114333 research outputs found
Sort by
3D-Printed Smart Reinforced Beam for Strain Monitoring
No full textThe automation of concrete constructions through 3D printing has garnered considerable attention in civil engineering due to significant advantages over conventional methods. Nevertheless, the widespread adoption of this technology faces substantial challenges stemming from inherent uncertainties associated with the additive manufacturing process. A solution is to functionalize the 3D printed components with self-sensing capabilities to monitor performance during construction and operation and thus assess quality in real-time. Here, we study the local functionalization of 3D printed components through a hybrid 3D printing process. To do so, we build on prior work in self-sensing cementitious composites by integrating graphite powder and carbon microfbers as conductive fllers into cement-based mixtures to generate substantial piezoresistive capabilities. The technology is demonstrated on a 3D printed reinforced concrete beam. The smart beam is fabricated using a self-sensing composite at the bottom, followed by a continuous transition to a traditional cementitious mix. The printed self-sensing layers serve as strain-responsive interfaces capable of mapping strain field evolution by continuously monitoring changes in electrical resistance. A series of quasi-static and dynamic tests were performed to characterize the strain-sensing performance of the developed composite specimens. Results demonstrate the successful integration of self-sensing cementitious materials into the 3DP fabrication process, highlighting their potential for real-time monitoring of construction quality, detection of load-path alterations, and early identification of structural defects.This article is published as Liu, H., Sousa, I.N.L., Laflamme, S., Doyle, S.E, D'Alessandro, A., Ubertini, F., 3D-Printed Smart Reinforced Beam for Strain Monitoring. Procedia Structural Integrity. 2026,78;1759-1766.https://doi.org/10.1016/j.prostr.2025.12.22
Green manure enhances ecological pest management by triggering systemic resistance in rice through reshaped rhizosphere microbiome
No full textEcological pest management (EPM) is gaining increased attention with concerns regarding human health and the environment. Planting green manure (GM) represents a significant practice in EPM; meanwhile, GM enhances crop production and reduces environmental footprints via its effect on the soil microbiome. GM's direct inhibitory effect on pests and its protective effect on natural enemies have been widely reported. However, the impact of GM's soil legacy effect on pests and the underlying molecular mechanisms remains poorly characterized. In this study, three-year field trials, greenhouse experiments, and multi-omics integration were conducted to address the gap. Compared to winter fallow treatment, GM significantly reduced the occurrence of rice major pests by 43.8–94.2 %, including Mythimna separata, Cnaphalocrocis medinalis, Chilo suppressalis, and rice planthoppers. The infestation rate of C. suppressalis, consumption by M. separata, and oviposition by Nilaparvata lugens were reduced by 64.3–87.4 %, 38.7–39.9 % and 45.3 %, respectively. Mechanistically, GM upregulated key defense-related genes and stimulated biosynthesis of flavonoids and alkaloids, alongside the accumulation of jasmonic acid and salicylic acid, indicating synergistic activation of induced systemic resistance in rice plants. Rhizosphere soil analysis revealed GM-driven enrichment of plant-beneficial taxa (Rhizophagus irregularis, Bradyrhizobium erythrophlei, Pseudolabrys sp.), alongside enhanced soil multifunctionality (N/C cycling) and nutrient mobilization. Our PLS-PM results supported a scenario in which GM-induced pest suppression is potentially mediated by microbiome-driven defense priming. Our findings provide fundamental insights into EPM and highlight how GM modulates the rhizosphere ecosystem and further enhances aboveground systemic resistance in rice. This study offers a potential solution for reducing synthetic inputs in crop production, which contributes to agroecosystem sustainability.This article is published as Sun, Jiaqi, Yangyang Hou, Yueqiu Liu, Lei Zhang, Dianjie Xie, Lin Ma, Jixing Xia et al. "Green manure enhances ecological pest management by triggering systemic resistance in rice through reshaped rhizosphere microbiome." Resources, Environment and Sustainability 23 (2026): 100285. https://doi.org/10.1016/j.resenv.2025.100285This work was funded by projects under the National Natural Science Foundation of China (32572941 and 32072420), the China Agriculture Research System of MOF and MARA (CARS-22), and the Government Procurement of Public Services of MARA (072507034)
Genotypic and phenotypic exploration of Cardinium hertigii-induced reproductive manipulation of Encarsia parasitoid wasps
