Journal of Engineering Research and Reports

Real-Time Decision Intelligence in Healthcare Project Delivery Using Adaptive Data Environments

Kelvin E. Rabbles, Oladapo Aiyenitaju — Wed, 01 Jul 2026 00:00:00 +0000

Introduction: Healthcare project delivery increasingly relies on data-intensive systems to manage complex infrastructure, multidisciplinary stakeholders and rapidly evolving clinical requirements. Traditional project reporting approaches are often unable to provide the timely insights required for proactive decision-making.

Aim: This review synthesises current evidence on real-time data architectures, performance impacts and governance mechanisms that support decision intelligence in healthcare project delivery.

Methods: A systematic review was conducted in accordance with PRISMA 2020 guidelines. Literature searches were performed in Scopus, Web of Science and IEEE Xplore databases, covering publications from 2015 to 2025. The SPIDER framework guided the search strategy, focusing on decision intelligence, healthcare infrastructure and real-time analytics. Following screening and eligibility assessment, 12 high-quality studies were selected from an initial pool of 400 records. Methodological quality was evaluated using the Mixed Methods Appraisal Tool (MMAT), and the evidence was synthesised through critical analysis of technical architectures, operational outcomes and governance frameworks.

Results: The findings identify Lambda Architecture as the predominant framework for integrating real-time stream processing with long-term data auditability. Adaptive data environments demonstrated substantial operational benefits, including process efficiency improvements of up to 32%, reductions in medication turnaround times by 26% and enhanced stakeholder coordination through centralised decision-support systems. Large-scale implementations reported significant financial savings and reduced infrastructure costs. Automated governance mechanisms, including machine learning-based data quality assurance and AI-driven compliance monitoring, achieved high levels of data integrity and security compliance, supporting reliable decision-making in complex healthcare environments.

Conclusion: Real-time decision intelligence transforms healthcare project management from a reactive reporting function into a proactive decision-support capability. Adaptive data environments provide a robust foundation for improving project performance, governance and organisational resilience. Despite challenges related to interoperability, organisational readiness and regulatory compliance, these technologies represent a critical pathway towards future-ready healthcare project delivery and infrastructure management.

Nano-silica-Modified Hydraulic Concrete: Mechanisms, Engineering Performance, Durability, and Challenges for Sustainable Water Infrastructure

Yibo Yin — Sat, 04 Jul 2026 00:00:00 +0000

Hydraulic concrete structures—including gravity dams, arch dams, sluice gates, navigation locks, hydropower tunnels, and coastal sea walls—operate under sustained hydrostatic pressure, hydraulic abrasion, and chemically aggressive environments that impose demanding requirements for impermeability, mechanical performance, and long-term durability. Nano-silica (NS), an amorphous silicon dioxide (SiO₂) material with primary particle sizes typically between 5 and 100 nanometres, has attracted substantial research interest as a supplementary cementitious material capable of simultaneously improving concrete strength, densifying microstructure, and substantially reducing water permeability. This article critically reviews the current state of knowledge on NS application in hydraulic concrete, encompassing NS synthesis routes and physicochemical characterisation, effects on fresh concrete workability and setting, cement hydration mechanisms and microstructural development, mechanical performance, impermeability, and durability under hydraulic service conditions. Evidence drawn from the literature consistently shows that NS additions of 1–3% by mass of cementitious binder yield the most favourable outcomes, owing to the dual action of pozzolanic reactivity and nano-filler densification. The review also examines NS in combination with other supplementary cementitious materials and assesses current and emerging field applications. Barriers to wider adoption—principally agglomeration, elevated water demand, cost premium, and absence of hydraulic-specific test protocols—are examined critically. A literature search spanning 2004 to 31 March 2026 was conducted across multiple academic databases. Key research gaps are identified, and priorities for future investigation are proposed with particular focus on mass concrete applications, long-term hydraulic performance monitoring, and life-cycle assessment.

Machine Learning-based Framework for Early Diabetes Risk Classification

Mirali Mammadzade — Fri, 26 Jun 2026 00:00:00 +0000

Aims: This study evaluated the effectiveness of supervised machine learning algorithms for early diabetes risk classification using demographic, behavioural, cardiovascular, and general health-related indicators. It also examined the variables most strongly associated with diabetes occurrence patterns.

Study Design: A quantitative experimental design was used, based on supervised multiclass classification and comparative machine learning evaluation.

