Search Results (117)

Trail running imposes highly variable biomechanical demands due to steep, irregular terrain that renders flat-road pacing models inadequate. We present a physics-based digital twin that integrates a terrain-adaptive grade-adjusted pace (GAP) model with individualised physiological calibration to predict finish time across heterogeneous trail-running races. The GAP core applies Minetti’s fifth-degree metabolic cost polynomial to map slope-dependent energy cost across the full range of uphill and downhill gradients encountered in trail racing. Segment-by-segment pace is further modulated by an altitude–VO₂max correction, a Banister TRIMP-based fatigue term, and a progressive pacing-decay factor. Course-elevation profiles are extracted from 1 Hz barometric altimeter data through a five-step normalisation pipeline. Individual parameters (sustainable VT₂ fraction α; pacing-decay slope μ) were calibrated by grid search against 13 race sessions. A sequential validation across four model-complexity stages showed R² increasing from 0.763 to 0.905. Leave-one-out cross-validation (n = 13) yielded R² = 0.864, MAE = 18.2 min, MAPE = 11.1%, and a small positive bias (+2.0 min). The framework demonstrates that integrating biomechanical terrain correction with individual physiological calibration substantially improves race-time prediction for trail running, offering a promising foundation for athlete-specific pre-race simulation. Full article

Private cloud object stores provide infrastructure isolation but leave application-layer data exposed to insider threats and compromised credentials. This paper presents an engineering integration of an Add-Rotate-XOR (ARX) block cipher and multi-bit Least Significant Bit (LSB) steganography into an end-to-end pipeline for private MinIO object storage. The cipher, KREA v2, is a SPECK-64/128 derived ARX construction with three application-driven choices: CRC32 key whitening, byte-aligned rotations ($\alpha =7$, $\beta =2$), and deterministic CTR-mode nonces. Mixed Integer Linear Programming (MILP) trail analysis matches SPECK-64/128’s minimum-trail weights through rounds 1–4. KREA v2 ciphertext meets standard keystream-quality preconditions (NIST SP 800-22 battery, 49.98% mean avalanche, Shannon entropy 7.9992–7.9998 bits/byte across realistic XML, JSON, video, and HTTP/2 payloads). Modified LSB (MLSB) embeds 3 bits per RGB channel with an XOR watermark at 37–38 dB Peak Signal-to-Noise Ratio (PSNR), providing $3\times$ standard-LSB capacity. Steganalysis uses chi-square and RS detectors plus a Convolutional Neural Network (CNN) detector (Yedroudj-Net) trained on 8000 BOSSBase-1.01 cover/stego pairs; CNN area under the ROC curve is ≥0.999 against the watermarked variant. The MinIO pipeline runs at 355.1 ms (68.6% network I/O) with 100% message fidelity. The XOR watermark increases RS detectability above 75% capacity; a 200-image ablation cuts median RS detection (0.289 to 0.000) and mean (0.342 to 0.130) in a sparse-keystream variant, prioritised for follow-on full-scale evaluation. The architecture is offered as a documented engineering integration with explicit security caveats and threat-model boundaries, not as a production-hardened cryptographic primitive. Full article

Tool-augmented large language model agents are increasingly proposed for network configuration, but routing protocols differ in the control-plane state each commanded router can observe. This difference creates a specific problem for multi-agent orchestration: agents may coordinate more, yet still fail when correct verification depends on peer- or remote-router evidence. We study this interaction through 350 controlled runs on RIP, OSPF, and BGP tasks implemented with FRRouting and Containerlab, comparing a single-agent baseline with multi-agent orchestration patterns across language models. Protocol-centric trace metrics, including spatial coverage, coordination tax, and cross-router verification gap, are combined with intent-property scores and model-balanced bootstrap analysis. The results show that observability explains performance more clearly than orchestration patterns: multi-agent templates trail the baseline on local RIP feedback, show only small and uncertain gains on single-area OSPF troubleshooting, and remain near zero on stricter multi-area OSPF and BGP tasks where peer-side verification gaps are often complete. The main contribution is therefore a protocol-centered account of when agentic orchestration helps, when it adds coordination cost, and why current architectures face a cross-router verification ceiling. Full article

