National University of Science and Technology

This study presents results of a systematic literature review (SLR) of the responsible use of artificial intelligence (AI) tools in higher education, identify patterns of ethical and irresponsible use, and propose a conceptual framework for predicting ethical AI adoption. Following PRISMA guidelines, was conducted on 60 peer-reviewed studies published between 2022 and 2026, sourced from Google Scholar. Studies were mapped against four research questions addressing AI tools used, their applications, reported unethical practices, and predictive modelling approaches. Results reveal that general AI, generative AI tools, and large language models dominate higher education contexts, primarily deployed for personalized learning, academic work, and teaching. Irresponsible practices were documented in one-third of studies, including academic integrity breaches (13.33%), algorithmic bias, and privacy violations. Critically, no existing study developed a real-time predictive model capable of monitoring ethical AI use, despite four studies demonstrating predictive modelling capabilities for other purposes. This study addresses a significant gap by proposing a novel conceptual framework that integrates AI tool deployment, user behaviour, governance measures, and predictive analytics to forecast ethical outcomes. The framework provides higher education institutions with a pathway toward data-informed, proactive governance of AI technologies.

<p>Motor evoked potential (MEP) latencies generated in transcranial magnetic stimulation (TMS) experiments are an important and fundamental physiological parameter for studying the human motor system. Manual annotation of MEPs can be performed, but this is time-consuming and subject to human error, which can lead to problems with automated methods for extracting MEP features. The purpose of this study is to compare existing methods for automated MEP latency determination in terms of their advantages and disadvantages and propose a new method that can improve results. The hypothesis is that we can propose a new automated annotation model that can improve results, or that this model will fail to automatically extract MEP features. Each existing method will be examined in terms of its advantages and disadvantages to identify targeted issues that will be addressed in the proposed model. The results show that all existing methods share a common problem: they are not reproducible to other datasets. This means that each model was focused on extracting MEP features under specific conditions and dataset formats. This study identified relevant challenges that we must consider when developing a new model for automatically annotating MEP features. Therefore, it is recommended to pay attention to every detail to achieve successful results.</p>

Iron in coals occurs in various forms (pyrite, ultradispersed particles, ions in the carbon matrix) and plays a key role in oxidation processes, spontaneous combustion, liquefaction, and gasification. Although extensive experimental data have accumulated over recent decades, a systematic generalization linking quantitative estimates of the surface density of catalytically active Fe centers to coal rank has been lacking. This study provides a systematic analysis of literature data on the content, occurrence forms, and catalytic activity of Fe centers in coals from different deposits and rank stages. Data from Russian (Agroskin, Epshtein, Larionov), German (Kreysa), American (Kaiser, Silbernagel), and Chinese (Chen) research groups over more than 40 years are analyzed and compared. A quantitative estimate of the surface density of active sites n_act (m⁻²) is derived from three independent experimental approaches: (1) kinetics of coal interaction with ozone; (2) influence of Fe additives on the activation energy of oxidation; (3) electrochemical oxidation data with Fe centers as mediators. Conversion of active site concentration (mol/g) to surface density (m⁻²) uses data on specific surface area accessible to water and dissolved reagents, with sensitivity analysis for this parameter. Based on the Epshtein classification, three coal groups are distinguished: Group 1 (minimal oxidation propensity, anthracites, lean coals) with n_act < 10¹⁴ m⁻²; Group 2 (pronounced oxidation propensity, gas-fat, fat, coking coals) with n_act = 10¹⁵–10¹⁶ m⁻²; Group 3 (high oxidation propensity, brown and long-flame coals) with a wider scatter (10¹⁵–10¹⁷ m⁻²) and lower center stability. The estimate for Group 2 coals is supported by all three methods: from ozonation data (n_act = 4×10¹⁵–1.7×10¹⁶ m⁻²); from Fe additive data (n_0 = 10¹⁴–10¹⁶ m⁻²); and from electrochemical data (n_eff = 10¹⁴–10¹⁶ m⁻² after accounting for surface accessibility). Sensitivity analysis shows that variation of specific surface area in the realistic range (3·10⁵ – 3·10⁶ m²/m³) changes the estimate by no more than a factor of 3, preserving the order of magnitude. Factors affecting center accessibility under in situ conditions (pore blocking by water, poisoning by sulfur, competition with methane, thermal deactivation) are analyzed, giving a conservative estimate for effective in situ density of n_eff ≈ (2·10¹⁴ – 5·10¹⁵) m⁻². The quantitative estimates obtained may be useful for parameterizing theoretical models of catalytic processes in coals, including resonant oxygen activation methods for methane mitigation in coal seams.

