Engineering

Abstract: Accurate daily rainfall forecasting is essential for water resource management, flood risk mitigation, and early warning systems in tropical mountain regions. This study systematically compared five univariate forecasting models drawn from three methodological families—one statistical model (ARIMA), two deep learning architectures (Long Short-Term Memory, LSTM; Gated Recurrent Unit, GRU), and two decomposition-based hybrid models (Discrete Wavelet Transform–LSTM, DWT-LSTM; Variational Mode Decomposition–GRU, VMD-GRU)—for one-step-ahead daily rainfall forecasting across 18 meteorological stations in Tungurahua Province, Ecuador, covering the period 2013–2025. Model performance was evaluated using a comprehensive multi-metric framework that combined continuous accuracy metrics (Mean Absolute Error, MAE; Root Mean Square Error, RMSE; Nash-Sutcliffe Efficiency, NSE; Kling-Gupta Efficiency, KGE) and categorical event-detection metrics (Probability of Detection, POD; False Alarm Ratio, FAR; Critical Success Index, CSI) applied at three precipitation thresholds (1 mm, P90, and P95). LSTM achieved the best overall performance for continuous prediction and rainfall-occurrence detection at the 1 mm threshold; GRU delivered comparable skill at lower computational cost, making it a practical alternative for operational applications. Station-level performance rankings varied considerably across the network, reflecting strong hydroclimatic heterogeneity and confirming that spatially uniform single-model forecasting strategies are suboptimal. All evaluated models failed to reliably detect high-intensity (P90) and extreme (P95) rainfall events, producing very low POD and high FAR values. Although ARIMA ranked first in relative terms for extreme-event detection, its absolute skill remained insufficient for operational purposes. Hybrid decomposition models (DWT-LSTM, VMD-GRU) did not outperform the simpler recurrent architectures and frequently produced negative NSE values. These results demonstrate that univariate approaches are adequate for routine daily rainfall forecasting but inadequate for reliable extreme-event prediction, underscoring the critical need for multivariate predictor frameworks, specialized loss functions, and probabilistic approaches in future research.

Abstract: Andean blackberry is a highly perishable fruit rich in anthocyanins and ellagitannins, and its short postharvest life significantly affects smallholder growers’ economies. Although crossflow microfiltration is a promising non-thermal stabilization technology for blackberry juice, its industrial application is limited by permeate flux decline during long-term operation, while most previous studies have focused on short processing times. This study evaluated the effect of high-shear homogenization prior to enzymatic depectination on flux decline, product quality, and techno-economic feasibility during CFM. Juice processed by conventional grinding followed by high-shear homogenization and enzymatic treatment was subjected to CFM at feed volumes up to 400 L and volumetric reduction factors close to 30. Homogenization significantly reduced particle size and suspended insoluble solids, resulting in higher permeate flux, improved flux stability, and greater productivity. Flux decline analysis showed that high-shear homogenization extended the stable filtration regime and delayed severe fouling, sustaining higher average permeate fluxes during long-term operation. Product quality was preserved, ensuring microbial reduction while improving anthocyanin and ellagitannin recovery and enhancing blackberry aroma. In addition, HS3+E reduced energy consumption and beverage production cost while achieving a positive NPV and a 21% IRR. Overall, homogenization improved the industrial feasibility of long-term CFM processing of Andean blackberry juice.

Abstract: Contemporary extended reality (XR) systems deliver high-fidelity visual and spatial-audio immersion, yet the cognitive and affective states of the user remain invisible to the virtual environment. We present PiEEG XR, an open-source platform that bridges consumer-grade electroencephalography (EEG) hardware with WebXR-based mixed reality to produce real-time, neural-adaptive avatar behaviour. The system acquires multi-channel EMG at 250 Hz with 24-bit resolution via Bluetooth Low Energy 5 (BLE5), derives five canonical frequency-band power features (delta, theta, alpha, beta, gamma) using a rolling-window Fast Fourier Transform, and maps the resulting affective state estimates to VRM avatar facial expressions and interaction affordances rendered through a React Three Fiber / Three.js pipeline on the Meta Quest browser. A lightweight OSC bridge enables simultaneous streaming to social VR platforms such as VRChat. We describe the system design, the EEG-to-expression mapping heuristic, the WebXR spatial anchoring strategy, and an OSC parameter protocol. PiEEG XR represents an accessible entry point for affective computing research in immersive environments and demonstrates that production-ready neural-adaptive avatars can be deployed entirely within an open browser stack.

