This study presented a novel mathematical functional-based algorithm designed to predict the risks of vehicular crashes by leveraging real-time traffic data collected from urban road networks. The proposed model integrated multiple critical variables, including traffic speed, vehicle density, visibility conditions, spatial coordinates, and time-of-day factors, to generate a comprehensive and dynamic assessment for foreseeing the likelihood of traffic crashes. The flexible functional framework enabled the incorporation of diverse traffic and environmental variables, thereby improving the accuracy and contextual sensitivity of risk predictions for road traffic. The model was calibrated and validated using real-world traffic data from five key locations in Islamabad, Pakistan, known for their varying traffic patterns. The results demonstrated that the model could effectively identify high-risk zones and specific time intervals during the day when the probability of crashes was elevated. For example, areas such as Inter-junction Principal (IJP) Road exhibited significantly higher risks of crashes during peak congestion hours, correlating strongly with increased vehicle density and reduced visibility. The study highlighted the potential of combining mathematical modeling with real-time data analytics to address the growing challenges of traffic safety in rapidly urbanizing cities. By providing spatially and temporally resolved estimations of risks, the proposed method enables urban planners and traffic authorities to implement proactive and targeted safety interventions, such as dynamic traffic signaling, speed regulation, and public awareness campaigns. This approach not only enhances urban traffic management but also contributes to reducing accident rates and improving overall road safety.
Reliable detection of road surface objects under foggy conditions remains a critical challenge for autonomous vehicle perception systems due to the severe degradation of visual information. To address this limitation, a novel framework was developed that integrates entropy-guided visibility enhancement, Pythagorean fuzzy logic, and structure-preserving saliency modeling to improve object detection performance in low-visibility environments. Visibility restoration was achieved through an entropy-guided weighting mechanism that selectively enhances salient image regions while preserving essential structural features critical for downstream detection tasks. Uncertainty and imprecision inherent to fog-degraded scenes were systematically modeled using Pythagorean fuzzy logic, enabling improved confidence estimation and robustness in object localization. A saliency mechanism that preserves structural characteristics further contributes to the accurate delineation of road-relevant elements. Extensive evaluations on multiple publicly available foggy road datasets were conducted, demonstrating substantial gains in detection performance, with notable improvements in accuracy, precision, recall, and F1-score metrics. Furthermore, enhancements in visual quality were verified using structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), natural image quality evaluator (NIQE), and blind/referenceless image spatial quality evaluator (BRISQUE) metrics. The computational efficiency of the proposed method was confirmed, supporting its applicability to near real-time deployment scenarios. Consistent performance was observed across varying fog densities, highlighting the framework’s scalability and generalizability. The integration of entropy-based visibility enhancement with fuzzy reasoning and saliency preservation offers a comprehensive and practical solution to the challenges of perception in visually degraded environments, contributing to the advancement of safe and intelligent transportation systems.
The rural Andean community of Yacubiana, Ecuador, currently lacks an adequate sanitation infrastructure, with domestic wastewater managed through individual septic tanks. These decentralized systems have exhibited significant infiltration issues, resulting in groundwater contamination, degradation of sensitive páramo ecosystems, and adverse public health outcomes. Furthermore, this environmental degradation has impeded the community’s potential for ecotourism-based development. To address these challenges, an integrated wastewater management strategy was developed, grounded in sanitary engineering principles and aligned with conservation priorities. The proposed framework encompassed four sequential phases: (i) a comprehensive analysis of existing data on water and wastewater practices within the community; (ii) a systematic evaluation of sanitation alternatives tailored to the community’s socio-environmental context and the ecological fragility of Andean paramos; (iii) the design of a selected sanitation solution in accordance with national and international technical standards; and (iv) a SWOT (Strengths, Weaknesses, Opportunities, Threats) analysis conducted with both technical experts in water resource management and local community representatives. This participatory evaluation aimed to identify strategic pathways for enhancing environmental stewardship, promoting circular water economies, and enabling sustainable tourism. The recommended intervention consists of a simplified, decentralized sewage collection system linked to a trickling filter-based treatment plant, designed for a hydraulic load of 2.79 L/s. The SWOT analysis revealed substantial institutional and infrastructural constraints, primarily due to limited governmental support; however, it also identified considerable ecotourism potential grounded in the area’s geological, ecological, and cultural assets. When implemented within a conservation-based framework, the proposed system is expected to support compliance with Sustainable Development Goals (SDGs) 3 (Good Health and Well-being), 6 (Clean Water and Sanitation), and 11 (Sustainable Cities and Communities). The methodological approach developed herein offers a replicable model for integrated wastewater management in rural, environmentally sensitive regions, providing a viable foundation for community-led, sustainable socio-economic development.
