Search Results (1,447)

Walnuts are commercially important tree nuts whose moisture content (MC) influences quality, shelf life, and post-harvest processing. This study evaluated the potential of low-cost acoustic sensing combined with machine learning for non-destructive MC prediction. Sixty in-shell walnuts were subjected to controlled drying at 40 °C for 26 h, with acoustic recordings and physical measurements collected every two hours. Acoustic signals were processed using Wavelet Soft Threshold Denoising (WSTD), Short-Time Fourier Transform (STFT), and Variational Mode Decomposition (VMD), and features were extracted from the resulting signals. Predictive models included generalized linear models (GLM), random forests (RF), gradient boosting machines (GBM), and Partial Least Squares (PLS) approaches. Following grouped walnut-level validation, the highest MC prediction performance was achieved by the model combining dimensional and acoustic descriptors (RF: R² = 0.836; GBM: R² = 0.826), while the model combining drying time and acoustic descriptors achieved moderate predictive performance (RF: R² = 0.762; GBM: R² = 0.760). Overall, the results provide proof-of-concept evidence that acoustic descriptors may complement physical measurements for non-destructive walnut moisture-content prediction. However, substantially larger independent datasets collected across multiple cultivars, production batches, acquisition conditions, and external validation studies will be required before practical industrial implementation can be considered. Full article

Ambient Assisted Living (AAL) environments aim to enable older people to remain active and lead an autonomous and independent life for as long as possible. Among the technologies that can be incorporated into these settings, socially assistive robots (SAR) seek to establish a more natural and intuitive means of interaction with people, whilst helping them to carry out everyday tasks. One of the main challenges facing the design of these robots is how to enable them to undertake more complex tasks. Recent advances in Large Language Models (LLMs) have opened new avenues for flexible robot deliberation, yet their integration into real-time robotic systems remains challenging due to latency constraints, reasoning reliability, and the complexity of coordinating multi-step tasks. This paper proposes a hierarchical multi-agent architecture for robot deliberation that addresses these challenges by combining LLM-based planning with structured execution mechanisms within the ROS 2 ecosystem. The proposed architecture employs a supervisor agent that decomposes high-level natural language instructions into prioritised subtasks, enabling a priority-driven execution model that dynamically adapts to task relevance, temporal constraints, and environmental feedback. Subtasks are delegated to a set of Single-Purpose Agents (SPAs), orchestrated via LangGraph state machines and coordinated through a priority-aware scheduling mechanism. A key design principle is the use of Behaviour Trees (BTs) as high-level callable tools through the Model Context Protocol (MCP), encapsulating closed-loop control strategies while enabling preemptive and priority-consistent execution. This reduces the number of LLM inference steps required per task and improves robustness under dynamic conditions. A further contribution concerns the deployment of fine-tuned, lightweight LLMs—on the order of 0.6 billion parameters—specifically adapted for both the supervisor and the individual SPA roles through parameter-efficient low-rank adaptation (LoRA). These models are trained on role-specific tool-calling datasets to specialise in constrained reasoning patterns and task-specific decision-making, enabling efficient, low-latency inference directly on edge hardware. The combination of fine-tuning and hierarchical priority control enhances both the determinism and responsiveness of the system while mitigating error propagation across agent interactions. The paper presents the full software architecture, a formal characterisation of the system as a priority-aware hierarchical policy over a graph of agent workflows, and an experimental evaluation in an Ambient Assisted Living scenario assessing task success rate, inference efficiency, responsiveness under competing priorities, and overall user experience. Because SPA execution is decoupled from the supervisor’s own reasoning loop, the architecture is designed to keep accepting, processing, and queuing new user queries while previously dispatched SPAs are still executing their tasks. Full article

Urban green spaces play a fundamental role in maintaining ecological stability, supporting biodiversity, regulating urban microclimates, and improving human wellbeing. Because photosynthetic activity is a key indicator of vegetation functionality and environmental adaptation, this study evaluated the photosynthetic performance of vegetation across eight urban green spaces within the campus of the University of Life Sciences “King Michael I” in Timișoara, Romania. Vegetation structure, spatial organization, and estimated photosynthetic rates of selected ornamental species were comparatively analyzed under the temperate climatic conditions of the study area. Two-factor analysis of variance (ANOVA) revealed that the spatial characteristics of the investigated green spaces had a highly significant effect on photosynthetic variability (F(7, 42) = 6.783; p = 0.000021), explaining approximately 46% of the total variance, whereas species identity accounted for only 13.3% and showed no statistically significant influence (p = 0.052863). The highest photosynthetic performance was recorded in structurally diverse green spaces characterized by dense tree canopy and balanced woody–herbaceous vegetation, whereas the lowest values occurred in highly urbanized areas with limited vegetation cover. These findings demonstrate that the structural organization and vegetation composition of urban green spaces strongly influence photosynthetic performance and support the use of integrated ecological assessment as a decision-support tool for sustainable and climate-resilient urban green infrastructure planning. Full article