Abstract: Smart-city infrastructure planning requires decision-support models that can evaluate urban mobility, power-grid readiness, land-use constraints, and accessibility under in-terdependent conditions. In electric-vehicle charging-station siting, traffic demand and grid capacity are not independent, since high demand becomes practically valuable only when adequate power infrastructure is available. This study proposes the Nonlin-ear Interaction-Einstein Aggregation (NI-EA) model as an interaction-aware MCDM framework for siting EV charging stations in smart cities. The model combines a linear baseline score, pairwise criterion interaction terms, a unit-interval nonlinear transfor-mation, and an Einstein aggregation component within one interpretable decision structure. Three diagnostic indicators, namely the Interaction Strength Index (ISI), the Weighted Interaction Strength Index (WISI), and the Active Interaction Strength Index (AISI), are introduced to characterize the interaction structure before final ranking. Benchmark cases and an illustrative smart-city EV charging-station application show that NI-EA preserves stable rankings when the decision structure is clear, improves score discrimination among close alternatives, and reveals meaningful ranking changes when traffic flow and grid capacity interact. The results indicate that NI-EA can sup-port transparent and reusable smart-city infrastructure decisions involving interde-pendent criteria. The framework is reusable across related smart-city decisions involv-ing mobility-grid interactions and urban energy infrastructure.

Abstract: ANSI/IES TM-30-24 uses 99 Color Evaluation Samples (CES) selected to provide broad, approximately uniform sampling of object-color space for evaluating color rendition in built environments. Although TM-30 is not intended as a dedicated skin-tone fidelity assessment method, the standard designates two CES, #15 and #18, as skin-tone samples. This study examines how this sampling structure intersects the color space region occupied by measured human skin tones. The analysis used 100 skin reflectance spectra from two independent datasets grouped into five lightness-based categories from Very Light to Very Dark. CAM02-UCS coordinates were computed under D65 using TM-30-specified viewing conditions, and Euclidean nearest-neighbor distances (ΔE′) to the 99 CES were calculated. In both datasets, the Very Dark category showed substantially larger nearest-CES distances than the Very Light category. In the ISSA dataset, Very Dark samples had a mean distance of 8.13 ± 2.17, compared with 3.99 ± 0.59 for Very Light samples. In the CIE dataset, the corresponding values were 7.96 ± 0.94 and 4.24 ± 0.85. Across the combined dataset, Very Dark samples were approximately 95% farther from their nearest CES than Very Light samples. The designated skin CES, #15 and #18, served as nearest neighbors for 51 of 100 samples, primarily in the Very Light, Light, and Medium categories. These findings provide a structural map of how measured skin tones relate to the TM-30 CES set and clarify the role of CES #15 and #18 within the broader skin-tone region.

Abstract: Determining the parameters of a manipulator for optimal performance is a challenging task. This is primarily due to the possible conflicting objectives, the various tasks that should be considered and the highly non-linear behavior that is involved. This work proposes an enhanced version of the Concept-based Design Space Exploration (C-DSE) approach for the design of manipulators. According to the C-DSE approach, prior to the search, the designers divide the set of feasible solutions into meaningful subsets, which are termed concepts. The design space exploration involves a simultaneous search for optimal solutions within each of the pre-defined concepts. This enhanced framework integrates: (1) Kinematics, dynamics and control co-design—simultaneous optimization of manipulator morphology and controller parameters; (2) Surrogate-assisted optimization using Gaussian Process (GP) and Neural Network (NN) models to reduce computational cost; (3) Comprehensive performance metrics across approximately 30 metrics; (4) Task-aware feasibility verification applying multi-level hierarchy; (5) Generative AI Integration with diffusion models and LLM-guided concept generation. The results demonstrate computational cost reduction of 96%, 40% task performance improvement and superior performance compared to state-of-the-art methods from 2020–2026.