To facilitate a rigorous evaluation of damage progression in in-service steel frame structures subjected to seismic loading, a seismic damage model that integrates the effects of atmospheric corrosion has been developed. Corrosion-induced deterioration significantly influences the structural integrity of bolted steel frames, yet its impact on seismic performance remains inadequately quantified. In this study, a performance-based seismic damage assessment framework has been established, wherein corrosion-related degradation is incorporated into the structural damage evolution process. Drawing on an extensive review of domestic and international research, a refined damage index classification system has been formulated to characterize varying levels of structural impairment. To validate the proposed model, a seismic collapse simulation was conducted on a 1:4 scaled-down steel frame specimen, enabling a comprehensive analysis of damage accumulation over different service durations. The results confirm that the developed model accurately captures the progressive deterioration and collapse behavior of corroded steel frames under seismic excitation. This study provides a quantitative basis for assessing the post-earthquake residual load-bearing capacity of in-service bolted steel frame structures, offering critical insights for structural resilience evaluation and maintenance planning.
A comprehensive statistical analysis was conducted to investigate the causes and prioritization of failure modes within a production line manufacturing leather covers for automotive interiors. The study was grounded in a Process Failure Mode and Effects Analysis (PFMEA), with a dual emphasis on evaluating the traditional Risk Priority Number (RPN) approach and the more contemporary Action Priority (AP) methodology, which has been increasingly adopted to enhance risk assessment sensitivity. Failure modes were classified and prioritized using both approaches, revealing notable differences in the ranking outcomes. To further elucidate the underlying contributors to these failure modes, causal factors were systematically categorized in accordance with the 5M+1E framework—Man, Machine, Method, Material, Measurement, and Environment—commonly employed in quality and reliability engineering. A cause-and-effect diagram was constructed to visualize the distribution of root causes across these categories. Descriptive statistics and correlation analyses were employed to quantify the relationship between each category and the prioritized failure modes. Particular attention was paid to examining the interdependencies among the core PFMEA parameters—Severity, Occurrence, and Detection—in order to determine their respective contributions to the variability in failure mode rankings. It was found that Severity exerted the most substantial influence on the prioritization outcomes under the AP model, while Occurrence was more dominant when the RPN method was applied. These findings suggest that the choice of prioritization method significantly alters the interpretation of risk and resource allocation for corrective actions. The integration of 5M+1E categorization with PFMEA metrics offers a structured pathway to enhance the diagnostic capability of reliability assessments and improve decision-making in failure prevention strategies. This approach is proposed as a more robust alternative to traditional analysis, enabling more precise targeting of corrective and preventive measures in high-precision manufacturing environments.
An artificial intelligence (AI)-powered agricultural advisory system, termed NDEMRI (Nurturing Digital Extension via Mobile and Responsive Intelligence), has been developed to provide evidence-based farming guidance to rural communities across sub-Saharan Africa through short message service (SMS). Designed for compatibility with basic GSM-enabled mobile phones and independent of internet access for end-users, the system integrates large language models (LLMs) via the ChatGPT API to generate contextually relevant, linguistically localized responses to a wide array of agricultural queries. A quasi-experimental evaluation was conducted in the northern regions of Cameroon over a four-month period, employing a matched control group methodology involving 831 treatment farmers and 400 controls. Statistically significant improvements were observed among participants using NDEMRI, with mean crop yields increasing by 16.6% and agricultural incomes rising by 23%, relative to the control group. Adoption of improved agronomic practices was notably higher among users of the system. A total of 2,487 unique messages were exchanged, covering themes such as pest management, planting schedules, soil health, and post-harvest storage, with 78% of users reporting that system responses were context-sensitive and adapted to local climatic and cultural conditions. The technical architecture is characterized by modular natural language understanding pipelines, embedded guardrails to minimize model hallucinations, and a reproducible framework for contextualization based on regional agricultural datasets. A detailed economic analysis demonstrated the financial sustainability of the intervention, with favorable cost-benefit ratios and scalability potential. These findings offer robust empirical evidence that the integration of accessible communication technologies with state-of-the-art AI can overcome infrastructural limitations, enhance decision-making in low-resource farming environments, and serve as a viable model for transforming agricultural extension services across the African continent.