Pediatric food allergy is an increasing public health challenge associated with severe allergic reactions, psychosocial burden, and healthcare inequalities. This study aimed to characterise molecular sensitisation patterns among Bulgarian children with suspected food allergy using component-resolved diagnostics and to evaluate family quality of life in a healthcare setting without access to epinephrine auto-injectors. A total of 1163 pediatric patients underwent molecular allergy testing using the ALEX² multiplex platform, while 90 caregivers completed structured quality-of-life questionnaires. A sensitisation threshold of >0.10 kUA/L was applied to maximise sensitivity for molecular profiling and co-sensitisation analyses. Tree nut allergens showed the highest sensitisation prevalence, predominantly involving Ana o 3, Pis v 1, Jug r 1, Cor a 9, and Cor a 14. Younger children showed a predominance of stable storage protein sensitisation, whereas older children more frequently demonstrated PR-10-associated cross-reactive profiles. Correlation analyses identified strong co-sensitisation clustering among phylogenetically related allergen families, particularly cashew-pistachio, walnut-hazelnut, mammalian milk, and fish allergens. Questionnaire responses revealed substantial psychosocial burden related to fear of anaphylaxis, chronic hypervigilance, restricted daily activities, and inadequate emergency preparedness. These findings highlight the importance of molecular risk stratification and underscore major unmet needs in pediatric food allergy management in Bulgaria. Full article

In response to the significant seasonal fluctuations in heating and cooling loads in greenhouses for high-value fruit trees in northern China, as well as issues such as heat accumulation on the ground-source side and high carbon emissions from coal-fired heating, this paper proposes a coupled energy supply system comprising a PVT system, a mine pit seasonal thermal storage unit, a ground-source heat pump and a cooling tower. Taking a 30,000 m² cherry greenhouse and an existing 15,000 m³ mine pit thermal storage reservoir in Weifang, Shandong Province, as the research objects, annual design-stage simulations with a 0.125 h time step were conducted using SketchUp-TRNBuild and TRNSYS. Discrete sensitivity analyses and engineering constraints were used to determine the PVT area and cooling tower outlet temperature. Heating demand mainly occurred from November to February, whereas cooling demand was concentrated from June to September. The selected 2452 m² PVT system supplied direct heating for 34 days, covered 23.05% of the seasonal heating demand, and achieved a storage efficiency of 69.17%. Without a cooling tower, the first-year soil temperature increased by 1.1 °C. With a 26 °C cooling tower outlet temperature, the soil thermal imbalance ratio decreased to 2.4%, and the 15-year soil temperature rise was limited to 0.28 °C. The recommended system required 1.3239 million kWh of net purchased electricity annually, reduced operating costs by approximately CNY 802,800 (USD 118,243) and operational emissions by 2027.8 tCO₂-eq per year relative to the baseline, and had a static payback period of 6.4 years. The annual operational emission reduction was linearly extrapolated over a 20-year assessment period under fixed weather, load, equipment performance, and grid emission assumptions, resulting in a scenario-based carbon reduction threshold of 40,556 tCO₂-eq. Net life cycle carbon savings would be possible if the additional emissions from construction, equipment replacement, and end-of-life treatment remained below this threshold. Full article

