Abstract: Deploying foundation models in embodied edge systems is fundamentally a systems problem, not just a problem of model compression. Real-time control must operate within strict size, weight, and power constraints, where memory traffic, compute latency, timing variability, and safety margins interact directly. The Deployment Gauntlet organizes these constraints into eight coupled barriers that determine whether embodied foundation models can run reliably in practice. Across representative edge workloads, autoregressive Vision-Language-Action policies are constrained primarily by memory bandwidth, whereas diffusion-based controllers are limited more by compute latency and sustained execution cost. Reliable deployment therefore depends on system-level co-design across memory, scheduling, communication, and model architecture, including decompositions that separate fast control from slower semantic reasoning.

Abstract: Engineer-to-Order (ETO) manufacturers face persistent cost and complexity challenges driven by product variety, including duplicate components, redundant variants, and inconsistent procurement setups. Although enterprise resource planning (ERP) and product lifecycle management (PLM) systems contain detailed Bills of Materials (BOMs) and procurement records, they typically lack portfolio-wide support for systematic cross-product commonality analysis without substantial manual effort. Prior approaches are either conceptual (e.g., indices and modularity frameworks) or ad hoc in practice, often relying on one-off spreadsheet analyses. This paper introduces the concept of Product Commonality Analysis Tools (PCATs) and develops and evaluates a lightweight PCAT in an action-research collaboration with a European ETO laser manufacturer. The PCAT operates on exported enterprise data to provide interactive portfolio-level views of component reuse and cross-product consistency. Its usefulness is evaluated through scenario-based think-aloud usability sessions and a functional comparison against Excel workarounds, standard ERP/PLM reporting, and vendor customizations. The results indicate that a lightweight PCAT can be integrated into existing ERP/PLM workflows with minimal disruption and can reduce the effort required to prepare reusable portfolio views for engineering and procurement reviews.

Abstract: Polyfluoroalkyl substances (PFAS) are synthetic compounds shown to be associated with metabolic disturbances in the experimental literature. Evidence of the relationship between PFAS and MetS from human epidemiological studies remains inconclusive and warrants further study. This study leverages a pooled index to examine associations between a mixture of PFAS and metabolic syndrome in a sample of adults in the United States. Using data from the National Health and Nutrition Examination Survey 2005-2018 (n= 8095), we examined the relationship between serum concentrations of perfluorohexanesulfonic acid, perfluorononanoic acid, perfluorodecanoic acid, perfluoroundecanoic acid and 2-(N-Methyl-perfluorooctane sulfonamido acetic acid, and MetS. We evaluate individual associations with logistic regression and joint associations in a pooled index (PI) model. One standard deviation increase in the PI was associated with 18% decrease in odds of MetS (OR: 0.82, 95% CI: 0.76, 0.89). In logistic regression models, higher PFAS concentrations were also associated with decreased odds of MetS in perfluorodecanoic acid (PFDA) (OR: 0.43, 95% CI: 0.28, 0.64) and perfluoroundecanoic acid (PFUA) (OR: 0.19, 95% CI: 0.11, 0.36). This study found an inverse association between serum PFAS concentration and MetS, in both pooled and individual models; however, given the cross-sectional design, these findings should be interpreted cautiously.

Abstract: We study the long-time behavior of nonlinear stochastic evolution equations in a separable Hilbert space driven by a Q-Wiener process. The linear part of the equation is generated by a strongly continuous semigroup with an exponential dichotomy, which provides fixed rates of decay and growth. The nonlinear drift and diffusion terms are globally Lipschitz and become small as time tends to infinity. Our main result shows that under these conditions, the mean-square Lyapunov exponents of the nonlinear system coincide with those of the linear part. In other words, nonlinear stochastic perturbations that decay in time do not change the main growth or decay rates of solutions in the mean-square sense. This result provides simple and verifiable criteria ensuring that the long-time Lyapunov behavior of the nonlinear stochastic equation is fully determined by the linear semigroup, even in the presence of time-dependent stochastic perturbations.