Place and Duration of Study: The experimental analysis was conducted using the Diabetes Health Indicators BRFSS2015 dataset between March 2026 and mid-May 2026.

Methodology: The dataset contained demographic, lifestyle, cardiovascular, and general health-related variables associated with diabetes conditions. Before model implementation, duplicate inspection, exploratory data analysis, feature standardisation, and variable consistency evaluation were performed to improve analytical stability. Diabetes status was used as the target variable in a multiclass classification framework. Logistic Regression, K-Nearest Neighbours, Naïve Bayes, and AdaBoost classifiers were implemented using Python-based machine learning libraries. The dataset was divided into training and testing subsets using an 80:20 ratio. Model performance was evaluated using accuracy, precision, recall, F1-score, and ROC-AUC. Feature importance analysis and ROC curve comparison were also performed to assess classification behaviour and variable contribution patterns.

Results: Logistic Regression achieved the highest ROC-AUC value of 0.814 and demonstrated stable discrimination across diabetes categories. AdaBoost achieved the highest accuracy score of 0.847 and produced competitive precision, recall, and F1-score values. K-Nearest Neighbours showed moderate classification capability, whereas Naïve Bayes demonstrated comparatively weaker classification consistency. Feature importance analysis identified HighBP, GenHlth, Age, BMI, CholCheck, and HighChol as influential variables.

Conclusion: The findings indicate that supervised machine learning methods can support early diabetes risk classification. Cardiovascular conditions, obesity-related indicators, and general health variables were important contributors to classification behaviour within the implemented framework.

An Internal Architecture-Based Strategy to Mitigate Delamination and Improve Flexural Capacity of Composite Beam Structures

Cihan Ciftci — Mon, 29 Jun 2026 00:00:00 +0000

Delamination is an important failure mechanism that can limit the flexural response of laminated composite beam structures, particularly when increased section depth raises the demand on interlaminar load transfer. This study investigated an internal architecture-based strategy for reducing visible delamination at first failure and improving the flexural response of additively manufactured composite beams. Six beam configurations were manufactured using nylon as the matrix material and continuous carbon fibre as the reinforcing phase. Two reference beams, with depths of 20 mm and 30 mm, were produced without internal nylon corridors. Four modified beams were produced by introducing single or multiple nylon corridor arrangements within selected carbon-fibre reinforced regions. All specimens had a width of 19 mm and were tested under three-point bending over a 100 mm span using displacement-controlled loading. The observed first failure load and mode, load-displacement response, and relative manufacturing cost were evaluated. For the 20 mm deep specimens, the modified B3 and B5 configurations increased the first failure load by approximately 12% compared with the corresponding reference beam. The B5 specimen also changed the observed first failure mode from delamination to flexural failure without visible delamination. For the 30 mm deep specimens, the B4 and B6 configurations increased the first failure load by approximately 45% and 42%, respectively, and both showed flexural failure without visible delamination at first failure. The modified configurations also reduced the software-estimated manufacturing cost by decreasing the amount of continuous carbon-fibre reinforcement. The results suggest that internal nylon corridor arrangements may improve material efficiency and influence the failure response of continuous carbon-fibre reinforced composite beams. However, the findings should be interpreted as preliminary because one specimen was tested for each configuration.

Functional Quality Evaluation of an Integrated Hospital Information System: A Case Study of a Private Clinic

Djiba Kourouma, Lancinet Saran Damang, Mamadou Aliou Diallo — Mon, 29 Jun 2026 00:00:00 +0000

This study evaluates the functional quality of an integrated hospital information system developed and deployed in a private clinic. The evaluation was based on a multidimensional approach combining functional validation, user satisfaction assessment and operational impact analysis. Functional validation was conducted through twenty-five test scenarios covering the main system modules, namely authentication, patient management, appointment management, consultation management, hospitalisation management and laboratory management. User satisfaction was assessed through a questionnaire administered to fifteen active users representing administrative, clinical and management roles. Operational impact was examined by comparing selected administrative indicators before and after system deployment.

The results showed that twenty-three of the twenty-five functional test scenarios were successfully validated, corresponding to an overall functional compliance rate of 92%. Authentication, patient management, appointment management and hospitalisation management achieved full compliance, whereas consultation management and laboratory management recorded compliance rates of 75%. The user satisfaction assessment involving fifteen active users revealed an overall satisfaction rate of 87%. Operational analysis showed a reduction in average patient record retrieval time from eight minutes to two minutes and a decrease in average monthly report generation time from 180 minutes to 30 minutes.