This study evaluated the validity of the leg stiffness metric provided by the Stryd running power meter against the Morin (2005) sine-wave spring–mass model. Twenty-three highly trained trail runners (11 women) completed a 12 min uphill time trial at +12% grade and one hour of submaximal level running. Leg stiffness was calculated from contact time, flight time, running speed, and leg length using Morin’s method, and compared with Stryd values. Agreement was assessed following the Dhahbi and Chamari Level-1 analytical framework, including intraclass correlation coefficient (ICC_2,1), Bland–Altman analysis, mean absolute percentage error (MAPE), and paired t-tests. Stryd and Morin estimates showed excellent agreement in both conditions: uphill running: ICC_2,1 = 0.96 (95%CI: 0.91–0.98), bias = −0.02 kN·m⁻¹, limits of agreement (LoAs) = [−0.61, 0.58] kN·m⁻¹, MAPE = 2.5% (p = 0.803); and level running: ICC_2,1 = 0.97 (95%CI: 0.93–0.99), bias = −0.04 kN·m⁻¹, LoAs = [−0.62, 0.54] kN·m⁻¹, MAPE = 2.6% (p = 0.505). The Stryd sensor provides valid leg stiffness estimates in highly trained trail runners on both level and inclined terrain. The negligible systematic bias and narrow limits of agreement support the use of Stryd for leg stiffness monitoring in field and laboratory settings. Full article

Dynamic fee market mechanisms, exemplified by EIP-1559, have been extensively studied for Ethereum’s 12 s block environment but remain uncharacterized at sub-second timescales. We present an agent-based simulation study of an EIP-1559 adaptation for Injective, a Layer 1 blockchain combining native EVM compatibility with CometBFT consensus, operating at 600 ms block times. Across twelve simulation runs (four parameter configurations × three demand scenarios), our analysis yields three findings: (1) temporal smoothing mechanisms (MA-25, 25-block trailing average) produce mixed effects in sub-second environments with up to 47% basefee overshoot during spam attacks and slight smoothing elsewhere, making per-block mechanisms preferable for consistent performance; (2) transitioning from 150M (66.66% target) to 300M (50% target) configuration reduces peak fees by 31% during variable demand; during spam attacks, the 300M configuration peaks 32% higher but recovers faster with block capacity as the primary driver for spam throughput; and (3) per-block mechanisms establish initial spam barriers within 17–32 s versus Ethereum’s 4–6 min, economically justifying lower minimum fees. We provide the first systematic sub-second EIP-1559 analysis and a parameter optimization framework for high-throughput chains. With proper tuning, dynamic fee mechanisms are compatible with high-throughput architectures. Full article

Generative AI systems increasingly mediate how students retrieve literature and generate citations, shifting methodological rigor toward the maintenance of an auditable evidence trail. This study audits the search stage of AI-assisted literature review work, focusing on retrieval performance and citation traceability rather than downstream screening or synthesis. Four widely accessible tools were compared across two retrieval postures, and Boolean queries were executed against Scopus and evaluated against a DOI-verified librarian baseline built from Scopus, Web of Science, and Google Scholar. Using a canonical prompt and a bounded top-k capture rule (k = 20), each bibliographic record was evaluated for DOI traceability, DOI resolution integrity, metadata accuracy, and run-to-run drift. Records were screened through staged title/abstract and full-text eligibility review, and the final set included 37 studies after quality appraisal was 37 studies. Across sixteen audit runs, natural-language prompting frequently produced under-target yields, recurrent integrity failures, and low overlap with the librarian benchmark. Boolean translation improved run completion and increased the proportion of auditable records, but reproducibility remained unstable across repeated runs. These findings show that correctness at the record level does not ensure stability at the evidence-set level. Limitations include the bounded tool set, the search-stage focus, and the absence of downstream screening or synthesis evaluation. Retrieval posture, therefore, emerges as a practical governance lever for AI-assisted literature review workflows and supports the use of a student-facing verification checklist anchored in DOI verification and transparent protocol capture. This research received no external funding. OSF registration: Open Science Framework, 10.17605/OSF.IO/U8NHT. The manuscript reports the final included set as n = 37, states no external funding, and lists the OSF registration DOI. Full article