Abstract: Photovoltaic technology selection in Gulf Cooperation Council (GCC) desert environments is affected by coupled dust, thermal, ultraviolet (UV), humidity and salinity stresses, which are not fully represented by static weighting and additive multi-criteria decision-making models. This study develops a GCC-calibrated nonlinear decision-support framework that integrates field-based environmental calibration, adaptive hybrid entropy–desert weighting and bipolar fuzzy Einstein aggregation. The framework is applied to compare PERC, TOPCon and heterojunction (HJT) photovoltaic technologies using calibrated evidence from Qatar, the United Arab Emirates, Saudi Arabia and Oman. Results show that dust tolerance receives the highest final hybrid weight (0.258), followed by thermal resistance (0.228), UV resistance (0.207), efficiency (0.173) and cost effectiveness (0.134). The nonlinear Einstein aggregation ranks HJT first (0.889), followed by TOPCon (0.861) and PERC (0.742). Benchmark comparison with TOPSIS, VIKOR and PROMETHEE II shows high rank agreement, while Monte Carlo perturbation analysis indicates that HJT preserves the first rank in 93% of scenarios. The proposed framework links PV technology selection with measured GCC desert stress behavior and provides a reproducible basis for technology prioritization in harsh solar-energy deployment environments.

Abstract: To improve the laser-induced damage resistance of 7075 aluminum alloy, a typical aerospace material, a laser-resistant coating material for the surface of 7075 aluminum alloy was prepared in this paper. The performance of 7075 aluminum alloy coated with this coating was analyzed and tested by combining numerical simulation and experimental verification. The test results show that under the same laser irradiation conditions and geometric dimensions, the breakdown time of 7075 aluminum alloy coated with the laser-resistant coating is prolonged by 541.6%, and its laser-induced damage resistance is significantly improved.

Abstract: This research establishes critical process–property relationships for the Fused Deposition Modeling (FDM) of rigid (PLA) and functional flexible (TPU) polymers. By systematically evaluating the influence of key manufacturing parameters—infill density, print orientation, and loading rate—this study quantifies the structural integrity and operational limits of 3D-printed components. Tensile characterization, conducted according to ASTM D638-14 standards, reveals that horizontal and laid manufacturing strategies produce superior mechanical properties, whereas vertical configurations exhibit significant anisotropy due to interlayer bonding constraints. Stress-strain analysis differentiates the brittle, high-stiffness response of PLA from the hyper-elastic and viscoelastic behavior of TPU, highlighting the latter's sensitivity to rate-dependent processing. A critical finding for advanced manufacturing is the role of support-induced interfaces: while negligible for rigid PLA, support structures are shown to be detrimental to TPU, reducing maximum elongation by over 70% and necessitating support-free design strategies for functional elastomeric parts. Higher infill densities consistently enhance structural performance, providing a clear roadmap for optimizing material efficiency. These findings provide a principled framework for Design-for-Manufacturing (DfM) optimization, enabling the reliable production of tailored components for aerospace, biomedical, and soft-robotic applications.

Abstract: The dense tree canopy in the complex orchard environment obstructs wireless positioning signals and generates NLOS interference, which reduces the positioning accuracy of agricultural mobile robots. This study investigates a localization method for agricultural mobile robots based on two UWB tags and an electronic compass. By analyzing the NLOS interference factors and error sources of UWB, a method for NLOS interference suppression and positioning correction employing two UWB tags tightly coupled with heading angle was proposed. The construction of the heading angle L2IB system and its comprehensive process were also introduced as follows. The proposed method constructs candidate localization domains for dual UWB tags based on multilateration and integrates the inter-tag distance and heading-angle constraints within an L2IB framework to suppress NLOS-induced errors and estimate the robot center position. Experiments were performed under four simulated scenarios, namely line-of-sight (LOS), single-anchor occlusion, multi-anchors occlusion, and single-tag occlusion. The proposed method was compared with the centroid and least-squares methods. The results demonstrate that the L2IB method effectively improves localization accuracy under NLOS conditions. Specifically, in the single-tag NLOS interference scenario, the MAE, RMSE, and maximum localization error were 3.7, 4.0, and 6 cm, respectively. These results indicated that the system could meet the positioning needs of most NLOS environments in the orchard. Therefore, the proposed method exhibits feasibility and provides a new alternative for high-precision localization of mobile robots in orchards under NLOS conditions.