No full textCardinium hertigii is an obligate intracellular bacterium which lives as a symbiont of many arthropods, including insects, nematodes, and arachnids, and is primarily transmitted vertically from an infected female host to her offspring. Cardinium has varying effects on its hosts, ranging from negative to conditionally beneficial, but it is often manipulative in nature. To ensure and enhance its own spread through host populations, Cardinium has evolved different methods of hijacking host reproduction. These include two reproductive manipulation strategies which bias the sex ratio of host offspring toward females (feminization and parthenogenesis induction, or PI), and one which results in symbiont-induced mortality of offspring between an infected male host and an uninfected female, called cytoplasmic incompatibility (CI). While inducing such reproductive manipulation phenotypes is not unique to Cardinium, studies have shown that Cardinium has likely evolved the factors used to cause PI and CI independently from other manipulative symbionts like Wolbachia pipientis. The factors used by Wolbachia to cause CI and PI have either been confirmed (CI) or proposed (PI), but currently little is known about potential Cardinium CI factors outside of genomic and transcriptomic evidence. Further, nothing is known of potential PI factors due to a lack of available genomes for Cardinium which cause PI. This dissertation aims to address this large knowledge gap through a variety of -omics and heterologous expression approaches.
In this dissertation, we first assemble and explore the transcriptome of the parasitoid wasp host of Cardinium strain cEper1 host, which will allow us to utilize metaproteomics to identify potential Cardinium CI factors. These candidates will then be characterized by heterologous expression in Saccharomyces cerevisiae. Two main cEper1 proteins were found to be toxic upon expression, providing evidence for their toxicity and potential to be involved in causing CI. We also utilize comparative genomics of other Cardinium strains infecting parasitoid wasps in the genus Encarsia to study potential shared and unique host interaction genes of the CI strain cEper1 compared to another CI Cardinium (cEina3), two strains associated with PI (cEper2, cEhis1), and one asymptomatic strain (cEina2). The genes identified in this study are great candidates for follow-up characterization. Overall, this dissertation provides the best evidence to date providing support for potential factors involved in Cardinium-induced reproductive manipulation
Toward reliable AI: Practical and certified defenses for adversarially robust vision systems
No full textDeep learning has achieved remarkable success across numerous domains, yet its vulnerability to adversarial perturbations, as first revealed in 2013, has significantly limited its applicability in critical and safety-sensitive applications. Beyond the practical concerns, this phenomenon is intrinsically intriguing, warranting dedicated study in its own right. These adversarial attacks, though ingeniously crafted, conclusively demonstrate that modern deep networks often exploit superficial statistical correlations rather than developing a true semantic understanding of the tasks they perform. This has drawn substantial attention from both theoretical and applied perspectives.
This dissertation, develops a unified framework for improving the reliability of vision models against adversarial threats, spanning both practical black-box defenses and formal certification methods. The central goal is to ensure that vision systems can maintain correct predictions under a wide range of perturbation models, deployment settings, and task complexities.
First, we present AlignFix, a practical black-box defense for image classification inspired by top-down feedback in biological perception. AlignFix exploits complementary feature biases in naturally trained and adversarially trained models to detect and correct adversarial perturbations on the fly, providing robustness against score-based, decision-based, and transfer attacks i.e, all practical black-box attacks.
Second, we introduce Open Vocabulary Certification (OVC), a fast certification framework for open-vocabulary vision-language models such as CLIP. OVC leverages pre-computed certificates for related prompts, along with embedding caching and distributional approximations, to accelerate randomized smoothing by up to two orders of magnitude while retaining tight robustness guarantees for novel prompts.
Finally, we propose one of the first direct certification frameworks for object detection, applied to YOLO-based runway detection in aviation settings. Using Interval Bound Propagation, our method jointly certifies classification, objectness, and localization heads, enabling provable robustness guarantees for both labels and bounding boxes under bounded perturbations.