Abstract: This study addresses the challenges of financial risk early warning by proposing a modeling approach based on spatiotemporal Transformers. The research first examines the multidimensional characteristics of financial risk, emphasizing its temporal dynamics and cross-regional interactions. It notes that many existing methods struggle to jointly capture temporal dependencies and inter-regional risk transmission patterns. To overcome these limitations, a unified spatiotemporal modeling framework is developed. The framework integrates temporal encoding, spatial adjacency information, and multi-head attention mechanisms to model long-range dependencies and regional spillover effects. In the model architecture, an embedding layer is employed to learn representations from multi-source financial indicators. A self-attention mechanism facilitates global feature interaction, while a graph convolution component further enhances the modeling of spatial relationships across markets. The final risk representation is generated through a feed-forward network with normalization layers, providing a structured basis for financial risk assessment and early warning analysis. Experimental evaluations include comparative studies and sensitivity analyses under varying missing data ratios, time window settings, and environmental conditions. The results indicate that the proposed method consistently outperforms several baseline models in terms of accuracy, precision, recall, and F1-score. Overall, the approach demonstrates strong robustness and practical applicability in complex financial settings, offering an effective tool for financial risk monitoring and decision support.

Abstract: Low-voltage distribution networks (LVDNs) serve as the final delivery layer of the electricity system, directly influencing reliability, public safety, customer service quality, and the integration of distributed energy resources. Despite their importance, LVDNs have historically received less monitoring than transmission and medium-voltage systems due to their scale, cost, and deployment complexity. Non-contact magnetic sensing has emerged as a promising alternative to invasive measurement methods for these networks. Among magnetic sensor types, giant magnetoresistive (GMR) devices are appealing because they offer high sensitivity, compactness, low power consumption, and compatibility with embedded electronics. This review assesses the current state of GMR-based monitoring for overhead and low-voltage applications, focusing on non-contact current measurement, fault detection, and fault classification. It first examines the operating characteristics of LVDNs and the unique challenge of detecting low- and high-impedance faults. Next, it outlines the physical principles behind GMR sensing, compares GMR with Hall, AMR, TMR, current transformer, and Rogowski-coil technologies, and discusses the use of multi-axis sensor heads to address cross-coupled magnetic fields in three-phase setups. Special focus is given to calibration, alignment, temperature effects, electromagnetic interference, packaging, wireless deployment, and data-driven classification. The review concludes that GMR sensors are well-suited for scalable, non-contact monitoring, but widespread adoption in the field will require better low-voltage fault datasets, standardized calibration procedures, long-term environmental testing, and closer integration with digital-twin and smart-meter infrastructures.

Abstract: In this work, the design and experimental validation of passive UHF RFID tag antennas are presented with the objective of evaluating the impact of chip placement and miniaturization approaches on tag performance. Four initial antenna layouts were developed by varying the position of the RFID integrated circuit within a coupling loop. Simulations and measurements confirmed that Antenna 1 achieved the best impedance matching, with a minimum reflection coefficient of −40 dB at 866 MHz and a power sensitivity of −16.3 dBm. Based on this reference design, a miniaturized version (Antenna 5) was obtained by integrating meander lines and capacitive end-loading, reducing the physical size while maintaining resonance at 866 MHz. Both structures were fabricated and evaluated using a Voyantic Tagformance measurement system, with read-range measurements performed under freespace conditions and in proximity to dielectric and metallic materials. The results demonstrated a maximum read range of 8.6 m for Antenna 1 in free space, while Antenna 5 preserved a read range of 6.3 m. In the presence of copper, Antenna 1 maintained a read range of 3 m, whereas Antenna 5 achieved approximately 0.5 m, confirming the robustness of the proposed designs in representative industrial environments.

Abstract: Beef production is widely recognized as a significant contributor to global greenhouse gas emissions, making robust and transparent environmental assessments essential for advancing sustainability within supply chains. This study applies a comprehensive cradle‑to‑grave Life Cycle Assessment (LCA) to evaluate the environmental performance of beef destined for export, following ISO 14040, ISO 14044 and ISO 14067 standards and the Product Category Rules for meat of mammals. Sixteen impact categories were quantified for 1 kg of vacuum‑packed beef using detailed primary data from a pasture‑based production system and a representative processing facility. The total climate change impact was 3.27×10¹ kg CO₂eq, with enteric methane and feed production jointly responsible for over 70% of overall impacts. Slaughtering and distribution were associated mainly with fossil energy use and ozone depletion, while soil carbon sequestration partially compensated biogenic emissions. The results were consistent with international benchmarks, highlighting the environmental advantages of pasture‑based systems, low fertilizer use, and stable land management. Key hotspots were identified in animal growth, feed efficiency, and manure management, with logistics also contributing notably. Overall, the study provides a high‑resolution environmental baseline that can support Environmental Product Declarations and guide targeted mitigation strategies across beef supply chains.