Invasive aspergillosis and mucormycosis are life-threatening fungal infections characterized by high early mortality, substantial toxicity, and intensive healthcare resource utilization. While multiple antifungal therapies are recommended in international guidelines, their comparative clinical and economic value remains uncertain, particularly in resource-limited settings. We developed a hybrid decision tree and Markov model to evaluate antifungal strategies, including liposomal amphotericin B, isavuconazole, voriconazole, posaconazole, and caspofungin, from the perspective of the Colombian healthcare system over a 6-month horizon. Health states included clinical response, treatment failure, toxicity, and death. Outcomes included quality-adjusted life-years (QALYs), costs, incremental cost-effectiveness ratios (ICERs), and net monetary benefit (NMB). Deterministic and probabilistic sensitivity analyses were performed. In invasive aspergillosis, isavuconazole provided a modest incremental benefit over voriconazole (+0.022 QALYs), driven primarily by improved tolerability, at an incremental cost of COP 3.4 million, yielding an ICER of COP 154.5 million/QALY. In mucormycosis, liposomal amphotericin B yielded greater survival benefits (+0.035 QALYs) but at substantially higher costs (ICER: COP 494.3 million/QALY). Across willingness-to-pay thresholds, isavuconazole and voriconazole demonstrated more favorable economic profiles, owing to lower toxicity and reduced hospitalizations. These findings highlight that the clinical–economic value of antifungal therapy is primarily determined by early survival gains, toxicity burden, and front-loaded costs. Strategies minimizing toxicity and hospitalization appear more efficient in constrained health systems, whereas survival-maximizing approaches may be justified in severe disease. This integrated analysis supports context-specific antifungal decision-making in the management of invasive fungal infections. Full article

The elevational divergence of plant CSR (Competitor–Stress-Tolerator–Ruderal) strategies represents a core theme in global change ecology, yet adaptive patterns across subalpine transition zones remain insufficiently understood. This study focused on 375 plant individuals sampled along an 1143–4361 m elevational gradient on Gongga Mountain. We divided elevations into three zones (low, mid, high) using tertile partitioning, measured leaf functional traits including leaf area (LA), specific leaf area (SLA), and leaf dry matter content (LDMC), and quantified CSR strategy scores with the StrateFy approach. We further explored how hydrothermal factors drive community- and intraspecific-level strategy variation. Results showed that CSR strategies differed significantly along the elevational gradient. The relative abundance of individuals with the C strategy was 53.4% at low elevations, peaked at 72.4% at mid elevations, and decreased to 43.6% at high elevations. By contrast, the S strategy was most abundant at high elevations (32.5%), while the R strategy consistently accounted for less than 5% across all zones. Life form-specific strategies varied strongly: herbs maintained a dominant C strategy (66.7%) at high elevations; trees shifted to S strategy dominance (60%) at high elevations; and shrubs displayed a balanced C–S strategy. Intraspecific plasticity in CSR strategies was pronounced: with increasing elevation, the dominant strategy shifted from C (33.3%) to S (66.7%), with a greater magnitude of variation than at the community level. Temperature was positively correlated with the C strategy (p < 0.001, R² = 0.53) and negatively correlated with the R strategy (p < 0.001, R² = 0.40). Precipitation was negatively correlated with the C strategy (p = 0.001, R² = 0.54) and positively correlated with the R strategy (p = 0.005, R² = 0.42), whereas both factors exerted weak effects on the S strategy. Our findings highlight that shifts in hydrothermal conditions are the primary drivers shaping CSR strategy distributions, and coordinated leaf trait variation serves as a key adaptation mechanism. These results improve mechanistic understanding of mountain plant adaptation and provide a scientific basis for alpine vegetation conservation and management. Full article

Persistent global warming significantly impacts crop phenology and productivity, with perennial fruit trees facing heightened challenges due to their long life cycles and complex heat stress accumulation. Apple and pear, which hold substantial economic and nutritional value, are particularly vulnerable to high temperatures, manifesting as accelerated phenology, impaired floral organ development, disrupted pollination and fertilization, and insufficient fruit coloration—all of which severely compromise fruit quality and commercial value. Although recent advances have been made in elucidating heat stress signal transduction and regulatory networks in model plants such as Arabidopsis and rice, research on heat stress responses in apple and pear remains limited. This review systematically synthesizes the physiological responses, gene expression regulation, and protective cultivation strategies under high-temperature stress in apple and pear, aiming to provide a theoretical foundation for thermotolerance breeding and the establishment of heat stress regulatory networks, thereby supporting sustainable production in the context of global warming. Full article