Background: Hydration and electrolyte strategies are critical in mountain ultra-endurance events, yet field-based evidence from trail races remains limited. This study examined the relationship between fluid intake, sodium consumption, and body mass changes in trail runners competing under real environmental conditions. Methods: A cross-sectional field study was conducted during La Misión Brasil 2024. Athletes of both sexes competing in the endurance race (35 km; EG: n = 15; age = 37.0 [29.5–46.0] years; 12 men and 3 women) and the ultra-endurance race (80 km; UEG: n = 13; age = 42.0 [37.0–46.0] years; 11 men and 2 women) were included in the study. Pre- and post-race body mass were assessed, and in-race fluid and food intake were collected using an adapted 24-h dietary recall. Water and sodium intake were expressed as total (L and mg, respectively) and per-hour (mL/h and mg/h, respectively) values. Environmental temperature and humidity were obtained from a local weather station. Group comparisons were performed using the Mann–Whitney U test, and associations were examined with Spearman’s correlation (p < 0.05). Results: EG (n = 15) and UEG (n = 13) showed similar absolute and relative body mass changes (2.6% to −3.0%; p > 0.05). EG runners presented greater weight loss rate (−270 vs. −115 g/h; p = 0.002), while UEG consumed higher total water (7.11 vs. 4.14 L; p = 0.008) and sodium (5789 vs. 2857 mg; p = 0.003). Water intake per hour was higher in EG (626 vs. 427 mL/h; p = 0.017). Body Mass Index was negatively correlated with hourly weight loss (r = −0.605; p < 0.001). Water and sodium intake per hour were positively correlated (r = 0.607; p < 0.001), though neither predicted hourly weight loss. Conclusions: Hydration responses may differ according to environmental stress and pacing demands. Changes in body mass may not necessarily reflect hydration adequacy, suggesting a possible multifactorial nature of hydroelectrolyte balance during mountain endurance events. Full article

This study investigates why companies sponsor individual athletes in sports with low media coverage and how such athletes secure sponsorship agreements. While sport sponsorship research has predominantly focused on mainstream sports and event-based contexts, limited attention has been given to individual athletes in niche sports. Using a qualitative research design, semi-structured expert interviews were conducted with Norwegian sponsors and elite athletes in long-distance running, trail running, and orienteering. The data were analyzed through qualitative content analysis, informed by the Sponsorship Motive Matrix and the Model of Athlete Brand Image. The findings indicate that sponsorship decisions are primarily driven by market-related motives, complemented by bond and society motives, with cost-effectiveness, authenticity, and value alignment playing important roles. Sponsors prioritize athlete performance, personality, and social media presence, while athletes emphasize financial support and performance optimization. Sponsorship activation is generally limited, and agreements are predominantly in-kind or hybrid. The study concludes that sponsorships in low-media-coverage sports are relational and selective, relying heavily on athlete-driven outreach and social media visibility. These findings extend existing sponsorship frameworks to an underexplored context and offer practical insights for sponsors and athletes in niche sports. Full article