Abstract: Modern e-commerce platforms must handle sudden and unpredictable traffic surges caused by flash sales, festive shopping events, and viral online activity. Traditional web architectures typically adopt one of two extremes: a tightly coupled monolithic design that provides low latency but becomes fragile under heavy load, or a loosely coupled microservices architecture that improves scalability and resilience but introduces communication overhead during normal operation. This trade-off forces system designers to choose between performance efficiency and scalability robustness. This paper introduces ATLAS (Adaptive Traffic-aware Loose–tight Architecture System), a next-generation adaptive web architecture that dynamically adjusts its coupling strategy based on real-time system conditions. ATLAS employs machine learning models to analyse operational telemetry, predict traffic surges, detect anomalies, and forecast potential system failures. Using these predictions, the architecture can automatically transform its runtime structure, switching between tightly coupled monolithic execution and loosely coupled microservices deployment as traffic conditions evolve. To improve reliability, ATLAS incorporates a self-healing recovery pipeline that autonomously detects service failures, isolates faulty components, and restores normal operation without human intervention. Through case studies of large-scale platforms such as Google Search, Amazon, and Flipkart, we illustrate how existing systems can evolve toward the ATLAS paradigm, enabling self-adaptive and resilient web infrastructures for the next generation of large-scale online services.

Abstract: Environmental noise is increasingly recognized as a major environmental development challenge, with road traffic identified as the dominant source of acoustic pollution across Europe. Noise barriers are among the most widely implemented mitigation strategies. However, their spatial distribution and adequacy remain poorly documented, limiting their effectiveness for sustainable territorial planning. This study develops the first georeferenced database of highway noise barriers in Andalusia (Spain) and applies a reproducible, transdisciplinary geospatial workflow integrating field surveys, remote-sensing tools, and Geographic Information Systems (GIS). A total of 110 barriers were mapped, classified by material, geometry, and surrounding land use, and analyzed in relation to dwellings, schools, and hospitals. Results show that 1.6% of the Andalusian highway network is currently protected by barriers, with strong territorial disparities: over 50% of all structures are concentrated along coastal metropolitan corridors, while extensive inland areas remain unprotected. Misalignments were also detected between barrier placement and officially reported high-exposure segments, indicating limited correspondence between infrastructural deployment and acoustic priorities. Beyond generating a comprehensive regional dataset, the methodology provides a scalable basis for national and European initiatives seeking to harmonize the mapping and assessment of noise-mitigation infrastructures. By offering an open-access, transferable framework, this work supports policy professionals, environmental managers, and planners in evaluating mitigation gaps and informing more equitable and sustainable transportation and land-use strategies.

Abstract: Background: Atrial fibrillation (AF) frequently coexists with heart failure (HF) and worsens clinical outcomes. However, predictors of AF in HF with preserved (HFpEF) and mildly reduced ejection fraction (HFmrEF) remain poorly defined. This study aimed to identify clinical, laboratory, and echocardiographic predictors of AF in these HF phe-notypes. Methods: This retrospective single-center observational study included 700 consecutive patients with HF hospitalized between January 2018 and December 2023. The median age was 74 years (IQR 66–80). Women predominated in the cohort (55.3% vs. 44.7%, p < 0.001). Based on echocardiographically assessed left ventricular ejection fraction, patients were stratified into groups with preserved (≥50%), mildly reduced (41–49%) and reduced (≤40%) ejection fraction. Predictors of AF were evaluated using univariate and multivariate lo-gistic regression analyses, and model discrimination was assessed using ROC analysis. Results: Strongest predictors of AF in our patients with HFpEF and HFmrEF were left atrial size (OR 1.114 per mm increase; 95% CI 1.054–1.177; p < 0.001), moderate and severe tricuspid regurgitation (OR 4.092; 95% CI 1.977–8.466; p < 0.001 and OR 6.957; 95% CI 2.482–19.499; p < 0.001), male gender (OR 1.680; 95% CI 1.076–2.621; p = 0.022) and advanced age (OR 1.070 per year; 95% CI 1.032–1.109; p < 0.001). Conclusions: In patients with HFpEF and HFmrEF, AF is strongly associated with atrial remodeling, with left atrial enlargement as the key structural predictor. Identification of high-risk patients using clinical and echocardiographic parameters may facilitate earlier AF detection and improved risk stratification.