Together, these contributions advance the development of vision systems that are both practically robust in realistic black-box settings and certifiably correct under formal threat models, enabling a path toward reliable, trustworthy AI in high-stakes visual perception tasks
Exploring how Black adolescent girls experience colorism in school
No full textThis qualitative study centered the lived experiences of eight self-identified Black and African American girls in an Urban-Midwestern middle school. Through lenses of self-awareness, sisterhood, and sociocultural consciousness, their voices were captured to illuminate how they navigated and resisted the layered impacts of colorism. Engagement in meaning-making permitted youth to challenge dominant narratives and redefine their identities. This meaning-making was a way to disrupt the status quo leading to collective empowerment, a sense of well-being, and agency. Utilizing Sista Circle Methodology—a culturally responsive, dialogic framework— while situated in Black Feminist Thought, these Black adolescent girls discovered historical strength, resilience, and a vision for future possibilities through storytelling and communal conversations. The data collected was interpreted using Narrative Analysis, and the findings revealed four central themes that impacted the youths’ self and group perceptions, peer relationships, and educational experiences: (1) racism, (2) colorism, (3) microaggressions, and (4) the tools and strategies used to confront colorist occurrences. As participants cultivated a critical consciousness of their marginalized positioning, they reclaimed cultural and social distinctiveness while affirming Black girlhood. Implications for future research and practice included the importance of examining how historical context and shared knowledge inform perception; the role of culturally responsive methodologies in affirming Black girlhood in academic and community spaces; the reimagining of curriculum; and the cultivation of navigational strategies that foster critical consciousness. Ultimately, this calls for proactive interventions that affirm Black girls’ identities while nurturing belonging, resistance, and self-defined success
Integrating structural and semantic understanding for robust knowledge graph construction: From knowledge graph completion to zero-shot entity linking
No full textKnowledge Graphs (KGs) have become an essential foundation for representing structured knowledge and supporting reasoning in artificial intelligence.
However, real-world KGs are inevitably incomplete and semantically inconsistent, limiting their ability to provide reliable knowledge for downstream applications.
This thesis addresses these challenges by advancing both Knowledge Graph Completion (KGC) and Entity Linking (EL)—two fundamental yet interdependent tasks that underpin the construction of accurate and self-evolving KGs.
The first part of this research focuses on the Few-shot Knowledge Graph Completion (FKGC) problem, where models must predict missing facts for relations with only a handful of reference examples.
Existing methods often suffer from suboptimal negative sampling and static entity representations.
To overcome these limitations, we propose RANA (Relation-Aware Network with Attention-Based Loss), which strategically selects relevant negative samples and introduces an attention-based loss to emphasize more informative contrasts.
A dynamic relation-aware entity encoder is further designed to generate context-dependent entity representations.
Extensive experiments demonstrate that RANA significantly outperforms state-of-the-art FKGC models on multiple benchmark datasets.
Building on this foundation, the second part of the thesis investigates how to integrate structural and semantic knowledge for general Knowledge Graph Completion.
Most prior methods rely solely on either KG embeddings or pre-trained language models (PLMs), resulting in incomplete representations.
To bridge this gap, we develop Bridge, a unified framework that jointly encodes entities and relations through PLMs and structural representation learning.
Bridge introduces a self-supervised fine-tuning strategy inspired by BYOL, constructing semantically consistent “views” of triples without altering their meaning.
This alignment enables effective fusion of textual semantics and graph structure, yielding state-of-the-art results across multiple KGC benchmarks.
The final part of the thesis extends from KGC to Entity Linking, focusing on zero-shot EL in the biomedical domain—a setting characterized by lexical divergence and annotation scarcity.
We propose a cost-aware hybrid framework that leverages large language models (LLMs) to synthesize semantically faithful, entity-centric variants, which are then used to fine-tune compact retriever–reranker models.
This design reduces reliance on expert annotation while maintaining computational efficiency.
Experiments show that the proposed framework achieves robust zero-shot generalization and improved performance under lexical variation, surpassing existing biomedical EL systems.
Together, these three studies advance the goal of constructing self-improving knowledge graphs, where completion enriches relational knowledge and linking ensures semantic accuracy.