Abstract: Industrial Internet of Things (IIoT) production environments rely on automated communication between control systems and embedded devices while operating under strict availability constraints that limit the deployment of conventional IT security controls. Despite extensive research on intrusion detection systems, empirical studies based on long-term observations of real industrial networks remain scarce. This paper presents a three-month passive monitoring study of a production-line IIoT network in a factory environment. A containerised instance of Zeek was deployed in promiscuous mode to collect flow-level and application-layer telemetry without interfering with operations. The results reveal highly deterministic communication patterns dominated by periodic HTTP polling between a central server and distributed devices. Analysis of connection and HTTP logs shows strong temporal regularity, low variability, and stable communication regimes. Although no confirmed malicious activity was observed, transient deviations were identified and attributed to operational events such as production stoppages. These findings demonstrate that production-line IIoT networks can exhibit predictable behaviour suitable for statistical anomaly detection. The main contributions are a longitudinal empirical characterisation of a real-world IIoT network, a methodology for behavioural baseline extraction using passive monitoring, and practical insights for safe deployment in operational environments.

Abstract: Landmine contamination remains a critical barrier to post-conflict recovery, necessitating detection methods that are safe, scalable, and cost-effective. This study investigates drone-based detection of relevant plants in a mixed vegetation as required for future plant-based biodetection of explosive remnants derived from landmines. Field experiments were conducted in Ukraine and Denmark using oilseed radish and winter rapeseed as indicator biosensors due to their suitability for genetic modification and resilience in diverse climates. High-resolution RGB and multispectral imagery were collected under varying conditions, including flight altitude, time of day, and land preparation. We evaluated plant visibility at different altitudes and trained models to detect the plants and developed strategies to streamline annotation. Practical deployment challenges, such as seeding and growth of indicator plants in non-cultivated and hazardous soils are also considered. Our findings demonstrate the potential of aerial biosensor monitoring and offer insights into the development of plant-based landmine detection systems suitable for real-world applications.

Abstract: This study assessed the current status of irrigation systems and water management practices in Rwanda’s irrigated agricultural zones focusing on Nasho Government Funded Irrigation (GFI) scheme in Kirehe District and Kagitumba Irrigation Scheme in Nyagatare District. A mixed descriptive approach was applied combining field observation with structured questionnaires administered through Kobo Toolbox to 224 respondents in Nasho and 188 respondents in Kagitumba. Field observations were used to evaluate the physical condition and functionality of irrigation infrastructure while questionnaires captured stakeholder perceptions, water management practices, institutional arrangements and operational challenges. Results show that both irrigation schemes are operational but function below optimal efficiency due to multiple constraints. In Nasho, irrigation performance is mainly affected by sedimentation in canals and reservoirs, pump inefficiencies and inadequate maintenance practices leading to unreliable water delivery. In Kagitumba, despite the use of modern center pivot systems performance is constrained by pipeline corrosion, pressure losses, sediment-laden water and uneven water distribution. Across both schemes, more than 80% of respondents reported frequent system failures while over 95% indicated the absence of formal irrigation scheduling practices. Water management remains largely reactive with limited preventive maintenance and weak technical capacity among users and institutions. The study concludes that improving irrigation efficiency in Rwanda requires integrated interventions combining infrastructure rehabilitation, strengthened maintenance systems, improved water governance and farmer capacity development to enhance sustainable water use and agricultural productivity.