Background: Trail running (TR) is an endurance discipline characterized by prolonged exercise, irregular terrain, and marked elevation changes, which increase eccentric muscular load and may induce muscular, neuromuscular, and cardiac damage. Objective: This study aimed to systematically review the evidence on muscular, neuromuscular, and cardiac damage associated with TR participation. Methods: This systematic review followed PRISMA 2020 guidelines and was registered in PROSPERO (CRD420251135043). Five databases (PubMed, Web of Science, Scopus, SportDiscus, and ScienceDirect) were searched up to 31 August 2025. Observational, longitudinal, prospective, and case studies involving healthy adolescent or adult trail runners were included. Outcomes comprised muscle damage biomarkers (e.g., creatine kinase, alanine aminotransferase), neuromuscular function (e.g., squat jump performance, maximal voluntary isometric contraction), and cardiac biomarkers (e.g., CK-MB, cardiac troponins, NT-proBNP). Methodological quality was assessed using the National Heart, Lung, and Blood Institute Study Quality Assessment Tool. Results were synthesized qualitatively. Results: Fifteen studies met the inclusion criteria, including a total of 247 participants. Post-race analyses consistently showed marked increases in muscle damage biomarkers and significant reductions in neuromuscular performance. Transient elevations in cardiac biomarkers were also observed, suggesting acute but reversible cardiac stress following TR events. Limitations: Evidence was limited by methodological heterogeneity, small sample sizes, and underrepresentation of female athletes. Conclusions: It was found that trail running induces substantial acute muscular, neuromuscular, and cardiac stress, particularly in events with high eccentric loading. Monitoring biochemical and neuromuscular markers may support training load optimization, recovery strategies, and injury prevention. Full article

This study aimed to assess the effect of a downhill-running (DR) bout on muscle damage biomarkers. It also examined whether training background and gait kinematics may influence DR-induced muscle damage and strength loss. Thirty-six experienced trail runners (25 men, 11 women), participants of a 106 km ultra-trail, performed a 5 km DR bout at 15% decline and at an intensity equivalent to their first ventilatory threshold. Muscle damage biomarkers (creatine kinase, lactate dehydrogenase, and myoglobin) were analyzed before and 30 min after the DR protocol, and also before and after the UT race. Isometric strength was assessed before and after DR, and gait parameters were recorded during DR. All muscle damage biomarkers increased following DR (d = 0.19 to 1.85). Lactate dehydrogenase concentrations after the race and DR were associated (r = 0.64). Athletes who habitually performed downhill repetitions showed reduced creatine kinase (182 ± 73 U/L vs. 290 ± 192 U/L; p < 0.05; d = 0.64) and greater squat strength retention (4 ± 10% vs. −9.1 ± 16.8%; p <0.05; d = 0.87). Ankle plantar flexion and squat strength retention were inversely correlated with vertical oscillation (r = −0.44) and step length (r = −0.37), respectively. In summary, lactate dehydrogenase response to a short DR bout could indicate an athlete’s readiness to handle ultra-trail-induced muscle damage, although further research is needed to confirm it. In addition, despite the exploratory nature of the study, regularly performing downhill intervals and adopting a more terrestrial gait pattern appear to soften strength loss and muscle damage response to DR. Full article

Altitude training camps are a popular measure to enhance endurance performance at sea level. This study elucidates the effects of acute altitude-induced hypoxia, running speed and surface on cadence, peak tibial acceleration (PTA), gait asymmetry and residual shock in distance running. Ten healthy, trained native lowlanders (6 males, 4 females; 28.2 ± 9.2 years; mean ${\dot{V}}_{{\mathrm{O}}_{2,\mathrm{p}\mathrm{e}\mathrm{a}\mathrm{k}}}$ of 54.9 ± 5.9 mL min⁻¹ kg⁻¹) participated in this study. They ran 1500 m bouts of at 50, 1000 and 2300 m above mean sea level on paved roads and natural trails at three different speeds. Those speeds were chosen to represent the most common training zones and were defined as $v_1=90\%\cdot v_{\mathrm{VT}1}$, $v_2=\frac{1}{2}\left(v_{\mathrm{VT}1}+v_{\mathrm{VT}2}\right)$ and $v_3=100\%\cdot v_{\mathrm{VT}2}$, with $v_{\mathrm{VT}1}$ and $v_{\mathrm{VT}2}$ denoting the speeds at the ventilatory thresholds 1 and 2. Based on the experimental results, cadence increased by +2.2 spm per +1 km h⁻¹ (p < 0.001) and fell by −1.1. spm per +1000 m of elevation (p < 0.001), whereas surface did not show any significant effect. Likewise, PTA was not affected by surface, but grew by 0.9 g per +1 km h⁻¹ (p < 0.001), and decreased by −0.6 g per +1000 m in elevation, with significant effects particularly at speeds beyond v_VT1 (p < 0.049). Absolute lateral asymmetry was not altered by elevation, surface or running speed. Mean shock attenuation increased with running speed by +2.5 percentage points per +1 km h⁻¹ (p < 0.001) but was independent of elevation and surface. In essence, running speed seems to be the predominant factor defining biomechanical loading, even under acute hypoxia and for varying surface conditions. Full article