These findings indicate that the system effectively supports the clinic’s administrative and medical information management activities. However, the study was conducted in a single private clinic with a limited number of participants, which may restrict the generalisability of the results. Future work should include additional healthcare facilities and broader software quality dimensions, such as security, reliability, maintainability and interoperability.

Ethical and Secure Deployment of Generative AI: Balancing Innovation, Data Privacy, and Enterprise Risk Governance

Tue, 30 Jun 2026 00:00:00 +0000

This study examines the ethical and secure deployment of generative artificial intelligence in enterprise environments, with emphasis on innovation, data privacy, and risk governance. It addresses the gap between the rapid organisational adoption of generative AI systems and the slower development of institutional mechanisms for managing ethical, privacy, security, and regulatory risks. The study identifies the absence of a validated, integrated governance instrument that combines ethical, privacy, security, and enterprise risk controls for the specific characteristics of generative systems. A desk-based mixed-methods approach was used, combining systematic literature review, document analysis, thematic synthesis, comparative evaluation of governance frameworks, and Analytic Hierarchy Process weighting. Evidence was drawn from peer-reviewed literature and authoritative international governance instruments. Eight governance dimensions were assessed: scope and coverage, risk classification, data privacy, ethics and accountability, security controls, legal enforceability, enterprise applicability, and adaptability to generative AI. The findings show that existing governance instruments provide useful but fragmented coverage when applied independently. Ethics and accountability emerged as the highest-weighted dimension, followed by data privacy, security controls, and enterprise applicability. The proposed framework integrates five pillars: ethics and accountability, privacy and data governance, security governance, enterprise risk management, and innovation and compliance alignment. The study concludes that responsible enterprise deployment of generative AI requires coordinated, multi-layered governance rather than reliance on isolated ethical, technical, or legal controls.

Analyzing the Response Current of a Series RL Circuit Using Graphic

Ming-Jong Lin — Thu, 02 Jul 2026 00:00:00 +0000

This study presents a MATLAB-based graphical approach for analysing the transient current response of a series resistor-inductor (RL) circuit. It focuses on decomposing the total current into natural-response and forced-response components under specified circuit parameters. Resistance, inductance and the electrical angle of the power-supply voltage are treated as key variables. The complete response-current expression for the first-order circuit equation is derived, and the natural, forced and complete response-current values are calculated and plotted in MATLAB. Rather than relying only on algebraic calculation, the work emphasises the simultaneous display of the natural, forced and complete current curves. Several verification cases demonstrate how changes in supply voltage and angular input affect the resulting current waveforms. An additional protection-relay example illustrates the practical relevance of the method for power-system transient analysis. The results indicate that graphical representation can clarify the behaviour of transient current components in a series RL circuit and support comparison between calculated values and plotted waveforms. The proposed MATLAB procedure offers a simple instructional tool for visualising response-current characteristics and examining the effect of circuit parameters on transient behaviour. The study is intended mainly for educational and explanatory purposes, particularly for readers seeking to understand the relationship between analytical calculation and graphical representation in basic RL circuit analysis.

Entropy Generation and Bejan Number Analysis in Micropolar Fluid Flow with Variable Thermal Conductivity

E. O. Fatunmbi, S. A. Odunlami, O. A. Olaiju, O. P. Durojaye, S. A. Adegbenro — Fri, 03 Jul 2026 00:00:00 +0000

This study examines entropy generation and Bejan number behaviour in steady, two-dimensional, incompressible micropolar fluid flow over a linearly stretching sheet with variable thermal conductivity. The flow is considered in the presence of a transverse magnetic field and a homogeneous porous medium, with viscous dissipation and heat generation/absorption included in the thermal formulation. The governing boundary-layer partial differential equations are transformed into coupled nonlinear ordinary differential equations by applying suitable similarity transformations. The resulting boundary value problem is solved numerically using the shooting technique combined with the fourth-order Runge-Kutta scheme, and the numerical formulation is validated against limiting cases reported in earlier studies. The effects of micropolar coupling, magnetic field strength, variable thermal conductivity, Prandtl number, and Eckert number are examined through velocity, temperature, entropy generation, and Bejan number profiles. The results indicate that increasing the magnetic parameter retards the velocity field while increasing the temperature distribution and entropy generation near the stretching surface. The micropolar parameter enhances microrotational effects and influences the balance between thermal and frictional irreversibilities. Variable thermal conductivity modifies heat diffusion and shows a comparatively limited effect on the overall entropy generation within the considered parameter range. The Bejan number analysis indicates that the relative dominance of heat-transfer and fluid-friction irreversibilities depends strongly on the governing thermal and magnetic parameters. These results provide a basis for assessing thermodynamic losses in micropolar fluid transport systems.