Accurate forecasting of Gross Domestic Product (GDP) growth remains essential for supporting strategic economic policy development, particularly in emerging economies such as Indonesia. In this study, a hybrid predictive framework was constructed by integrating fuzzy logic representations with machine learning algorithms to improve the accuracy and interpretability of GDP growth estimation. Annual macroeconomic data from 1970 to 2023 were utilised, and 19 input features were engineered by combining numerical economic indicators with fuzzy-based linguistic variables, along with a forecast label generated via the Non-Stationary Fuzzy Time Series (NSFTS) method. Six supervised learning models were comparatively assessed, including Random Forest (RF), Support Vector Regression (SVR), eXtreme Gradient Boosting (XGBoost), Huber Regressor, Decision Tree (DT), and Multilayer Perceptron (MLP). Model performance was evaluated using Mean Absolute Error (MAE) and accuracy metrics. Among the tested models, the RF algorithm demonstrated superior performance, achieving the lowest MAE and an accuracy of 99.45% in forecasting GDP growth for 2023. Its robustness in capturing non-linear patterns and short-term economic fluctuations was particularly evident when compared to other models. These findings underscore the RF model's capability to serve as a reliable tool for economic forecasting in data-limited and volatile macroeconomic environments. By enabling more precise GDP growth predictions, the proposed hybrid framework offers a valuable decision-support mechanism for policymakers in Indonesia, contributing to more informed resource allocation, proactive economic intervention, and long-term development planning. The methodological innovation of integrating NSFTS with machine learning extends the frontier of data-driven macroeconomic modelling and provides a replicable template for forecasting applications in other emerging markets.
Sovereign large language models (LLMs), emerging as strategic assets in global information ecosystems, represent advanced AI system developed under distinct national governance regimes. This study examines how model origin and governance context influence AI-generated narratives on international territorial disputes. The study compares outputs from three prominent sovereign LLMs - OpenAI’s GPT-4o (United States), DeepSeek-R1 (China), and Mistral (European Union), across 12 high-profile territorial conflicts. Statistically significant differences in each model's sentiment distribution and geopolitical framing are identified using a mixed-methods approach that combines sentiment analysis with statistical evaluation (chi-square tests and analysis of variance, ANOVA) on responses to 300 standardized prompts.
The findings indicate model provenance substantially shapes the tone and stance of outputs, with each LLM reflecting distinct biases aligned with its national context. These disparities carry important policy and societal implications: reliance on a single sovereign model could inadvertently bias public discourse and decision-making toward that model's native perspective. The study highlights ethical considerations such as transparency and fairness and calls for robust governance frameworks. It underscores the need for careful oversight and international cooperation to ensure that sovereign LLMs are deployed in a manner that supports informed and balanced geopolitical dialogue.
The proliferation of Generative AI (GenAI) tools has introduced new dynamics in user behaviour, environmental perception, and digital sustainability. This study, based on a primary questionnaire survey of 1,005 GenAI users aged 18 and above from India, investigates the frequency of GenAI usage and its relationship with climate change awareness, environmental concern, and willingness to adopt energy-efficient digital practices. Using regression-based models, the research reveals a pattern of indirect dependence: lower GenAI usage is related with a greater inclination toward environmentally responsible behaviours, such as transitioning from non- sustainable platforms and adopting energy-efficient digital services. In contrast, frequent GenAI users tend to perceive climate change as temporally distant and of lower immediate importance.
The study also examines how the frequency and nature of social media usage influence users’ attitudes toward sustainable technology choices. These findings provide valuable insights for policymakers, AI educators, digital strategists, and sustainability advocates aiming to foster environmentally conscious technology adoption in emerging economies like India.