Abstract: The supercritical CO₂ extraction of essential oils from Origanum vulgare L., Matricaria chamomilla L., and Moringa oleifera Lam. was kinetically interpreted using a logistic mass transfer approach under different combinations of pressure and temperature. Extractions were performed in a fixed-bed SFE system operated for 210 min using high-purity CO₂ under pressures ranging from 100 to 500 bar and temperatures between 30 and 60 °C, depending on the vegetable matrix. The logistic model was parameterized through the total extractable mass (m_t), the characteristic time associated with the maximum extraction rate (t_m), and the kinetic slope parameter b. The highest extraction yields were obtained at 300 bar and 45 °C for oregano (2.807 g), 100 bar and 40 °C for chamomile (5.006 g), and 500 bar and 60 °C for moringa (5.433 g). Simultaneously, increasing pressure and temperature systematically reduced, decreasing from 16.737 to 8.75 min in oregano and from 15.01 to 9.73 min in moringa, indicating an intensification of convective-diffusional transport mechanisms. The model adequately reproduced the experimental extraction curves, particularly in Oregon, where SSD values remained below 0.03 under all evaluated conditions. Unlike highly parameterized phenomenological approaches, the proposed logistic formulation represented the extraction dynamics using kinetically interpretable parameters without requiring experimentally inaccessible internal coefficients. The results demonstrate that logistic modeling constitutes a mathematically simplified but kinetically robust alternative for the comparative analysis and preliminary optimization of supercritical extraction systems applied to aromatic and medicinal plant matrices.

Abstract: Alzheimer's disease (AD), a leading cause of dementia worldwide, is a neurological disorder characterized by progressive cognitive decline. AD is also considered a significant socioeconomic burden. While definitive diagnostic tools such as positron emission tomography (PET) imaging and cerebrospinal fluid (CSF) biomarker analysis offer high sensitivity and specificity, they are limited by high cost, invasiveness, and limited accessibility. Consequently, these gold standard approaches hinder their applicability for large-scale screening and longitudinal follow-up. Recent advances in blood-based biomarkers hold promise in capturing systemic molecular changes associated with AD. In particular, transcriptomic signatures derived from RNA sequencing (RNA-seq) are promising in capturing systemic molecular changes associated with AD. Gene expression profiles in peripheral blood reveal underlying pathological processes. These pathological processes can be listed as synaptic dysfunction, neuroinflammation, and metabolic dysregulation. Together with the high-dimensional datasets and AI approaches enable the identification of robust predictive models which has the assistance of estimating AD-related biomarker status. We further discussed the integration of multiple omics data, including genomics, proteomics, and metabolomics to improve biomarker robustness. We also addressed key challenges related to reproducibility, repeatibility, cohort heterogeneity, and clinical application. And we outline future directions of standardized, scalable, and clinically applicable diagnostic machineries.

Abstract: The integration of generative artificial intelligence (AI) into agricultural extension services presents a transforma- tive opportunity to address the unique challenges faced by smallholder farmers, particularly in resource-constrainedsettings. While traditional extension services often struggle with scalability and personalized support, generative AI offers potential solutions through dynamic content generation, real-time decision-making assistance, and adaptive learning tools. This systematic literature review examines the efficacy of generative AI in enhancing agricultural extension services, focusing on its applications, benefits, and limitations for smallholder farmers. We synthesize existing research across multiple dimensions, including AI-driven farmer support, IoT-enabled monitoring, andclimate-smart agriculture, to identify gaps and trends in the current knowledge landscape. A rigorous methodol- ogy was employed to select and analyze relevant studies, ensuring a comprehensive evaluation of both theoreticalframeworks and practical implementations. The findings reveal that generative AI can significantly improve access to tailored agricultural advice, optimize resource allocation, and mitigate climate-related risks; however, challengessuch as digital literacy, infrastructure limitations, and ethical concerns remain critical barriers to widespread adop- tion. The review also highlights the disproportionate focus on high-income regions, underscoring the need for moreinclusive research in low-resource agricultural systems. By consolidating these insights, we provide actionable rec- ommendations for policymakers, researchers, and practitioners to harness generative AI’s potential while addressingits socio-technical constraints, thereby fostering equitable and sustainable agricultural development.