Screening-Level Deterministic Estimates of Fault-Controlled Reservoir Prospects in the Mag Field, Niger Delta Basin

Bello Akpoku Macquen, Francis Omonefe — Fri, 03 Jul 2026 00:00:00 +0000

This study presents an integrated seismic interpretation and deterministic volumetric screening of fault-controlled reservoir prospects in the MAG Field, shallow offshore western Niger Delta Basin, Nigeria. The study reassessed the field’s structural framework, calibrated reservoir markers to seismic reflections, generated time and depth structural maps, and identified additional hydrocarbon opportunities within the limits of the available dataset. The dataset comprised seventeen 2D seismic lines, referenced 3D seismic information, and well data from MAG-01 and MAG-02. Review of the seismic coverage showed that the available 3D data do not extend across the MAG well locations, limiting their direct use for detailed reservoir-scale volumetric evaluation. Consequently, the nearest 3D seismic line was integrated with the 2D seismic framework to provide structural context. Well-to-seismic calibration using MAG-01 produced a fair-to-good tie, with a correlation coefficient of approximately 0.58. A second-order polynomial time-depth relationship derived from MAG-01 was applied for depth conversion. Interpretation indicates that the MAG Field is a NW-SE-trending collapse-crest rollover anticline bounded by a major northern growth fault and a southern antithetic fault. Three reservoir intervals, D-5 sand, E sand, and E-1 sand, were mapped. Deterministic screening identified five reservoir-prospect combinations, with the largest estimates in the E-1 East and E Sand North-East prospects. The results support preliminary reassessment of MAG Field prospectivity but require improved seismic coverage, velocity control, and uncertainty-based volumetric evaluation before prospect maturation.

Strengthening Cybersecurity and Data Governance in Nigeria’s Aviation Industry: Implications for Blockchain, Artificial Intelligence AI, and Big Data Systems

Okeke Chuba Henry, Okoro Prosper Onyekachi — Wed, 08 Jul 2026 00:00:00 +0000

Nigeria’s aviation industry is increasingly adopting digital systems, including blockchain, artificial intelligence and big data analytics, for passenger processing, cargo tracking, predictive maintenance, operational monitoring and regulatory oversight. These technologies may improve efficiency and decision-making, but they also introduce cybersecurity and data governance risks that can affect safety, confidentiality, compliance and operational resilience. This study examined cybersecurity readiness and data governance practices associated with the use of blockchain, artificial intelligence and big data systems in Nigeria’s aviation sector. A mixed-method research design was adopted. Quantitative data were collected from aviation stakeholders using structured questionnaires, while qualitative insights were obtained from key informant interviews. The study focused on regulatory agencies, airport authorities, airline personnel, ICT managers and senior management staff involved in digital aviation systems. Out of 220 questionnaires administered, 198 were completed and returned, representing a 90% response rate. The findings showed a modest level of cybersecurity preparedness among respondents, with a grand mean of 3.00. Data governance practices were also found to be present but not fully mature, with a grand mean of 2.93. The most frequently reported challenges were inadequate cybersecurity financing, lack of experienced cybersecurity personnel, weak regulatory enforcement, legacy information systems and poor collaboration among institutions. Alignment with international cybersecurity standards was also moderate, with a grand mean of 2.92. The study concludes that stronger institutional coordination, improved funding, staff development, periodic cybersecurity assessment and sector-specific data governance frameworks are required to support secure digital transformation in Nigeria’s aviation sector.

Deep Learning Architectural Enhancements for Robust Facial Recognition in Occluded Scenarios

Afolabi Awodeyi, Omolegho A. Ibok, Omafovbe Imonikosaye, Abosede E. Adeoye — Wed, 08 Jul 2026 00:00:00 +0000

Background: Facial recognition systems often experience reduced performance when key facial regions are obscured by masks, sunglasses, scarves or hand overlays.

