integrates Adaptive Synthetic Sampling (ADASYN) with machine learning techniques to improve classification accuracy. The ADASYN algorithm is employed to generate synthetic samples for the minority class, thereby balancing the dataset and enhancing model performance. A Random Forest classifier is utilized to build the prediction model due to its robustness and efficiency. In addition, a secure authentication mechanism is implemented to ensure the confidentiality of sensitive patient information. The proposed system is evaluated using standard performance metrics such as accuracy, precision, recall, and F1-score. Experimental results demonstrate that the integration of ADASYN significantly improves prediction accuracy compared to traditional approaches. The system provides a reliable, efficient, and privacy-aware solution for early diabetes detection, making it suitable for real-world healthcare applications.

In precision agriculture, the timely and precise identification of plant leaf diseases is of prime importance in reducing crop loss and in optimizing treatment plans. This work introduces a new deep learning architecture specifically developed for the diagnosis of rice leaf diseases based on multi-resolution image features and meta-learning strategies. The suggested system combines dual-resolution convolutional neural networks (CNNs), attention-based feature augmentation, and a model-agnostic meta-learning (MAML) framework to provide resilience to different data and environment types. A dataset of 2,627 rice leaf images from six disease classes Bacterial Leaf Blight, Brown Spot, Leaf Blast, Leaf Scald, Narrow Brown Spot, and Healthy is used. All images are preprocessed with histogram equalization and Gaussian filtering, and resized into two scales of 128×128 for preserving global structures and 512×512 for detailed texture information. These images are separately processed by shallow and deep CNNs to extract complementary feature maps. A Squeeze-and-Excitation (SE) block is incorporated to carry out channel-wise feature recalibration, elevating discriminative capacity with attention-guided feature fusion. In order to counter the problem of limited annotated examples in real-world applications, a MAML-based meta-learning strategy is utilized. The model is learned over synthetic few-shot tasks that are created to mimic various field conditions like occlusions, lighting variations, and partial leaves. This allows the classifier to rapidly adjust to new variations with little more data. The last classification network is trained using traditional supervised learning to achieve optimal performance on actual test sets. A large number of experiments illustrate that the new model much surpasses traditional single-resolution CNNs and baseline classifiers. The model reaches top-classification accuracy, precision, recall, and F1-score, and shows robust resistance under low-data scenarios. Also, adopting Grad-CAM visualization verifies that the model always pays attention to biologically meaningful areas of the leaf surface. The whole pipeline from image loading and preprocessing to feature extraction, classification, and result visualization is packaged in a friendly Graphical User Interface (GUI) coded using Python's Tkinter library. This renders the system usable and feasible for deployment by farmers and agricultural experts.

Ocean pollution from plastic waste and submerged debris poses a severe threat to aquatic ecosystems, marine biodiversity, and overall water quality. Traditional underwater waste detection methods depend on manual diver inspection and surface monitoring, which are time-consuming, costly, and unable to provide continuous realtime coverage of large water bodies. Existing commercial solutions either require expensive underwater hardware, constant internet connectivity, or fail to address the specific visual challenges of turbid underwater environments. This paper presents a Deep Learning-Based Smart Underwater Trash Detection System, an intelligent, costeffective, and field-deployable platform that leverages computer vision to automate the identification and classification of submerged waste materials. The system employs the SMVYOLOv11n architecture — a lightweight yet high-accuracy variant of YOLOv11 enhanced with attention mechanisms — to detect trash categories including plastic bags, bottles, fishing nets, and metal debris in real-time video streams and static images. Data preprocessing techniques including contrast-limited adaptive histogram equalization (CLAHE), underwater color correction, and image augmentation are applied to overcome challenges inherent to underwater imaging such as light attenuation, color distortion, and low visibility. The trained model achieves a trash detection accuracy of 88–95%, processes frames in under one second, and operates entirely on a local computing device without requiring cloud services. A user-friendly interface displays detection results with bounding boxes, confidence scores, and category labels, while automated alerts notify operators of pollution levels. The system is designed for integration with underwater drones and ROVs for scalable, autonomous marine cleanup op

The objective of this study was to evaluate the effect of moisture content on the engineering properties of whole and hulled millets, specifically pearl millet, sorghum, finger and proso varieties. Moisture-dependent changes in physical, mechanical and thermal characteristics were systematically investigated to optimize post-harvest handling and processing. Standardized experimental protocols were employed to determine bulk density, porosity, true density, angle of repose and frictional properties across a controlled range of moisture levels. Results demonstrated that bulk density decreased with increasing moisture, while porosity and true density exhibited corresponding variability, indicating structural changes in grain packing behaviour. Frictional coefficients against common structural surfaces increased with moisture, suggesting greater resistance during conveying and storage. Mechanical property hardness was strongly influenced, with higher moisture reducing brittleness and breakage rates but simultaneously increasing susceptibility to microbial spoilage. Hulled grains consistently showed lower bulk density and higher porosity compared to whole grains, reflecting the impact of husk removal on physical structure. Overall, moisture content emerged as a critical determinant of millet storability, processing efficiency and mechanical integrity. The study concludes that maintaining optimal moisture levels is essential to minimize post-harvest losses, enhance processing performance and support a resilient and sustainable millet supply chain.