Species delimitation and biodiversity assessment in Chironomidae are frequently hindered by morphological conservatism and widespread cryptic diversity. Neozavrelia (Diptera: Chironomidae), a genus inhabiting specialized freshwater microhabitats such as springs and hygropetric seepages, remains poorly understood with respect to its genetic diversity and the environmental processes underlying lineage differentiation. Clarifying lineage diversity is a prerequisite for evaluating the ecological factors associated with genetic differentiation. In this study, we compiled a global COI DNA barcode dataset comprising 137 sequences from newly generated specimens and publicly available records in the Barcode of Life Data Systems (BOLD). Genetic diversity was evaluated using neighbor-joining (NJ) analysis and Automatic Barcode Gap Discovery (ABGD), whereas environmental and geographic drivers of genetic differentiation were assessed using Mantel tests, distance-based redundancy analysis (db-RDA), variation partitioning, and hierarchical partitioning. A total of 26 molecular operational taxonomic units (MOTUs) were identified, with most lineages forming well-supported clades in the NJ tree. Genetic differentiation was significantly associated with both environmental and geographic distances, although environmental variables explained a substantially greater proportion of the observed variation. Hierarchical partitioning identified slope, annual cloud cover, mean precipitation amount of the coldest quarter, and frost frequency as the principal environmental predictors of lineage differentiation. Our findings reveal substantial COI genetic diversity within Neozavrelia and suggest that environmental heterogeneity, particularly variation in water availability and thermal conditions, is significantly associated with lineage differentiation. This study expands the global COI DNA barcode reference library for Neozavrelia and provides new insights into ecological and evolutionary factors associated with diversification in freshwater chironomids. Full article

Background: Machine learning (ML) is increasingly applied in medicine, underscoring the need for transparent and clinically relevant models. In gastrointestinal oncology, most ML studies rely on raw imaging data, which limits clinical adoption due to poor interpretability and the difficulty of collecting high-quality, large-scale video and image datasets in routine practice. Endoscopic ultrasound (EUS) plays a central role in the evaluation of pancreatic cancer; however, structured EUS features remain underused in predictive modeling. Objective: To assess the performance and interpretability of ML models for diagnosing pancreatic ductal adenocarcinoma (PDAC) using routinely collected EUS variables. Methods: We conducted a retrospective multicenter study using data from two Italian hospitals (n = 641) for model training and internal validation and from a third hospital (n = 120) for external validation, collected from 2015 to 2023. Decision trees, random forests, naïve Bayes and other classifiers were developed and evaluated. Model performance was assessed in terms of discriminative ability, calibration, and selective prediction. Results: All models demonstrated high discriminative performance (AUC ≥ 0.90). Decision trees provided the most favorable balance between interpretability and accuracy (balanced accuracy = 0.87; sensitivity = 0.89). Calibration and selective prediction analyses confirmed the robustness of the models. Conclusions: These findings demonstrate the feasibility of implementing interpretable yet high-performing ML models for PDAC diagnosis in real-life endoscopic settings. Full article

The problem of path planning is one of the most crucial and challenging issues in the fields of intelligent systems and autonomous robotics. A robot’s ability to move quickly and easily from a starting position to a goal position without hitting anything is directly related to how useful the robot is in real life. This paper compares five advanced path-planning algorithms: A* (A-Star), D* Lite (Dynamic A-Star Lite), RRT* (Rapidly exploring Random Tree Star), PRM* (Probabilistic Roadmap Star), and APF-D (Adaptive Potential Field with Dynamic Awareness). The paper addresses the difficult problem in dynamic environments where objects enter, exit, and move around continuously within the robot environment, as it moves through the environment, which is becoming more prevalent in the real world, such as in warehouses, hospitals, and urban and outdoor environments. Six performance measures, namely path length, computation time, memory, optimality ratio, success rate, and replanning latency, are used to test our five algorithms on a standard simulator in Matlab. Experiments are conducted in four different conditions, from very quiet to very dynamic, with a high number of obstacles. Results indicate that A* fails to perform well in dynamic environments and performs nearly optimally in static environments, while APF-D and D* Lite adapt to changes in the environment much better. An experimental study was carried out by 50 independent simulations in static and dynamic environments, where in each simulation, the hybrid solution was evaluated. The APF-D algorithm showed a success rate of 92.8% in highly dynamic environments, which is better than that of A* (58.8%), RRT* (76.8%), and PRM* (70.5%), whereas the success rate of D* Lite was found to be 89.3%. Additionally, APF-D decreased the average time taken for replanning by around 25% in comparison to other graph-based algorithms. Full article