Abstract: Deep learning-based tumor segmentation has achieved strong performance on benchmark datasets, yet models often degrade when deployed in new hospitals. This decline is largely driven by domain shift, including differences in scanners, acquisition protocols, reconstruction settings, patient populations, and annotation styles. In high-stakes clinical workflows, such instability limits real adoption because a model that performs well in one center may fail silently in another. This paper presents a preprint-ready methodological framework for domain shift-robust segmentation in multi-hospital MRI and CT tumor imaging. The proposed design combines four complementary ingredients: strong segmentation backbones from the U-Net family, domain-generalization through intensity, style, and frequency-based augmentation, self-supervised pretraining on unlabeled multi-site data, and optional label-free test-time adaptation for target hospitals. The manuscript emphasizes a deployment-oriented evaluation protocol that prioritizes worst-site reliability, boundary safety, calibration, and failure analysis rather than average Dice alone. We describe an experimental plan with leave-one-hospital-out validation, targeted ablations, uncertainty analysis, and stress tests under artifact corruption. The expected pattern is that self-supervised pretraining and frequency-aware augmentation reduce the gap between in-domain and out-of-domain performance, improve worst-site Dice, and lower extreme boundary errors measured by Hausdorff distance. The central argument is that robustness should be treated as a first-class objective in medical image segmentation and that multi-center validation, transparent reporting, and clinically meaningful error analysis are necessary before deployment.

Abstract: Anthocyanins and betalains are hydrophilic plant pigments with numerous physiological and ecological functions. The biosynthesis routes of anthocyanins and betalains differ with anthocyanins being synthesized from phenylalanine via the general phenylpropanoid pathway, whereas betalains are derived from tyrosine. Although the precursors phenylalanine and tyrosine are present in all plants, there is no known plant where both these pigments are co-accumulated. Most plants synthesize anthocyanins, while certain families in the order Caryophyllales produce betalains. There is apparent mutual exclusion of these two plant pigments. Over the past five decades, evidence accumulated supporting this theory of mutual exclusion of the two pigments. However, recently published reports claim the presence of anthocyanins in well-known betalain-pigmented plants. Here, we explore the causes of such claims and provide recommendations for future studies on the topic.

Abstract: By using Abel's transformation we study the correlation of generalized divisor functions d_k(n) and obtain the correct main term order of the asymptotic estimate for the correlations.

Abstract: The armless snake eel, Dalophis imberbis, is a fossorial rare species; being considered as a non-target fishery resource with elusive behavior, knowledge on its distribution and biology results scarce. This study reports three new documented occurrence records of D. imberbis along the northern and southeastern coastal areas of Sicily (central Mediterranean Sea) during 2025. Specimens were collected at depths ranging from 43 m to an unusually shallow depth of 5.4 m. Environmental parameters have been collected through a multiparametric probe and integrated with products from the Copernicus Marine Service (CMS), providing new insights which highlight the presence of the species in relatively warm (17.6-20.8 °C) and moderately oxygen undersaturated (6.9-8.5 mg/L) waters. A global distributional analysis was performed by aggregating the field data with literature records and datasets from the Global Biodiversity Information Facility (GBIF), refining the distribution of the species in the Mediterranean and Atlantic sectors. This work underscores the importance of combining traditional surveys with big-data repositories and remote sensing to monitor rare marine biodiversity.