Generative artificial intelligence (AI) is reshaping writing practices, assessment models, and learner agency across educational contexts, yet little is known about how AI‑supported assessment functions at early stages of language development. This study investigates AI‑mediated writing development among A1 learners and integrates these findings with higher‑education research to establish a cross‑proficiency understanding of AI’s impact on writing. Using an expanded dataset of 34 learners who produced 204 naturalistic writing samples, a mixed‑methods design was employed combining CEFR‑aligned analytic scoring with AI‑assisted assessment, qualitative error analysis, and cross‑level comparison with over 80 peer‑reviewed studies. Findings reveal a stable developmental asymmetry: learners demonstrate strong communicative intent and topic engagement but persistent weaknesses in linguistic accuracy, cohesion, and mechanics. While AI enhanced scoring consistency and error‑pattern detection, it did not resolve deeper linguistic or rhetorical challenges. The study proposes a cross‑proficiency framework showing that AI reshapes—but does not eliminate—developmental constraints in writing, offering implications for pedagogy, assessment design, and responsible AI integration.

This study examines the relationship between various types of abuse—physical, verbal, sexual, and economic—and the socio-demographic characteristics of women victims. Using a sample of 200 respondents, the study applies Analysis of Variance (ANOVA) to determine significant differences across variables such as age, education, residence, partner occupation, income, and age at marriage. The findings indicate that verbal abuse shows significant variation across multiple socio-demographic factors, while physical and sexual abuse remain largely uniform. Economic abuse is significantly associated with education, occupation, and age at marriage. The study concludes that socio-economic vulnerabilities and structural inequalities play a crucial role in influencing abuse intensity.

The world faces some of the biggest problems, such as food insecurity, environmental degradation, and resource wastage, which can only be solved through sustainable agricultural practices. This has encouraged the green agriculture cooperation between China and Pakistan (ACCP), emphasising the integration of advanced technologies. In this regard, artificial Intelligence (AI) and green practices present a revolutionary approach to advancing sustainability in agriculture. This study proposes to assess the influence of AI capability (AIC), farmer green values (FGV), green innovative intentions (GIN), and energy-use reduction (ERD) towards the "adoption of green production" (AGP). Further, it evaluates the mediation of the sustainability of the supply chain (SCC) on these relationships and the moderation of eco-innovation (ENI) in the association between SCC and green production adoption. A quantitative research approach was applied, and the data was gathered via a survey questionnaire that was administered online through Qualtrics. The sample consisted of 294 farmers/ employees, working under ACCP projects and SEM was used for analysis. The study results showed that FGV, GIN and ERD have a significant impact on AGP (p < 0.05) while AIC was found to have an insignificant association with AGP (p > 0.05). Supply chain sustainability mediates these effects (p < 0.05), and ecoinnovation enhances the influence of sustainability on green production (p < 0.05). The study makes practical contributions by highlighting the need to integrate AI, sustainability practices, and ecoinnovation in agriculture. Thus, policymakers and agricultural firms can use these findings effectively to develop agendas promoting increased green production and international cooperation. The information presented provides a model for other areas interested in enhancing food yield while minimising adverse environmental impact through technology and new advancements.

Automated epileptic seizure detection from electroencephalogram (EEG) signals is essential for reducing the diagnostic burden on neurologists and enabling timely clinical intervention. This paper presents an end-to-end four-class seizure detection framework integrating Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) and a Radial Basis Function Support Vector Machine (RBF-SVM). EEG epochs from the Bangalore EEG Epilepsy Dataset (BEED), comprising 8,000 balanced samples across Normal (Eyes Open), Normal (Eyes Closed), Pre-ictal, and Ictal classes, are decomposed into Intrinsic Mode Functions (IMFs) using CEEMDAN. A 40-dimensional statistical feature vector— encompassing energy, kurtosis, Shannon entropy, standard deviation, log energy, mean absolute value, waveform length, and zero-crossing rate—is extracted from five IMFs per epoch. The proposed CEEMDAN-SVM pipeline achieves a hold-out accuracy of 91.17% with 5-fold cross-validation accuracy of 90.48%, outperforming the EMD-SVM baseline by 14.0 percentage points and surpassing K-Nearest Neighbours (82.5%) and Random Forest (83.1%) classifiers under identical experimental conditions. End-to-end inference latency remains below 300 ms per epoch, demonstrating real-time feasibility. A dual-interface deployment strategy comprising a MATLAB-based research GUI and a browser-based NeuroSense clinical interface further validates the translational readiness of the proposed system. The findings establish CEEMDAN as a robust decomposition stage for practical, low-latency epileptic seizure analytics.