Aims: The study aims to investigate architectural enhancements to standard Convolutional Neural Network (CNN) models, namely Residual Network (ResNet), Mobile Neural Network (MobileNet) and Inception Network, to improve facial recognition accuracy under occluded conditions such as masks, sunglasses and scarves.

Study Design: An experimental comparative study was conducted to evaluate baseline CNN architectures against architecturally enhanced versions incorporating attention mechanisms, feature fusion layers, dropout and batch normalisation.

Place of Study: Southern Delta University, Ozoro, Delta State, Nigeria.

Methodology: A combined dataset of approximately 25,000 facial images was assembled from benchmark sources, including Labeled Faces in the Wild (LFW) (Huang et al., 2007), the AR face dataset (Martinez & Benavente, 1998) and the CelebFaces Attributes (CelebA) dataset (Liu et al., 2015), supplemented with synthetic occlusion variants (masks, sunglasses, scarves and hand overlays) generated through data augmentation. ResNet-50, MobileNet and Inception backbones were enhanced with self-attention and spatial attention layers and multi-scale feature fusion modules, then trained using the Adam optimiser (learning rate 0.001, batch size 64, up to 120 epochs and early stopping after 10 stagnant epochs), with dropout (0.3-0.5) and batch normalisation applied for regularisation. Performance was assessed using precision, recall, F1-score, False Acceptance Rate (FAR) and False Rejection Rate (FRR), comparing enhanced models against their baseline counterparts.

Results: The enhanced ResNet attained 88.4% precision, 86.7% recall and an 87.5% F1-score, compared with 74.2%, 71.8% and 73.0%, respectively, for the baseline ResNet. MobileNet with feature fusion reached an F1-score of 85.3%, compared with 70.6% for baseline MobileNet. The enhanced Inception model achieved 89.1% precision, 87.4% recall and an 88.2% F1-score. Error analysis showed that the enhanced ResNet reduced FAR to 3.6% from 7.9%, while MobileNet with feature fusion reduced FRR to 5.1% from 10.2%. Overall, enhanced models improved precision, recall and F1-score by 8-12% across occlusion scenarios relative to baseline CNNs.

Conclusion: Integrating attention mechanisms and feature fusion layers with training optimisations such as dropout and batch normalisation substantially strengthens the robustness of CNN-based facial recognition systems under occlusion. These architectural enhancements show potential for biometric authentication and controlled security applications. Deployment in sensitive domains, including law enforcement, should be preceded by comprehensive fairness evaluation, privacy safeguards, legal compliance and human oversight, although further work is needed to assess computational efficiency and real-time adaptability.

Cobalt Adsorption Kinetics in Fixed-Bed Columns: A Hybrid ANN–Yan Approach

Victor Etuk, Otobong Ukoyo, Perpetual Bassey — Thu, 09 Jul 2026 00:00:00 +0000

A hybrid predictive–mechanistic model was developed for continuous Co²⁺ removal from wastewater using fixed-bed adsorption columns packed with cellulose-derived composite adsorbents. The Artificial Neural Network (ANN) model used a feed-forward backpropagation algorithm optimised through a trial-and-error grid search to determine network topology, with an independent validation vector using early-stopping protocols to reduce overfitting. Experimental breakthrough data generated under varying influent concentrations (50–150 mg L⁻¹), flow rates (4–12 mL min⁻¹), bed heights (4–12 cm), contact times, adsorbent dosages, and adsorbent configurations were used to construct the modelling dataset. A total of 1,842 observations obtained from ten breakthrough experiments were analysed and divided into training (70%, n = 1,289) and testing (30%, n = 553) subsets. Artificial Neural Network (ANN) modelling was used to capture nonlinear relationships among operating variables, while Weibull, Yoon–Nelson, and Yan equations were evaluated for breakthrough characterisation. The ANN achieved strong predictive performance on the independent testing dataset (R² = 0.9324, RMSE = 0.2815, MAE = 0.1887), outperforming Ridge Regression, Random Forest, and Support Vector Regression benchmark models. Comparative kinetic analysis demonstrated that the Yan model provided the best representation of breakthrough behaviour, yielding R² values between 0.9931 and 0.9996 and lower error-function values than the Weibull and Yoon–Nelson models. Combining this data-driven model with the regularizing mechanistic boundaries of the Yan model demonstrates a superior predictive accuracy (R² values approaching unity) over isolated classical kinetics. A Yan-derived kinetic correction factor was subsequently integrated with ANN predictions to account for breakthrough progression and adsorption-bed saturation. The resulting ANN–Yan hybrid framework improved predictive performance (R² = 0.9648, RMSE = 0.2141, MAE = 0.1433) while enhancing the physical interpretability of adsorption-performance prediction. Feature-importance analysis identified contact time, influent concentration, and bed height as the dominant variables influencing Co²⁺ removal efficiency. Although further validation using independent datasets and real wastewater systems is required, the proposed model provides a promising tool for adsorption modelling, process optimisation and engineering design support of cobalt-contaminated wastewater treatment systems.