Abstract: Based on the Multi-Sensor Reanalysis Version 2 dataset, this study analyzes variations in monthly mean total ozone column (TOC) over Northeast China (40–53°N, 115–135°E) during 2015–2024. The study area in winter lies in the transition zone between high polar and low subtropical TOC in East Asian mid-latitudes. Key results indicate that the TOC over Northeast China is consistently higher than the zonal mean TOC of the same latitude band and seasonal cycle demonstrates TOC maximum (minimum) in February (August), one month (two months) earlier than for the Northern Hemisphere midlatitudes. The important role of Brewer–Dobson circulation and quasi-stationary wave (QSW) structure in the TOC distribution over Northeast China is confirmed by the 10-year climatology for January–March. The QSW pattern is characterized by the TOC decrease from the northeastern (~415 DU) to southwestern (~330 DU) parts of the region. The strongest positive (negative) correlations approaching r = 0.9 (r = –0.8) exist between TOC and ozone concentration (temperature) at 50 hPa and 100 hPa, as well as at the surface. These findings can be applied to analyze the ozone observations and stratosphere–surface couplings in the Northeast China region.

Abstract: Background/Objectives: Spontaneous intracranial hypotension (SIH) is caused by spinal cerebrospinal fluid (CSF) leakage and is typically diagnosed by clinical presentation and characteristic MRI signs; however, objective tools for monitoring physiological alterations and treatment response remain limited. Cine phase-contrast MRI (PC-MRI) enables noninvasive quantification of aqueductal CSF dynamics, yet reliable analysis is challenging since the cerebral aqueduct is extremely small and susceptible to low contrast, partial volume effects, and ROI-dependent measurement variability—particularly in SIH where CSF pulsatility is often reduced. Methods: We propose an end-to-end automated framework that integrates (1) a cascade localization–segmentation strategy (Tiny YOLOv4 detection followed by MultiResUNet segmentation on a cropped ROI form YOLOv4 result; (2) physiology-informed pulsatility-based segmentation (PUBS) to refine anatomical masks into functional flow ROIs, and (3) one-dimensional convolutional neural networks (1D-CNNs) to learn waveform morphology biomarkers from 32-phase cardiac-cycle velocity waveforms. The study includes 39 participants (11 controls; 28 pre-treatment SIH; 20 post-treatment recovery). Results: The cascade model significantly improves segmentation robustness compared with a full-image baseline, achieving higher Dice scores and markedly lower boundary errors across cohorts (e.g., pre-treatment SIH HD95: 1.66 ± 0.74 px vs. 15.37 ± 44.98 px). PUBS refinement reduced quantification deviation from expert manual references in SIH (mean relative error: 7.4% to 5.6%) and improved diagnostic performance for multiple hemodynamic parameters (e.g., downward mean flow AUC: 0.747 to 0.792). Importantly, waveform morphology learning substantially outperformed conventional scalar metrics: peak systolic CSF velocity (PSV) showed limited discrimination (AUC 0.69), whereas the full morphology AI model achieved AUC 0.96. A single trait-like morphology feature reached AUC 0.98 with 100% specificity, while a state-dependent feature normalized after recovery, indicating complementary utility for diagnosis and longitudinal monitoring. Conclusions: This fully automated, reproducible, and physiologically informed pipeline demonstrates that SIH-related information is not only reflected in flow magnitude but also encoded in subtle aqueductal CSF waveform shape patterns, supporting morphology-based biomarkers as a promising tool for SIH assessment and follow-up.

