Search Results (2,229)

Accurate prediction of pile base resistance is essential for the safe and economical design of deep foundations, particularly in soft soils where load-transfer mechanisms are highly nonlinear and uncertain. This study develops a comparative, probabilistic, and interpretable machine learning framework for predicting pile unit base resistance using five input variables: applied load, settlement, effective pile length, axial stiffness, and SPT value. A Gaussian Process Regression model with an automatic relevance determination (ARD) Exponential kernel achieved the best performance, with RMSE = 262.11 kPa, R² = 0.943 on an independent test set, and 95% prediction intervals with 96.46% coverage. Beyond record-level evaluation, a leave-one-pile-out validation (the first grouped validation applied to this database) showed harder generalization to entirely unseen piles, driven mainly by a per-pile level offset rather than shape mismatch (within-pile correlation = 0.975). A sequential next-stage scheme, calibrating this level from a pile’s early loading stages, then predicted its remaining segments with consistently strong agreement (Willmott’s d = 0.76–0.83), supporting practical extension of partial load tests. Interpretability was assessed using ARD, SHAP, permutation/ablation importance, and partial dependence/accumulated local effects analysis, identifying settlement as the dominant predictor. The framework combines accuracy, calibrated uncertainty, interpretability, and validated segment-level extrapolation for reliability-oriented pile assessment. Full article

Chiranthodendron pentadactylon Larreat (“flor de manita”) is traditionally used for gastrointestinal, cardiovascular, and neurological disorders. Objective: This study comprehensively integrated phytochemical characterization, in silico toxicological profiling, an exposure-informed Integrative Toxicity Prediction Model (ITPM), uncertainty analysis, and acute testing of a fresh flower aqueous extract (FFAE). Methods: Thirty-six phytochemicals were screened for hepatotoxicity, nephrotoxicity, Ames mutagenicity, carcinogenicity, hERG inhibition, reproductive-effects alerts, and acute oral toxicity. Five organ- or effect-specific endpoints were integrated into an exploratory Toxicological Alert Score (TAS), whereas reproductive alerts and predicted LD₅₀ were reported separately. Seventeen compounds quantified in the FFAE were compared with a 20-compound literature-informed profile through 50,000 Monte Carlo iterations. Male CD-1 mice received a single FFAE dose of 300–5000 mg/kg and were observed for 14 days. Results: Positive predictions occurred for hERG inhibition in 33 compounds (91.7%), hepatotoxicity in 29 (80.6%), nephrotoxicity and carcinogenicity in 28 each (77.8%), and mutagenicity in 21 (58.3%). Six compounds had predicted LD₅₀ values ≤ 1000 mg/kg. The proportional ITPM toxicity-evidence index was 6.91% for the experimental profile and 19.34% for the literature-informed profile. Monte Carlo means were 6.98% (95% uncertainty interval, 5.65–8.44%) and 19.32% (16.76–22.37%), respectively. No mortality, overt clinical signs, or treatment-related body weight differences occurred up to 5000 mg/kg. Conclusions: Compound-level predictions identified priorities for confirmatory testing, while the extract-specific model yielded a lower comparative toxicity-evidence proportion, and the animal study showed low overall acute toxicity under the evaluated conditions. These findings do not establish general or long-term safety, calibrated risk probabilities, or biological neutralization of toxicological liabilities. Full article

Agile outcome models commonly map Jira variables directly to a retrospective label and then explain the prediction through fragmented feature attributions; they rarely separate domain concepts, team clustering, unresolved work, and concept-label coupling. This study evaluates a domain-informed, concept-bottleneck-style explainable AI architecture for retrospective diagnosis of Agile Epic outcomes. A frozen Jira export of 10,000 unique issue-level records was linked to a pre-specified analytical cohort of 180 Epics across 14 teams. Six experts rated efficiency, effectiveness, sustainability, and contextual risk, while outcomes were recorded as Successful, Challenged, or Unsuccessful. Because the outcome labels and concept ratings were informed by the same Jira evidence, the models estimate consistency with an expert labelling procedure, rather than independent project success. Under five-fold group-aware cross-validation, the fixed-configuration flat LightGBM achieved macro-F1 = 0.864 ± 0.053 and the fixed-configuration HMXAI/CBM-style model achieved 0.843 ± 0.084. These descriptive primary scores are not a joint nested-model-selection comparison. The proposed method, therefore does, not demonstrate a performance improvement; its contribution is an inspectable diagnostic structure. Performance fell materially on the resolved-only subset (LightGBM macro-F1 = 0.645), and model-specific nested, leave-one-team-out, calibration, uncertainty, correlation, and intervention analyses further bound the claims. Concept interventions were not uniformly monotone, so the concept layer is domain-interpretable in form but not yet user-validated as actionable. The study contributes a transparent audit of when concept-level diagnosis can complement flat classification and when circularity, censoring, and shortcut learning restrict interpretation. Full article