Multi-material additive manufacturing (MMAM) enables integration of multiple materials within single products, but existing design methodologies lack systematic frameworks linking detailed consolidation decisions to product-level functional requirements while preserving functional independence. This paper presents a methodology that extends the conventional design process model with a reverse-traced workflow connecting part-level decisions to higher-level product architecture. By tracing how Design for MMAM (DfMMAM) affects design decisions in reverse, designers can identify the best opportunities to use MMAM based on their project scope. The methodology introduces a Level of Process Integration (LPI) framework based on design novelty that structures redesign scope according to whether changes affect part geometry, component assembly, or function allocation, enabling designers to balance consolidation benefits against validation complexity at each level. Sequential decision-making workflows systematically determine which functions can be co-located within unified components while maintaining functional independence through zone-specific design parameters. The methodology is illustrated through a qualitative case study on trail running shoe design across three integration levels, identifying substantial consolidation potential while establishing the foundation for future quantitative validation. Unlike existing approaches limited to part-level redesign, this framework traces detailed consolidation decisions back to product architecture trade-offs, clarifying redesign scope and validation rigor required at each integration level. By operationalizing the relationship between functional decomposition, physical architecture, and MMAM capabilities, this framework provides designers with structured decision pathways to balance consolidation benefits against redesign complexity at each design phase. Full article

A special class of complex adaptive systems—biological and social—thrive not by passively accumulating patterns, but by engineering coherence, i.e., the deliberate alignment of prior knowledge, real-time updates, and teleonomic purposes. By contrast, today’s AI stacks—Large Language Models (LLMs) wrapped in agentic toolchains—remain rooted in a Turing-paradigm architecture: statistical world models (opaque weights) bolted onto brittle, imperative workflows. They excel at pattern completion, but they externalize governance, memory, and purpose, thereby accumulating coherence debt—a structural fragility manifested as hallucinations, shallow and siloed memory, ad hoc guardrails, and costly human oversight. The shortcoming of current AI relative to human-like intelligence is therefore less about raw performance or scaling, and more about an architectural limitation: knowledge is treated as an after-the-fact annotation on computation, rather than as an organizing substrate that shapes computation. This paper introduces Mindful Machines, a computational paradigm that operationalizes coherence as an architectural property rather than an emergent afterthought. A Mindful Machine is specified by a Digital Genome (encoding purposes, constraints, and knowledge structures) and orchestrated by an Autopoietic and Meta-Cognitive Operating System (AMOS) that runs a continuous Discover–Reflect–Apply–Share (D-R-A-S) loop. Instead of a static model embedded in a one-shot ML pipeline or deep learning neural network, the architecture separates (1) a structural knowledge layer (Digital Genome and knowledge graphs), (2) an autopoietic control plane (health checks, rollback, and self-repair), and (3) meta-cognitive governance (critique-then-commit gates, audit trails, and policy enforcement). We validate this approach on the classic Credit Default Prediction problem by comparing a traditional, static Logistic Regression pipeline (monolithic training, fixed features, external scripting for deployment) with a distributed Mindful Machine implementation whose components can reconfigure logic, update rules, and migrate workloads at runtime. The Mindful Machine not only matches the predictive task, but also achieves autopoiesis (self-healing services and live schema evolution), explainability (causal, event-driven audit trails), and dynamic adaptation (real-time logic and threshold switching driven by knowledge constraints), thereby reducing the coherence debt that characterizes contemporary ML- and LLM-centric AI architectures. The case study demonstrates “a hybrid, runtime-switchable combination of machine learning and rule-based simulation, orchestrated by AMOS under knowledge and policy constraints”. Full article