Abstract: The rapid advancement of generative large language models (LLMs) has sparked significant interest in their poten- tial to transform higher education, particularly in fostering student engagement. While these models offer novelopportunities for personalized learning and interactive experiences, their integration into academic settings remains underexplored, with varying implications for pedagogy, ethics, and institutional policy. This systematic literaturereview examines the role of generative LLMs in enhancing student engagement across multiple dimensions, includ- ing their impact on learning outcomes, academic writing, subject-specific applications, and ethical considerations.We synthesize existing research to identify key trends, challenges, and gaps in the current understanding of how these technologies are reshaping educational practices. A rigorous methodological approach was employed to select and analyze relevant studies, ensuring a comprehensive evaluation of the field. The findings reveal that generative LLMs can significantly influence student engagement by facilitating adaptive learning environments and supporting creative problem-solving; however, concerns about academic integrity, equitable access, and pedagogical alignment persist. The review also highlights emerging tools and systems designed to integrate LLMs into education, alongside institutional and student perspectives on adoption. Based on the synthesized evidence, we discuss future directions for research and policy, emphasizing the need for balanced frameworks that harness the benefits of generative AI while addressing its risks A. Chowdhury et al., 2025. This study contributes a structured overview of the current landscape, offering insights for educators, researchers, and policymakers navigating the evolving intersection of AI and higher education.

Abstract: The experimental determination of the relationships between the stress distribution zone in the soil layer and the parameters of tillage working bodies is a labor-intensive process. Therefore, preliminary mathematical modeling of this process is recommended to mi-nimize the total number of experiments. The research was conducted using the principles of classical mechanics and soil mechanics. Using an equation proposed by J. Boussinesq,a graphical-analytical method was developed to evaluate the stress state in the soil layer induced by a dihedral wedge. This method incorporates both the geometric parameters of the dihedral wedge and the physico-mechanical properties of the soil. A direct pro-portional relationshipwas established between the length of the dihedral wedge and the total area of the deformed soil mass. Specifically, increasing the length of the dihedral wedge by 83% (from 0.05 to 0.30 m) resulted in an 80% increase in the area of the de-formed soil mass (from 0.02 to 0.10 m²). The proposed graphical–analytical method can be employed in the design of tillage implements.

Abstract: Sugarcane harvester performance varies substantially with field geometry, crop, and operator factors, yet separating these sources from telematics data while preserving engineering interpretability remains a methodological gap. This study models field efficiency (Eff) and harvesting capacity (Ca) separately from JDLink telematics, aligning model structure with each target's response behavior. Operational data covered 105 plots across four seasons (2019/20–2022/23) from three John Deere chopper harvesters in eastern Thailand. Six engineering-relevant predictors were retained after multicollinearity screening, and linear (MLR), additive nonlinear (GAM), and tree-based models were compared under 5-fold grouped cross-validation by BaseField (87 groups). Eff was assigned to GAM (R²CV = 0.621 ± 0.114) on the basis of its threshold-like response to turning frequency; Ca was retained for MLR (R²CV = 0.681 ± 0.121), with GAM essentially tied. Train–validation gaps were substantially smaller for additive models (0.096–0.118) than for tuned tree-based candidates (GBR 0.210–0.302, RF 0.322–0.358). Turning frequency (TF) and perimeter-to-area ratio (PAR) were the strongest predictors, and a constant-turn-time partial-out test indicated that TF's univariate effect on Eff is largely mediated by the time-budget identity. Tactical interventions (path planning, operator training, machine–field allocation) are immediately feasible, although strategic field-layout change remains constrained by smallholder land tenure.

Abstract: Emerging materials often face challenges in market adoption due to limited comparability and reliability of measurement-based material information, despite their potential to drive technological innovation. While standardization is widely recognized as an important mechanism for market diffusion, existing approaches provide limited insight into how material specifications facilitate the comparative evaluation of material characteristics and their use in market decision-making. This study proposes a complementary perspective that interprets standardization as an infrastructure for organizing the generation, sharing, and evaluation of measurement-based material information across industry, standard development organizations (SDOs), and markets. Within this framework, the study distinguishes between two complementary types of standards for material specifications. Type A standards enable the structured disclosure of measured characteristic values and associated measurement uncertainties, allowing application-specific evaluation without predefined acceptance criteria. In contrast, Type B standards define predefined characteristic values and compliance criteria, providing a basis for conformity assessment, certification, and quality assurance. These two types may be understood as complementary mechanisms that fulfill different functions of comparability and compliance under varying technological and market conditions in emerging material systems. Consequently, they contribute to both innovation-oriented market evaluation and quality-assured market acceptance.