Edge-Intelligence Framework for Binary Predictive Maintenance Diagnosis in Industrial Manipulator Workstations

Samuel David Tommy, Thomas Okechukwu Onah, Ndukwe Okoro Agha — Fri, 10 Jul 2026 00:00:00 +0000

Modern industrial manipulators operating within human-in-the-loop environments generate highly variable, noise-contaminated telemetry that can reduce the reliability of centralised predictive maintenance models. Cloud-based diagnostic architectures are constrained by communication bandwidth limitations, security vulnerabilities, and non-deterministic transmission latencies. To address these challenges, this paper introduces an autonomous, licence-free edge-intelligence framework designed for real-time in situ kinematic fault diagnosis without external computational or cloud dependencies. The studied architecture couples an event-driven Node-RED data orchestration engine with a synchronised, multi-channel sliding-window preprocessing pipeline. This architecture was empirically validated on an active production floor using multivariable telemetry, comprising motor current, axial vibration, and joint temperature, captured over a continuous 250-second sequence encompassing 50 distinct human-driven manual execution cycles. Through preprocessing via an 80%-overlapping sliding-window protocol (W = 10 s, S = 2 s), the framework generated 116 serialised feature windows, comprising 80 windows for training and 36 windows for independent testing. Raw telemetry was compressed into a low-dimensional feature space that isolates structural degradation markers from human-induced operational transients. Evaluated using a block-stratified 5-fold cross-validation scheme to prevent temporal data leakage, the native edge-compiled binary decision tree achieved classification accuracy and precision of 100% on the independent evaluation blocks. By restricting the classifier split depth, the diagnostic logic compiles directly into standard nested C-style conditional loops. Hardware profiling on a target ARM Cortex-M7 microcontroller demonstrated ultra-low-overhead execution, requiring an inference latency of less than 1.2 microseconds and a memory footprint below 8 KB, leaving 99.2% of the on-chip RAM free for core control loops. This framework provides a verifiable blueprint for low-overhead, high-precision hardware-in-the-loop implementation of binary fault isolation within a localised mechatronic workstation environment.

Generative AI and Multimodal Fraud Intelligence for Financial Cybercrime Detection in Digital Banking Platforms

Sat, 11 Jul 2026 00:00:00 +0000

The rapid digitalisation of banking has increased exposure to financial cybercrime, while existing fraud-detection methods remain limited by single-modality data, class imbalance, concept drift, and poor explainability. This study developed and evaluated the Generative AI-Enhanced Multimodal Transformer Framework (GAMT-Fraud), an explainable artificial intelligence model that integrates transactional, behavioural, and network data for fraud detection. The framework combines a multimodal attention transformer, gradient-boosted learning, and variational autoencoder-based anomaly detection, while synthetic minority augmentation addresses data imbalance. Using a quantitative experimental design, the framework was trained and validated on the Institute of Electrical and Electronics Engineers Computational Intelligence Society (IEEE-CIS) and PaySim benchmark datasets through stratified data partitioning, five-fold cross-validation, and bootstrap significance testing. Performance was evaluated using precision, recall, F1-score, area under the ROC curve (AUC), Matthews correlation coefficient, and precision-recall area. Results showed that GAMT-Fraud consistently outperformed conventional machine-learning and deep-learning baselines across both datasets, achieving statistically significant improvements in fraud-detection performance. Shapley-value-based explainability further enhanced transparency and regulatory compliance by providing interpretable decision insights. The study demonstrates that integrating generative AI, sequential learning, and relational analysis within a unified framework can improve fraud-detection effectiveness. It contributes a scalable, explainable, and auditable fraud-intelligence architecture and provides a replicable foundation for future research in multimodal and adversarial financial fraud detection. The framework is presented as an experimental and auditable proof of concept rather than evidence of immediate real-world deployment.