Abstract: The global monetary order is shifting and the shift is not subtle. The U.S. dollar, long unchallenged at the center of international finance, now faces serious pressure from three converging forces: geopolitical fractures, financial sanctions, and deliberate moves by central banks to diversify their reserve holdings. This study examines, with empirical rigor, how de-dollarization has progressed and how gold's monetary role has evolved between 2000 and 2026. Drawing on high-frequency data from the IMF's COFER database, the World Gold Council, and the International Financial Statistics database. The analysis integrates multiple authoritative sources to build a coherent picture of reserve composition globally. To capture dynamics that standard regression would miss, the study uses a three-part methodological framework quantile-on-quantile (QQ) regression, causality-in-quantiles testing, and descriptive trend analysis. Together, these tools allow us to examine how gold and equity markets behave under different conditions: calm periods, bull runs, and crises. The dollar's share of global reserves has fallen to 57.74%a meaningful retreat from its earlier dominance. Even more notable. Official gold holdings worldwide now stand at $3.909 trillion, nearly matching the $3.920 trillion held in U.S. Treasury securities by foreign governments. For the first time in modern monetary history, gold and Treasuries are effectively at parity in central bank portfolios. However, this gold accumulation is not broadly distributed. It is concentrated in a handful of emerging economies notably Russia (1,894 tonnes), China (1,807 tonnes), Turkey (705 tonnes), and India (523 tonnes). Importantly, most of these countries are not systematically reducing dollar holdings alongside their gold purchases. The exception is where geopolitics directly drove the decision, as in Russia's case. The QQ regression uncovers a clear and counterintuitive pattern: gold's relationship with equity markets is U-shaped. Gold delivers its strongest gains at the extremes either as a safe-haven asset during sharp stock market crashes (β = −3.37 at τ = 0.10, θ = 0.10), or during exceptional equity bull markets (β = +3.16 at τ = 0.95, θ = 0.90). In between, in ordinary or muted market conditions, gold's role is far less pronounced. Causality-in-quantiles testing reinforces this picture. Stock market returns predict gold outcomes only when gold is already performing exceptionally well that is, in the upper tail of gold's distribution (τ = 0.90, p = 0.046; τ = 0.95, p = 0.013). In normal conditions, no such predictive link emerges. This tells us something important: gold does not respond to markets uniformly it responds selectively, and only under specific regime conditions. Taken together, these findings point to a monetary system in gradual but genuine transition. It is becoming more pluralistic but carefully so. For most countries, this reflects portfolio diversification, not a deliberate campaign to dethrone the dollar. Gold's role, meanwhile, is not universal it is regime-dependent, activated by crisis or exceptional growth, and largely dormant in between. The dollar remains dominant however it is no longer unquestioned.

Abstract: Background: The rapid evolution of omnichannel retailing has reshaped retail supply chains (SCs) by tightly coupling replenishment, fulfillment, and service decisions across multiple demand channels under inventory, lead-time, and capacity constraints. These interdependencies create complex coordination challenges, particularly when demand shocks interact with limited operational capacity. Methods: To address these challenges, this study develops a centralized Hierarchical Reinforcement Learning (HRL) control framework that makes decision timing explicit: replenishment and allocation are optimized weekly, while fulfillment and lateral inventory rebalancing are controlled daily. Policies are learned using Proximal Policy Optimization (PPO) in an actor–critic architecture with bounded stochastic policies suitable for constrained action spaces. To mitigate the curse of dimensionality often encountered in HRL, we introduce a capacity-aware state–action encoding mechanism that compresses the control interface into structured summary signals. Demand shocks are modeled using two specifications: a mixed regime where half the products follow uniform demand and half follow a Merton-type jump-diffusion process, and a fully shock-driven regime. Results: The framework is evaluated against forecast-driven base-stock and greedy fulfillment heuristics, as well as a perfect-information oracle. Results show that the proposed encoding improves learning efficiency and scalability, achieving higher profit and service performance than the full-observation alternative. Conclusions: Overall, hierarchically timed control outperforms heuristic baselines while remaining below the oracle bound, with the largest gains observed when demand shocks coincide with binding fulfillment and transfer capacities.

Abstract: This Data Descriptor presents an anonymized, shuffled dataset of creatinine-normalized urinary metabolite measurements from 73 Bulgarian children with autism spectrum disorder (ASD), released to support reuse in secondary analyses and cross-cohort comparisons. Spot urine results are provided as individual-level values after creatinine normalization; for trimethylamine, values below the limit of quantification (LOQ) were replaced with LOQ/2. The deposit contains measurements for 24 urinary markers grouped into three functional classes (neurotransmitters and aromatic amino acid precursors; one-carbon/methylation and vitamin-related metabolites; and energy metabolism/organic acids with microbiome-related amines). The release includes the results table in both XLSX and CSV formats, a reference limits and units file for contextual interpretation, a data dictionary, a README, a changelog, and SHA-256 checksums for integrity verification. Cohort-level demographics and additional sampling details are described in the companion publication cited in the main text. Dataset: https://doi.org/10.5281/zenodo.18614881 Dataset License: Creative Commons Attribution 4.0 International.