Trail running is a demanding endurance sport where performance prediction models often rely on laboratory testing or pre-race data, limiting their practical application. This study presents a real-time predictive model for marathon and ultra-trail races, based on variables recorded during the race, including uphill/downhill pace-times, terrain difficulty coefficients, and partial rankings. A total of 947 runners from the ‘Trail Valle de Tena’ event (Spain) were analyzed to develop equations that estimate total race time using only the first third of the race. The model incorporates weighted time (WT_n), pacing variability (WTV_n,n+2), and checkpoint percentile rank (CPR_n), showing strong predictive power (adjusted R² > 0.95) across sexes and race modalities. These variables reflect the runner’s ability to both overcome elevation and maintain consistent pacing, offering insights into fatigue management and performance optimization. The model enables coaches and athletes to monitor race progression, adjust strategies in real time, and potentially reduce injury risk through better control of effort intensity. Unlike laboratory-based models, this approach is fully applicable in field conditions and does not require prior testing. Further validation in similar endurance events is recommended to confirm its utility as a practical tool for training and competition planning. Full article

Background/Objectives: This exploratory study aimed to evaluate the impact of an 80 km ultra-marathon trail running event on changes in faecal bacterial composition, and to investigate whether any correlations exist between exercise-associated gastrointestinal symptoms (Ex-GIS) with faecal bacterial profiles. Such events represent a unique physiological stressor and may impact the composition of the gut microbiota. Studying this impact may provide insights into acute (i.e., <24 h) gut microbiota changes under extreme conditions. Methods: Thirteen endurance athletes (n = 7 males, n = 6 females) aged 41 ± 8 years completed the 80 km Margaret River (Australia) ultra-marathon race in 2022. Faecal samples were collected pre- and post-race. Faecal bacterial profile, as per relative abundance (RA) of operational taxonomic units and the determination of α-diversity (Shannon Equitability Index (SEI)), was achieved by 16S rRNA amplicon gene sequencing. Changes in RA% and SEI pre- to post-race were assessed by the Wilcoxon signed-rank test. Correlations between Ex-GIS with bacterial profile and changes pre-, during, and post-ultra-marathon race were determined by Spearman’s rank correlation coefficients. Results: Bacterial calculations of phyla (n = 5), family (n = 23), and genus (n = 41) were detected for RA (≥0.5%). A significant decrease pre- to post-race of Actinobacteriota (p = 0.035) phyla, Bifidobacteriaceae (p = 0.007), and Clostridiaceae (p = 0.010) family, and Blautia (p = 0.039) and Subdoligranulum (p = 0.023) genus was determined; meanwhile, Oscillospiraceae (p = 0.016) and Monoglobaceae (p = 0.039) family significantly increased pre- to post-race. No other bacterial group changes were observed. No correlations were observed between pre- to post-ultra-marathon RA change and Ex-GIS. Conclusions: The completion of an 80 km ultra-marathon did not invoke substantial changes in the gut microbiota as determined by faecal bacterial profiling. Very strong and strong correlations were observed between certain bacterial groups and Ex-GIS; however, no significant correlations were observed between pre- to post-ultra-marathon changes in RA ≥ 0.5% and Ex-GIS. Full article