By tackling the challenges of few-shot learning, cross-modal representation, and zero-shot generalization, this thesis contributes to the development of scalable, data-efficient, and semantically grounded methods for intelligent knowledge graph construction and evolution
Seroprevalence of porcine coronavirus antibodies in Iberian pigs and wild boars from central-western Spain
No full textPorcine coronaviruses (PoCoVs) are common etiological viral agents of enteric and respiratory disease in swine, but most epidemiological information derives from intensively managed herds. Data from outdoor systems and wildlife remain scarce, despite the potential role of free-range production and wildlife-livestock interactions in sustaining virus transmission. In southern Spain, the traditional Dehesa agroforestry system supports Iberian pigs that share space and resources with wild boars and other species, creating interfaces where cross-species circulation may occur. To address this gap, we assessed the seroprevalence of five PoCoVs, including porcine hemagglutinating encephalomyelitis virus (PHEV), porcine epidemic diarrhea virus (PEDV), porcine deltacoronavirus (PDCoV), transmissible gastroenteritis virus (TGEV), and porcine respiratory coronavirus (PRCV), in Iberian pigs and wild boars from central-western Spain. A total of 260 Iberian pig sera and 564 wild boar sera collected between 2016 and 2020 were tested using indirect ELISAs for PHEV, PEDV, and PDCoV and a blocking ELISA for PRCV/TGEV. Antibodies to PHEV were highly prevalent in Iberian pigs (68.0%) and detected at lower levels in wild boars (22.6%), a pattern consistent with endemic exposure in domestic pigs and sporadic circulation in wildlife. PEDV antibodies were identified in 8.5% of Iberian pigs and 2.8% of wild boars, with higher prevalence in pigs during 2016 followed by a sharp decline, suggesting past but not ongoing activity. PDCoV antibodies were rare overall (3.1% in pigs, 2.6% in wild boars) but reached 41.6% in pigs from Cáceres in 2017-2018, indicative of a localized event. PRCV antibodies were widespread in Iberian pigs (67.3%), with higher prevalence in Badajoz compared to Cáceres, while wild boars showed rising seropositivity in Ciudad Real/Toledo by 2022 (up to 33.3%). No TGEV antibodies were detected in either host population, supporting the predominance of PRCV in the region. These findings demonstrate that PHEV and PRCV are enzootic in free-range Iberian pigs, while PEDV and PDCoV circulate at low levels, and wild boars are more likely incidental than reservoir hosts. The detection of antibodies in the absence of clinical outbreaks underscores the silent nature of PoCoV circulation in extensive systems and highlights the importance of integrating wildlife-livestock interfaces into surveillance and biosecurity strategies in Mediterranean production landscapes.This article is published as Encinas, P., Real, G.d., Magtoto, R. et al. Seroprevalence of porcine coronavirus antibodies in Iberian pigs and wild boars from central-western Spain. Porc Health Manag 12, 3 (2026). doi: https://doi.org/10.1186/s40813-026-00488-3
Biosecurity Practices on Small- and Medium-Scale Dairy Farms in Northern Kosovo: A Risk-Based Scoring Assessment
No full textBiosecurity plays a central role in preventing disease transmission in dairy production systems and animal welfare. However, quantitative data on biosecurity implementation in smallholder and medium-scale dairy farms remains inconsistent, especially in developing countries. This study provides a structured assessment of on-farm biosecurity practices in northern Kosovo using a standardized, risk-based scoring approach. A cross-sectional survey was conducted on 55 dairy farms using the unmodified Biocheck.UGent™ dairy questionnaire. External and internal biosecurity scores were calculated through predefined, weighted algorithms and analyzed using non-parametric descriptive statistics. Farm-level results were subsequently compared with international reference values derived from the Biocheck.UGent™ global database. The median biosecurity scores for Kosovo farms were 47.8% for external biosecurity and 29.0% for internal biosecurity, indicating uneven implementation with pronounced weaknesses in measures designed to limit within-herd transmission. The lowest-scoring domains were purchase and reproduction and feed and water within external biosecurity, and working organization and equipment, calf management, and calving management within internal biosecurity. In contrast, visitors and farmworkers, control of vermin and other animals among external measures, and adult cattle management among internal measures, showed relatively higher