Spacecraft and launch-vehicle structures require lightweight multipoint monitoring under combined mechanical, thermal, and environmental loads. This study presents an analytical fiber Bragg grating (FBG) sensor-network workflow integrating reference-grating temperature compensation, regional strain assessment, opposite-surface curvature sensing, wavelength-division-multiplexing allocation, and strain-to-shape reconstruction. The deterministic compensation case is used only as a self-consistency check, whereas practical robustness is assessed through 10,000 Monte Carlo trials incorporating packaged-coefficient mismatch, temperature nonuniformity, wavelength noise, strain-transfer variation, drift, and calibration uncertainty. The calibrated estimator achieved a median strain mean absolute error of 1.73 με and a 95th-percentile error of 4.22 με. The defined finite-element benchmarks produced a maximum engine-mount truss strain of 456.2 με under the defined loads and a median full-field panel-reconstruction normalized root-mean-square error of 1.29% for 18 sensing locations with 2 με noise. Conservative WDM analysis yielded 54, 13, and 16 channels for three operating envelopes, and the prescribed random-vibration spectrum produced 6.78 g_rms. These results demonstrate a reproducible numerical proof of concept and define practical limits for compensation, spectral allocation, curvature interpretation, and inverse reconstruction; they do not constitute experimental validation or flight qualification. Full article

Underground mine object detection is challenged by low illumination, blur, dust scattering, and repetitive tunnel clutter, which jointly corrupt backbone features, entangle DETR queries, and weaken localization for small objects. Existing enhancement-based and detector-internal methods do not explicitly propagate degradation reliability across features, decoder queries, and box refinement. We propose Spectral-DETR, a detector-internal reliability framework built on RF-DETR. Its central design is a cross-stage reliability pathway that connects Degradation-Aware Frequency Decomposition (DAFD), Degradation-Adaptive Query Contrastive Denoising (DQCD), and Salience-Calibrated Uncertainty with Learned Uncertainty Estimation (SCU+LUE). On Mine-Objects (14 classes, 3081 images), Spectral-DETR achieves an average precision of 0.917 at an intersection-over-union threshold of 0.5 and 0.493 when averaged over thresholds from 0.5 to 0.95, exceeding YOLOv9m by 1.6 and 0.8 percentage points, respectively, under the dataset-specific evaluation protocol. In controlled RF-DETR validation, the three reliability stages improve these two measures from 0.883 to 0.913 and from 0.472 to 0.486, respectively. Spectral-DETR obtains corresponding values of 0.848 and 0.571 on ExDark and 0.973 and 0.495 on ScienceDB. DQCD and SCU remain training-only losses with no inference cost. Full article

Stable discrimination can coexist with unreliable probabilities, abstention policies, and uncertainty sets under distribution shift. We evaluate this mismatch with CRIT-AID, an executable multi-domain reliability-audit framework that separates model fitting, probability calibration, operating-rule calibration, and testing and then transports source-derived rules unchanged to target domains. Across four public tabular domains, stable discrimination did not imply stable probability quality, selective operating points, or conformal uncertainty. On identical ACS 2024 records, changing the income target definition left AUROC nearly unchanged, while ECE differed by 0.083; prevalence-intercept alignment reduced this difference to −0.004, showing that target semantics can alter probability reliability without materially changing ranking. Across 27 primary 90% conformal conditions, label-conditional calibration improved worst-class coverage in 18 but worsened it in 9 and usually increased prediction-set size. LightGBM sensitivity changed absolute discrimination without removing the mismatch among reliability dimensions. These findings show that discrimination alone is insufficient evidence for reliable AI decision support: audits should test the transportability of probability mappings, operating rules, class-specific validity, and uncertainty informativeness when deployment conditions or target meanings change. Full article