The mason bee Osmia excavata Alfken (Hymenoptera: Megachilidae) is an essential pollinator for rosaceous fruit trees in northern China, but its populations are declining because of severe infestations by the mite Chaetodactylus hirashimai Kurosa (Acari: Sarcoptiformes). This mite feeds on pollen within mason bee nest tubes, thereby impairing larval development. However, the effects of different floral resources on the fitness of C. hirashimai remain unclear. Herein, we evaluated the development, survival, and reproduction of C. hirashimai reared on four diets (cherry pollen, cherry bee bread, apple pollen, or apple bee bread), by using an age-stage, two-sex life table. Chaetodactylus hirashimai successfully completed its life cycle on all diets. Cherry pollen supported shorter pre-adult development, longer adult longevity, and higher fecundity than apple pollen. Bee bread diets resulted in significantly lower pre-adult mortality (11.76–14.81%), and higher fecundity (270–278 eggs/female) and net reproductive rate (R₀ = 115–120 offspring/individual), than raw pollen (44.64–44.83% mortality; 117–179 eggs/female; R₀ = 40–50 offspring/individual). No significant differences in the intrinsic rate of increase (r) and finite rate of increase (λ) were observed among diets. Our findings demonstrated that cherry pollen is a highly favorable resource for C. hirashimai, and bee bread further enhances mite population growth. This study highlights the ecological conflict between cherry pollination and mite infestation, and these life-table results can provide preliminary laboratory support for subsequent research on integrated mite control in fruit orchards. Full article

Maximal lifespan is one of the key life history traits, subject to variation and selection, with important implications for evolutionary and ecological strategies. Despite the keen interest in this topic and numerous studies on vertebrate lifespan and its drivers and correlates, there is little research on aquatic gastropods, for which information on life history and ecology is scarce. We used a dataset of 133 species, including the literature data on lifespan (the response variable) and 55 (explanatory) variables covering morphological, life history, ecological and biogeographical characteristics. We applied parametric regression models (generalised linear models, generalised additive models), comparative phylogenetic regression (Phylogenetic Generalised Least Squares) and nonparametric machine learning regression analyses (random forests, regression trees) to examine relationships between longevity and the other variables. We found that both parametric and non-parametric models converged towards similar results. Our findings support life history theory and metabolic scaling predictions, indicating that iteroparous, large-sized species that occupy environmentally stable habitats exhibit extended longevity. Further research should include a wider range of potential drivers to elucidate the underlying mechanisms of longevity and test mechanistic ageing hypotheses. Full article

Leucoptera malifoliella (Lepidoptera: Lyonetiidae) is a quarantine pest of apple and other Rosaceae fruit trees whose range is expanding into new territories. Its minute size and morphological overlap with closely related Lyonetiidae make routine identification unreliable, especially for larvae and damaged specimens. We compared COI sequences from six common small Lepidoptera species found in orchards and designed the species-specific primer pair SXW-F/SXW-R. The resulting polymerase chain reaction (PCR) assay amplifies an ~500 base pairs (bp) fragment exclusively from L. malifoliella; no product was detected in any of five non-target species. The reaction tolerates annealing temperatures of 50–58 °C and consistently detects the target across all life stages (first- to third-instar larvae, pupae, adults) and all adult tissues tested (antennae, head-thorax, abdomen, wings, legs). Detection sensitivity reaches 0.03 ng/μL—approximately one-thousandth of the DNA content of a single adult. This is the first species-specific COI (SS-COI) method reported for L. malifoliella. It furnishes a rapid, specific, and sensitive diagnostic tool for quarantine inspection, field monitoring, and integrated pest management (IPM) programs. Full article

As a widely used encapsulant for power electronic devices, silicone gel is continuously exposed to high voltage and high temperature during service, which seriously impairs the reliability and service lifetime of power modules. This work investigates the electrothermal coupling failure mechanism of conventional silicone gel under high-voltage and high-temperature environments; the influences of different pulse electric field edge times and aging stages on various properties of the material are investigated, including electrical treeing characteristics, breakdown field strength, amplitude of charge-excited molecular vibration, leakage current, and cone penetration. It is revealed that the failure mechanism of silicone gel is attributed to the synergistic effect between dynamic charge damage induced by the pulse edge electric field and the degradation of the solid–liquid two-phase structure at high temperatures. This research provides theoretical support and experimental basis for material composition modification, structural optimization, and improving the encapsulation life of power electronic devices. Full article