Accurately diagnosing emotional and psychological disorders is essential for prompt mental health interventions, especially in intelligent healthcare systems. This paper proposes a deep learning model that uses convolutional neural networks (CNN) and long short-term memory (LSTM) networks to classify emotional states based on physiological inputs like EEG and ECG. Bayesian optimisation improves the model's learning efficacy and generalisation ability by adjusting hyperparameters. In comparison to conventional machine learning models such as Support Vector Machines (SVM), random forest, and standalone deep learning models (CNN and LSTM), the proposed CNN-LSTM architecture increases classification accuracy by 25%, to 92.1%. Its exceptional performance is demonstrated by its AUC-ROC score of 0.96, accuracy of 0.93, recall of 0.91, and F1-score of 0.92. These results show that the model can distinguish between several emotional states, including neutral, tense, and concerned. A real-time application is used to investigate the potential of wearable EEG-based brain-computer interface (BCI) devices for continuous emotional monitoring. The findings indicate that the proposed framework might be a helpful tool for the early detection and tailored management of mental health conditions in intricate healthcare environments.
Enhancing the productivity of forage crops while maintaining soil health remains a critical objective in sustainable agriculture. Excessive application of inorganic nitrogen (N) fertilizers, particularly urea, has contributed to soil degradation and environmental concerns, prompting the need for biologically sustainable alternatives. In this study, the effects of substituting urea with bioorganic fertilizer on soil quality and forage yield in an intercropping system of Pennisetum purpureum and Macroptilium atropurpureum were investigated. A randomized block design (RBD) was employed with six substitution treatments: no fertilizer (T), 0% substitution (S0), and 25% (S1), 50% (S2), 75% (S3), and 100% (S4) substitution of urea-N with bioorganic fertilizer. Each treatment was replicated four times, resulting in 24 experimental plots. Parameters evaluated included soil properties, populations of nitrogen-fixing bacteria (NFB) and phosphorus-solubilizing bacteria (PSB), and growth and biomass characteristics of the forage association. Substitution treatments significantly improved soil fertility indices. The highest soil organic carbon (SOC) (3.23%) was observed in S3, while total N content (Total N) in S2, S3, and S4 exceeded that of T and S0. Available phosphorus (P) was greatest in S3 and S4, and the highest cation exchange capacity (CEC) (24.08 me 100 g-1) was recorded in S4. The S2 and S3 treatments yielded the highest leaf dry weights (1.55 and 1.49 kg plot-1, respectively), stem dry weights (1.84 and 1.70 kg plot-1), and total dry forage weight (3.38 and 3.19 kg plot-1). Leaf-to-stem ratios and leaf areas in S2 and S3 were comparable to S0 and significantly greater than T. The lowest leaf area-to-total forage ratios (14.39 and 15.05 m² kg-1) were also observed in these treatments. It was demonstrated that 50% and 75% substitution levels of urea-N with bioorganic fertilizer not only enhanced soil quality parameters but also significantly increased forage productivity compared to exclusive urea application. These findings underscore the potential of bioorganic fertilizer as a sustainable alternative to inorganic N sources, contributing to improved soil health, higher forage yields, and more resilient agroecosystems.
The Bandar tradition observed in Negeri Rutah represents a culturally embedded mechanism of informal economic exchange, whereby financial contributions are voluntarily extended by community members to support families with sons entering marriage. This study has revealed that such a system operates not only as a means of reducing the financial burden associated with wedding ceremonies but also as an instrument for reinforcing communal bonds, intergenerational solidarity, and the continuity of intangible cultural heritage. Despite the absence of formal financial records or institutional oversight, contributions are managed through a trust-based system underpinned by mutual reciprocity and collective memory. The persistence of the Bandar tradition in contemporary society has been examined through the lens of social accounting, with a particular focus on its potential alignment with modern principles of accountability, transparency, and cultural resilience. Through qualitative field research, it has been demonstrated that the practice continues to function effectively within the community, sustained by deep-rooted social norms and communal expectations. However, challenges such as urban migration, generational shifts in value systems, and external economic pressures have been identified as potential threats to its long-term sustainability. The integration of culturally sensitive social accounting frameworks has therefore been proposed as a viable strategy for safeguarding this tradition against socio-economic disruption while preserving its core values. The study contributes to a broader discourse on the intersection of indigenous cultural practices, informal economies, and contemporary accountability systems, offering a model through which traditional mechanisms can be adapted without compromising their cultural integrity.