scores, although all remained below international reference levels. When compared with the global overall biosecurity reference median of 76.7% derived from the Biocheck.UGent™ database, the biosecurity performance of the surveyed dairy farms in Kosovo was substantially lower. This gap does not indicate a complete absence of biosecurity measures but rather an uneven application, with the most pronounced deficiency observed in routine practices that govern within-herd disease transmission. The use of a risk-based scoring system allowed these weaknesses to be identified in a structured manner and placed the Kosovo results within an international benchmarking framework. In this context, the approach functions as a practical diagnostic tool, enabling farmers and veterinarians to prioritize feasible, epidemiological-relevant improvements within small- and medium-scale dairy production settings.This article is published as Mehmedi, B.; Voca, D.; Youngs, C.R.; Saegerman, C.; Sinani, A.; Behluli, B.; Heta, S.; Cana, A. Biosecurity Practices on Small- and Medium-Scale Dairy Farms in Northern Kosovo: A Risk-Based Scoring Assessment. Agriculture 2026, 16, 442. https://doi.org/10.3390/agriculture1604044
Physiology and disorders of the transition period in modern sows
No full textThe transition period is a critical timeframe for sows, as substantial physiological and metabolic adaptations occur during the shift from late gestation to the start of lactation. Furthermore, the period surrounding farrowing represents the greatest risk for sow removals, with a substantial proportion of sow mortality and culling occurring during this short interval. Postpartum dysgalactia syndrome (PDS) represents one of the most important conditions affecting sows during early lactation, and PDS is characterized by insufficient milk production resulting in negative consequences on piglet growth and survival during the pre-weaning period. Therefore, PDS represents a critical threat to sow longevity, as premature culling of sows often occurs as a consequence of reduced sow productivity. Therefore, one major focus of this dissertation was to elucidate the physiology underlying development of PDS by evaluating circulating metabolites and inflammatory markers, in addition to characterizing alterations to microbial populations within the fecal and reproductive tract microbiota. Collectively, the results from these studies demonstrate alterations to circulating inflammatory markers and metabolites, suggesting inflammation and metabolic disturbances may contribute to the underlying pathophysiology of PDS. Furthermore, pelvic organ prolapse (POP) also represents a critical threat to sow longevity, as POP is a major reason for sow mortality during late gestation and early lactation. Therefore, we tested the efficacy of bacitracin methylene disalicylate (BMD) administration as a potential strategy to reduce the incidence of POP in sows. Although this strategy was not successful at reducing POP incidence, an effect of BMD on stillborn piglets was observed. In summary, results generated from the studies detailed throughout this dissertation provide novel information to improve our understanding of challenges facing the transition sow, specifically in regard to PDS and POP
Direct observation of vortex liquid droplets in the iron pnictide superconductor CaKAs4Fe4 at 0.5Tc
No full textType-II superconductors under magnetic fields are in a quantum coherent non-dissipative state as long as vortices remain pinned. Dissipation appears when vortices depin, eventually driven by thermal fluctuations. This can be associated to a melting transition between a vortex solid and a vortex liquid. This transition is almost always observed very close to Tc when probed by macroscopic experiments. However, it remains unclear how the vortex solid responds to thermal fluctuations at the scale of individual vortices far from the melting transition. Here we use scanning tunneling microscopy (STM) to visualize vortices in CaKAs4Fe4 (Tc ≈ 35 K). We find vortex liquid droplets—localized regions in space where vortices strongly fluctuate due to thermal exctiation—at temperatures as low as 0.5 Tc. Our results show that the onset of dissipation at the local scale occurs at temperatures considerably below Tc in type-II superconductors.This is a preprint from Marqués, Oscar Bou, Jose A. Moreno, Pablo García Talavera, Mingyu Xu, Juan Schmidt, Sergey L. Bud'ko, Paul C. Canfield, Isabel Guillamón, Edwin Herrera, and Hermann Suderow. "Direct observation of vortex liquid droplets in the iron pnictide superconductor CaKAs4Fe4 at 0.5Tc." arXiv preprint arXiv:2601.18247 (2026). doi: https://doi.org/10.48550/arXiv.2601.18247