Evaluation and Development of a Linear Regression-based 4G Long Term Evolution Path Loss Model for Rural Wireless Networks in Kurutie, Nigeria

Tue, 14 Jul 2026 00:00:00 +0000

This study evaluated existing empirical path loss models and developed a linear regression-based propagation model for LTE networks operating at 2400 MHz in Kurutie, Delta State, Nigeria, a rural environment with dense vegetation and variable settlement patterns. Field measurements were collected along two routes using the Network Cell Info mobile application under clear weather conditions during two time slots to capture possible signal variations. Measured Reference Signal Received Power values were converted to path loss and compared with predictions from the Free Space Path Loss (FSPL) and Stanford University Interim (SUI) models, together with a site-specific linear regression model. The results showed clear differences between the empirical predictions and measured data. For Route 1, the SUI model overestimated path loss (mean = 140.23 dB; MBE = 33.86 dB), whereas FSPL underestimated attenuation (mean = 86.61 dB; MBE = -19.75 dB). The linear regression model produced lower errors, with RMSE = 14.29 dB, MAE = 9.38 dB, and MBE = -0.01 dB, explaining 29.7% of the variance after adjustment for sample size. However, five-fold cross-validation showed unstable performance (R² = -0.122 to 0.615). For Route 2, distance was not a statistically significant predictor of path loss (p = 0.126; R² = 0.020), indicating the influence of site-specific environmental factors. The findings support the need for local calibration and additional environmental variables in rural LTE path loss modelling.

Adaptive Dual-Transmit Energy Mechanism for Lifetime Optimization in Wireless Sensor Networks: A Simulation Study

Dominic S. Nyitamen, Maryam Amal Abdullahi — Wed, 15 Jul 2026 00:00:00 +0000

Wireless Sensor Networks are constrained by limited node energy, making energy-efficient routing protocols essential for extending network lifetime. This paper presents an Adaptive Dual-Transmit Enhanced LEACH protocol that dynamically adjusts intra-cluster transmission amplification factors based on per-round cluster geometry. The approach addresses the fixed amplification factor used in Enhanced LEACH by computing an adaptive intra-cluster distance threshold from the mean member-to-cluster-head distance and deriving a dynamic amplification reduction factor that increases as clusters become more compact. This enables greater energy savings when network conditions permit. Performance was evaluated through thirty-trial Monte Carlo simulations on a 100-node, 200 × 200 m network. The adaptive mechanism reduced the effective free-space amplification coefficient by 62.97% and achieved cumulative intra-cluster energy savings of 5.589 J. Statistical analysis showed significant improvements over the base protocol, including increases of 8.8% in First Node Death, 2.9% in Last Node Death, 1.5% in throughput, and 1.5% in energy efficiency (212.47 versus 209.28 packets/J). Scalability analysis for network sizes ranging from 50 to 200 nodes further confirmed that the proposed approach introduced no scalability penalties while providing progressively greater benefits at higher node densities. The results indicate that adaptive transmission amplification can enhance network lifetime and energy utilisation without materially increasing protocol complexity, since the adaptive computation adds only a constant-time, per-round arithmetic step to the existing clustering procedure.

Risk Assessment and Analysis for the Operation of a Conveyor Equipped Water Hyacinth Harvester in the Niger Delta Region

A. Clement Idiapho, O. David Maworo, O. Anho Lawrence — Thu, 16 Jul 2026 00:00:00 +0000

Water hyacinth infestation in the Niger Delta obstructs navigation, affects fishing and aquatic-resource use, and creates operational difficulties for mechanical removal. Conveyor-equipped harvesters provide a practical means of collecting the biomass, but their use in brackish, vegetation-dense waterways introduces risks related to loading, stability, mechanical components, electrical systems, and operator safety. This study assessed the operational risks of a conveyor-equipped water hyacinth harvester using Failure Mode and Effects Analysis. Potential failure modes were identified for the storage deck, stability system, cutter assembly, electrical control panel, hull structure, conveyor chain and mesh, propulsion system, geared motor, and operator work area. Severity, occurrence, and detection scores were assigned, and the corresponding Risk Priority Numbers were calculated to rank the identified risks. Storage deck overloading had the highest Risk Priority Number of 300. Cutter blade fracture or exposure recorded 270, while excessive heel or trim and conveyor chain breakage or jamming each recorded 240. The principal contributing factors were biomass overload, uneven loading, corrosion, fatigue, and dense vegetation. The findings indicate that operational risk can be reduced through load monitoring, defined storage limits, controlled loading sequences, improved cutter guarding, corrosion-resistant conveyor components, automatic shutdown systems, and routine inspection. The assessment provides a structured basis for prioritising safety measures for water hyacinth harvesting operations in the Niger Delta.