Afforestation can contribute to climate-change mitigation when tree establishment is matched to ecological context and sustained by long-term management, but its net climatic effect is not uniformly cooling. Existing syntheses often emphasize carbon sequestration while treating biophysical, hydrological, disturbance, and socioeconomic evidence separately. Here, we provide a structured integrative review. We distinguish afforestation from reforestation, natural regeneration, forest restoration, and improved management. We explicitly assess evidence from global modeling, remote sensing, meta-analyses, long-term observations, and regional case studies. Global forests cover about 4.14 billion ha in 2025, while annual net forest loss remained about 4.12 million ha yr⁻¹ during 2015–2025. Global forests were a sink of about 3.5 ± 0.4 Pg C yr⁻¹ in the 2010s, but this existing-forest sink should not be interpreted as an afforestation-specific removal rate. Humid tropical and subtropical regions generally have the greatest potential for net climatic cooling. In contrast, afforestation at snow-covered high latitudes may cause substantial albedo-driven warming, while water-limited regions require careful species selection and conservative planting densities. Soil carbon gains are most consistent on former croplands and other low-carbon degraded lands, but responses on carbon-rich grasslands are highly variable. Long-term benefits further depend on disturbance resilience, permanence, land competition, financing, and credible monitoring. Additionally, we identify five priorities for the future: climate-smart adaptive silviculture, digital forestry with field-calibrated uncertainty, permanence and disturbance-risk accounting, sustainable forest bioeconomy, and integrated international governance and finance. Full article

Sustainable rural cultural tourism requires effective approaches to improving the visibility, accessibility, and decision feasibility of dispersed cultural resources, particularly in destinations where service information is fragmented across online and offline channels and tourists face substantial uncertainty in coordinating transport, accommodation, and cultural experiences. This study examines how artificial intelligence (AI) agents can support the sustainable digital transformation of rural cultural tourism by alleviating information asymmetry, releasing latent tourism demand, and facilitating calibrated human–machine trust. Drawing on human–machine trust theory and the Stimulus–Organism–Response framework, this study conceptualizes AI agent functionality through three dimensions: AI Information Quality (AIQ), Information Extensibility (IE), and AI Planning Autonomy (APA). Travel Planning Risk Awareness (TPRA), Human–Machine Trust (HMT), Planning Satisfaction (PS), Rural Cultural Tourism Attractiveness (RCTA), and Rural Cultural Tourism Intention (RCTI) are further incorporated into an integrated model comprising four pathways: information empowerment, autonomy–risk awareness tension, trust boundary, and demand release. Using the Ctrip AI Travel Assistant as the research context, 413 valid questionnaire responses were analyzed through a hybrid Structural Equation Modeling–Artificial Neural Network approach. The results support 12 of the 14 hypotheses. AIQ significantly influences IE (β = 0.530), PS (β = 0.304), and HMT (β = 0.380). HMT functions as a central mechanism connecting AI empowerment with tourism decision-making and exerts the strongest effect on RCTA (β = 0.485), reaching 100% normalized importance in the corresponding ANN model. TPRA positively affects HMT (β = 0.262), indicating that risk awareness can facilitate rational and calibrated trust rather than simply inhibiting AI acceptance. RCTA (β = 0.281) and PS (β = 0.218) jointly promote RCTI through the complementary mechanisms of destination pull and planning push. The findings demonstrate that AI agents can contribute to the sustainable development of rural cultural tourism by improving information accessibility, strengthening responsible human–AI collaboration, and transforming fragmented cultural resources into credible and actionable travel-planning options. This study provides implications for sustainable destination marketing, responsible AI travel-service design, rural revitalization, and the long-term development of rural cultural tourism, while clarifying trust as a psychological gate in AI empowerment. Full article

This exploratory cluster-randomized mixed-methods study examined reading development and learner experience across AI-adaptive, teacher-differentiated, and non-differentiated EFL reading instruction. Eighty-seven university-level EFL learners, classified by CEFR proficiency, completed a 12-week intervention across four intact classrooms: two AI-adaptive, one teacher-differentiated, and one non-differentiated. Reading development was assessed using equated parallel-form IELTS-format tests calibrated through Rasch modeling and CEFR cut-score mapping. Learner experience was assessed using multi-item measures with satisfactory internal consistency and preliminary evidence of structural validity for cognitive load, engagement, instructional comfort, and perceived ownership of reading achievement, supplemented by qualitative reflections. Because instructional condition was assigned at the classroom level and only four clusters were available, reading development patterns were interpreted primarily through descriptive comparison of the recoverable model-implied pre-to-post patterns, with participant-level mixed-design ANOVA and cluster-adjusted mixed-effects modeling reported as secondary analyses. The pooled AI-adaptive classrooms showed a larger model-implied pre-to-post increase than the comparison classrooms, while both secondary analyses showed the same directional pattern but substantial uncertainty. Descriptively, learners in the AI-adaptive classrooms reported lower cognitive load and higher engagement, instructional comfort, and perceived ownership of reading achievement. These findings provide preliminary classroom-based evidence of favorable patterns associated with AI-adaptive personalization but do not establish causal or generalizable instructional effects. Larger cluster-randomized studies are needed to separate instructional effects from classroom-level influences and examine proficiency-related differences. Full article