Globally, the construction industry produces around 40% of all annual waste, making it a major contributor to environmental challenges within the built environment. While digital tools and circular economy practices have shown promise in reducing construction and demolition waste, most research focuses on developed countries. This paper seeks to understand the current state of digitalization in construction and demolition waste (CDW) management, identify key themes and trends, and provide guidance for adopting digital technologies to support sustainable construction waste management in developing countries. The PRISMA methodology and the archival research methods were used to identify papers published in the last 10 years through Scopus, Web of Science, and Google Scholar. The results show that of the 60 papers that were identified most of the research is concentrated in Europe and Asia. BIM was also found to be the most digital tool used in managing construction and demolition waste. More advanced technologies, such as machine learning, artificial intelligence, computer vision, and robotics, are mainly applied at demolition and recycling stages to support waste identification and material recovery. Based on the findings, the study outlines a phased approach to digital CDW management for developing countries, starting with BIM-based waste prevention and progressing towards more integrated, standardized, and scalable digital solutions that support sustainable industrialization in the built environment. As a mapping-focused study, this review emphasized the classification of digital tools and their applications; future studies should focus on empirical evaluations and comparative analyses to provide deeper performance-based insights into digital solutions for CDW management.

Sustainable feedstock management remains a major challenge in laser beam powder bed fusion (PBF-LB), where conventional reuse strategies are typically limited to sieving and blending rather than full material regeneration. Ultrasonic atomization (UA) offers a fundamentally different powder production route based on capillary-wave instabilities induced at the surface of a molten metal by high-frequency vibrations. In contrast to turbulence-driven atomization, droplet formation in UA is primarily governed by ultrasonic frequency and intrinsic thermophysical properties of the melt, enabling quasi-deterministic particle formation with high sphericity and reduced satellite formation. In this study, ultrasonic atomization was investigated as a closed-loop route for converting PBF-LB-manufactured 316L stainless steel parts into reusable powder. Printed rods were remelted and atomized under controlled variation of electric current and vibration amplitude. The resulting powders were characterized in terms of morphology, internal microstructure, particle size distribution, chemical composition, and gas impurity content. UA produced highly spherical particles with reduced internal porosity and improved flowability compared to the initial gas-atomized powder, while preserving the principal alloying elements. An increase in oxygen content was observed after recycling, attributed to selective high-temperature oxidation under residual oxygen in nominally inert conditions. The results establish a mechanistic framework for transforming consolidated PBF-LB material into secondary feedstock and identify key parameters governing structural and compositional stability in closed-loop recycling.

This review systematizes data on the phase composition and key properties of compounds in the W–B system, including thermodynamic stability, crystal structure, and hardness. The current understanding of the binary W–B phase diagram and the stability of individual borides is discussed, alongside the influence of defects and non-stoichiometry on their properties. The main methods for synthesizing these materials and producing coatings based on them are summarized. Potential applications of tungsten borides are highlighted, particularly for high-temperature environments, cutting tools, and protective and functional coatings. Finally, key directions for future research are outlined, focusing on the refinement of phase equilibria, the scaling of production methods, and the development of W–B-based materials with tailored performance characteristics.

To increase printing speed and quality, a route consisting of using two sequential coextruders to form impregnated fiber immediately before feeding it to the printer. Such an approach, aimed at allowing the use of the most common industrial 12K carbon fibers for additive manufacturing, prevents damage to composite fibers during transportation, storage, and loading. An elaborate system was used to prepare carbon fibers impregnated with polypropylene, ethylene vinyl acetate, and their blends. The used scheme allows the production of composite fibers containing from 60 to 80 wt. % of carbon fibers. It was found that the elastic modulus of the composite fibers is close to those for raw carbon fibers and does not depend on the used polymer. It shows that the used carbon fiber path in the polymer melt and two sequential calibrating nozzles result in a high degree of orientation of the elementary filaments in the fiber at impregnation and maintain the elastic properties of the carbon fiber in the resulting composite. The tensile strength of the composite fibers depends on the polymer content in the composite fiber; the highest tensile strength was observed for fibers impregnated with ethylene vinyl acetate when increasing the coextrusion temperature up to 220 °C, which results in a composite fiber with a polymer content of 30 wt. %. A decrease in the polymer content in composite fibers results in a decrease in strength.

The potential damages of low-rise building foundations on liquefiable sites were studied using the Midas GTS-NX analysis. Both free-field and foundation responses for single raft and piled raft foundations were monitored on layered sands using the UBCSand model. It was found that a single raft can be easily uplifted by surrounding groundwater and soil liquefaction occurred. The piled raft foundation has better seismic resistance compared to the single raft. The foundation load will enlarge the ground settlements but reduce the horizontal ground displacements underneath the foundation. The 2D analysis can provide rational predictions of the piled raft responses despite the foundation geometry not being fully simulated. Using a thicker raft can eliminate the angle variable of the foundation. Considerable amounts of foundation settlements were found for single rafts and piled rafts according to the design specification. Longer piles will help to reduce the foundation settlement.