Adversarial Robustness of Machine Learning-based Fraud Detection Systems: An Empirical Evaluation of Attack Impact and Mitigation in Fintech Environments

Fri, 17 Jul 2026 00:00:00 +0000

This study evaluated the adversarial robustness of machine learning-based fraud detection systems by comparing classifier vulnerability profiles and assessing adversarial training as a mitigation strategy. Using the IEEE-CIS Fraud Detection dataset, comprising 590,540 transactions with a fraud incidence of 3.5%, four classifiers—logistic regression, random forest, gradient boosting, and a feed-forward neural network—were trained under identical preprocessing and class-weighting conditions and then subjected to Fast Gradient Sign Method and Projected Gradient Descent attacks at a perturbation budget of 0.02. Adversarial examples were constructed directly using closed-form and backpropagated gradients for the differentiable classifiers and using a logistic regression surrogate for the non-differentiable ensembles, before adversarial training was applied as a post-attack mitigation stage. Logistic regression proved the most adversarially vulnerable architecture, sustaining a 31.12-percentage-point recall loss under Projected Gradient Descent, while adversarial training subsequently restored its recall from 0.39 to 0.999 at an accuracy cost of 0.10 percentage points. Random forest and gradient boosting were not degraded by the surrogate-based attack, indicating that comparative robustness claims for tree-based ensembles require attack methods suited to their non-differentiable structure rather than transfer-based evaluation alone. Within the scope of this single-dataset evaluation, the findings support the adoption of adversarial training for gradient-based fraud detection models and suggest that robustness claims should be accompanied by disclosure of the attack methodology used to establish them.

Process Simulation and Economic Evaluation of an Ammonia Recovery Plant Using a Counter-current Air Stripping and Absorption Process

Etinyene Uko Ukpong Essien, Ini-Obong S. Asuquo — Sat, 18 Jul 2026 00:00:00 +0000

Ammonia-laden effluent from fertiliser and urea processing facilities poses a significant environmental hazard, causing eutrophication, dissolved-oxygen depletion, and acute aquatic toxicity when discharged untreated. This study presents the process simulation, design, and economic evaluation of an ammonia recovery plant that combines counter-current air stripping with sulphuric-acid absorption to remove ammonia from urea-desorber effluent water while recovering ammonia into a sulphuric-acid absorption stream with potential for ammonium sulphate production. A steady-state flowsheet, comprising a stripping column, two centrifugal compressors, two centrifugal pumps, and an absorption column, was developed in ASPEN HYSYS® 8.6 using the Non-Random Two-Liquid (NRTL) property method. At the simulated operating conditions, the stripping column reduced the ammonia mass fraction in the treated water from 0.0593 to 0.0005, corresponding to a tray efficiency of 90%. Expressed on the same basis as common regulatory limits, this mass fraction corresponds to approximately 500 ppm by mass, which exceeds rather than satisfies the frequently cited 50 ppm ammonia discharge guideline, indicating that further polishing treatment would be needed to meet that standard. The absorption column recovered the stripped ammonia from the rich air stream into 98 wt% sulphuric acid, reducing the ammonia content of the exhaust air to a mass fraction of 0.0004 while recovering ammonia into a sulphuric-acid absorption stream with potential for ammonium sulphate production. A capital and operating cost evaluation, escalated to 2017 values using the Chemical Engineering Plant Cost Index, gave a total capital expenditure of $5.73 million against an annual operating cost of $58.55 million and annual product/avoided-cost revenue of $66.23 million. Discounted cash-flow analysis over a 20-year horizon, at a 15% discount rate, returned a net present value of $53.43 million, an internal rate of return of 58%, and a payback period of 1.6 years. These results indicate that counter-current air stripping coupled with acid absorption is both an environmentally effective and economically attractive route for ammonia recovery from fertiliser-industry effluent, converting a compliance liability into a profitable process unit.