Accurate finite element models are essential for structural condition assessment, yet nominal material properties and idealized boundary conditions can produce systematic discrepancies between numerical and measured dynamics. This study proposes a physics-constrained sparse response-surface framework that combines Elastic Net basis selection, mechanically prescribed monotonicity, adaptive sample enrichment, and identifiability-aware uncertainty assessment within a transparent finite element model-updating procedure. A scaled steel truss was tested using millimeter-wave radar, and the first three vertical natural frequencies were identified by stochastic subspace identification. The resulting sparse polynomial surrogate was independently validated before bounded inversion and ANSYS back-substitution. The mean frequency error decreased from 5.55% to 0.82%. Jacobian and bootstrap analyses further showed that several combinations of material and boundary parameters can reproduce similar modal responses, so the updated parameters are best interpreted as a coupled equivalent calibration state rather than unique direct measurements. The proposed framework therefore improves physical consistency and computational efficiency while explicitly retaining the uncertainty associated with weakly identifiable parameter directions. Full article

Ultra-high-pressure and extreme-pressure metrology plays a critical role in advanced manufacturing, geoscience, high-pressure physics, and frontier materials research. As pressure ranges extend from hundreds of MPa to several GPa and beyond, pressure generation and measurement are increasingly constrained by material strength, structural deformation, the state of pressure-transmitting media, sealing reliability, sensor drift, and incomplete traceability chains, imposing higher requirements on pressure metrology. This review systematically examines measurement technologies and traceability routes for ultra-high-pressure and extreme-pressure ranges. The development of controlled-clearance piston gauges is summarized, with emphasis on uncertainty reduction, range extension, and calibration automation. Drop-weight-based primary standards for dynamic pressure are also reviewed as an established route for the traceable calibration of high-amplitude, millisecond-scale hydraulic pressure pulses. Progress in secondary ultra-high-pressure standards is reviewed, including ultra-high-pressure gauges and piezoresistive, resonant, fiber-optic, and triboelectric sensors, with focus on range extension, high-accuracy measurement, environmental adaptability, and emerging pressure-sensitive mechanisms. Measurement methods for extreme pressure, including ruby fluorescence, Raman spectroscopy, X-ray diffraction, phase-transition points, and equations of state, are compared in terms of applicability and limitations. Current challenges in ultra-high-pressure and extreme-pressure metrology are further discussed, together with the prospects of quantum pressure sensing based on nitrogen-vacancy centers in nanodiamonds, providing a reference for future research and metrological infrastructure development. Full article

Accurate State-of-Health (SOH) prediction is essential for deploying retired electric vehicle batteries into reliable second-life energy storage systems. However, this task is challenged by sparse and noisy operational data from onboard Battery Management Systems (BMS). This study systematically evaluates a spectrum of deep learning architectures for SOH forecasting under BMS-style data constraints derived from laboratory cycling data: a BiLSTM on aggregated cycle statistics (Model A), preliminary zero-shot transfer to a single unseen cell (Model B), a waveform BiLSTM with full intra-cycle voltage, current, and temperature trajectories (Model C), a baseline TCN (Model D) and a probabilistic TCN-GPR hybrid (Model E). All models are constrained to identical low-fidelity BMS-style variables extracted from the NASA battery aging dataset. Model C achieves the lowest point accuracy error of 0.46% ± 0.18% MAE across five random seeds, demonstrating that high-resolution waveform inputs capture degradation signatures, notably voltage plateau morphology, transient dynamics, and implicit SOC information, that aggregated features irreversibly lose. Model D using the same waveform inputs and evaluation protocol as Model C, achieves a MAE of 2.99% at a single seed, providing direct architectural comparison evidence that the BiLSTM’s position-sensitive temporal summarization outperforms GlobalAveragePooling1D under these conditions. Model E achieves a higher MAE of 2.12% ± 0.33% but uniquely provides calibrated predictive distributions of 99.4% ± 1.2% coverage, NLL = −1.877 ± 0.038, with approximately uniform 95% predictive intervals (mean width 19.83% SOH across 34 test cycles at seed = 42), reflecting the near-constant posterior variance produced by the large optimized GPR length-scale under the frozen two-stage training design. A paired t-test confirms that Model C statistically significantly outperforms Model E on point accuracy (p < 0.01). Isotonic regression recalibration reduces mean calibration error from 0.138 to 0.010, demonstrating that shape-level miscalibration is correctable post hoc. The central implication for second-life battery deployment is a clear accuracy–uncertainty trade-off: Model C is preferred when point estimates suffice, while Model E is essential for risk-aware decisions requiring confidence intervals. Full article

This paper presents the metrological characterization of low-cost sensors integrated into a multi-sensor system for Indoor Environmental Quality monitoring, developed within the MIRABLE project. The analysis focuses on two domains: the thermal domain, using Sensirion SHT45 and SEN55 temperature sensors; and the Indoor Air Quality (IAQ) domain, using an Infineon photoacoustic spectroscopy (PAS)-based sensor for carbon dioxide (CO₂). All tests were conducted under controlled laboratory conditions using calibrated reference instruments. In the thermal domain, the influence of sensor integration within the device case was investigated at temperature (T) between 15 °C and 35 °C and relative humidities (RH) between 30 %rh and 60 %rh. The results revealed self-heating effects in the desk unit, causing temperature biases of up to 0.6 °C. In the IAQ domain, the repeatability and the impact of T and RH on CO₂ measurements were evaluated. T was identified as the main influencing factor; whereas, RH had a negligible effect. These results were supported by statistical analysis ANOVA. A correction strategy based on concentration intervals is proposed for operation between 15 °C and 25 °C, with an expanded uncertainty (k = 2) of (3.26–6.42) ppm for the with-case configuration. These results support the reliable use of the MIRABLE system. Full article

Large mineral flotation workshops in severe cold regions represent special industrial indoor environments characterized by the coexistence of limited ventilation and intense heat release. Such conditions generate pronounced spatial thermal stratification and localized heat accumulation within the workshop, leading to uneven worker thermal exposure and increased thermal discomfort and heat stress risk. However, conventional thermal comfort models were primarily developed for ordinary buildings with relatively stable thermal environments. Their applicability to large industrial workshops remains insufficiently validated. Nine representative monitoring points were arranged in the summer operating areas of the workshop, and thermal comfort surveys were conducted among 35 workers who had adapted to the local climate and working environment. The predicted mean vote (PMV) model was used as the baseline assessment framework, while an adaptive predicted mean vote (aPMV) model was further calibrated using field-based thermal sensation information. Monte Carlo simulation was employed to evaluate uncertainty propagation under field-data constraints, and Sobol sensitivity analysis was conducted to identify the dominant factors affecting thermal comfort predictions. The results demonstrated that the conventional PMV model exhibited a clear warm prediction bias under the investigated industrial conditions. After adaptive correction, the deviation from the field-based TSV was reduced by 82.93%, indicating improved agreement with workers’ actual thermal perception. Sensitivity analysis identified metabolic rate as the dominant contributor to aPMV output variance, with first-order and total-effect Sobol indices of 0.530 and 0.535. The proposed framework provides a scenario-specific approach for thermal comfort assessment in the investigated flotation workshop and offers preliminary methodological references for similar large-scale flotation workshops. Full article

Concurrent high river water temperature (RWT), low dissolved oxygen (DO), and low river flow (RF) can degrade water quality and aquatic habitat, but their joint probability is rarely quantified in a fully trivariate framework. This study develops a semiparametric trivariate drawable-vine (D-vine) copula model that uses Gaussian kernel density estimation (GKDE) margins with parametric pair-copulas to estimate compound thermal–oxygen–low-flow hazards. The framework provides conditional exceedance probabilities and AND- and OR-joint return periods (RPs) for monthly synchronized RWT–RF–DO states. The analysis uses 117 synchronized monthly triplets from the Tungabhadra and 152 from the Yamuna. Kendall’s $\tau$ values for RWT–RF and RF–DO are $-0.12$ and $+0.12$, respectively, at the Tungabhadra, compared with $+0.24$ and $-0.24$ at the Yamuna, indicating contrasting basin-specific dependence pathways. Candidate D-vine orderings are evaluated through permutation-based centred-variable selection and compared with a minimum-spanning-tree heuristic. The final selection uses information criteria, scoring rules, and calibration diagnostics. The selected structure is RF-centred for the Tungabhadra and DO-centred for the Yamuna. Bootstrap analysis supports the stability of the principal dependence structure but shows higher uncertainty in some conditional tail components. For an RWT threshold of $30°\mathrm{C}$, conditioned on DO and RF below their fifth percentiles, the fitted exceedance probability is approximately 0.90 at the Tungabhadra and 0.99 at the Yamuna. Within the available records, the Yamuna shows higher fitted co-occurrence probabilities and generally shorter trivariate AND-joint recurrence intervals than the Tungabhadra. The findings support risk-based screening of warm, low-flow, oxygen-stressed periods while emphasizing the need for local recalibration of margins, dependence structures, and management thresholds. Full article