143174 research outputs found
Sort by
Shimmer Series- Reflecting the past and looking into the future
No full textNew exploration of craft techniques responding to the theme of Scale, explored through both large and small works to examine how shifts in proportion affect form, surface and the viewer’s relationship to an object.
Represented by Gallery FIVE at Collect, the collective also present a collaborative project with Goldsmiths’ Fair, bringing together past and present artists through a shared installation of miniature works in precious metals.
Presented at 'Collect' International Art Fair, at Somerset house in London.
Collect, presented by Crafts Council, is the only annual art fair on the UK and international circuit dedicated to museum-quality contemporary craft and design
Optimising Harbour Construction Projects for Environmental Sustainability: A Hybrid Artificial Intelligence Approach
No full textopen access articleHarbour sedimentation represents a major challenge to the environmental sustainability and operational efficiency of coastal infrastructure, as frequent dredging activities increase maintenance costs, ecological disturbance, and carbon emissions. Conventional physical and numerical sediment transport models, while widely applied, are computationally intensive and often unsuitable for early-stage, sustainability-oriented design optimisation. To address these limitations, this study proposes a hybrid artificial intelligence-based optimisation framework integrating Artificial Neural Networks (ANNs), Genetic Algorithms (GAs), and Particle Swarm Optimisation (PSO) for sustainable breakwater and harbour layout design. Hydrodynamic simulations using the Coastal Modelling System (CMS) were conducted to generate a comprehensive dataset describing sediment transport behaviour under varying geometric and structural configurations. An ANN surrogate model was trained to capture nonlinear relationships between breakwater parameters and accumulated sedimentation volume, while GA-based global optimisation and PSO-based validation and local refinement were employed to identify optimal design solutions. Comparative assessment demonstrated consistent convergence of ANN–GA and ANN–PSO solutions within the same design region, with a maximum deviation of 8.46% between design variables and a sedimentation difference of 2.4%. The hybrid ANN–GA–PSO framework achieved the lowest predicted sedimentation volume, representing an improvement of approximately 2.3% relative to the ANN–GA baseline. The proposed framework supports Integrated Coastal Structures Management (ICSM) by enabling proactive, design-stage reduction in long-term sediment accumulation and dredging requirements, offering a scalable pathway toward sustainable and digital-twin-enabled harbour planning
OBIMAP (One-Bead Interchain Multipeptide Assembly Platform)
No full textA significant advancement in Merrifield’s classic solid-phase peptide synthesis (SPPS) that greatly expands the scope of accessible peptide structures is reported here. Building upon the one-bead, one-compound (OBOC) concept, this approach enables the simultaneous synthesis of multiple peptides on a single bead, followed by a novel solid-phase interchain assembly reaction to produce the final peptide product. This method, the one-bead interchain multipeptide assembly platform (OBIMAP), successfully generates diverse peptide architectures, including linear, cyclic, and bicyclic structures─ranging from minimal cyclic dipeptides to small proteins─many of which are inaccessible through conventional SPPS. OBIMAP demonstrates superior efficiency in both time and product purity compared to traditional methods. Crucially, it eliminates the need for solution-phase fragment condensation, a common but cumbersome step commonly used in synthesizing therapeutic peptides (30–60 amino acids). In addition to enhancing conventional SPPS methodologies, the OBIMAP enables access to novel classes of peptide architectures, including highly constrained peptides that were previously considered synthetically inaccessible
Atomically thin gold embedded in inkjet-printed PVA hydrogels: flexible catalysts for ambient phenol degradation
No full textInkjet-printed gold nanotape (AuNT) structures embedded in polyvinyl alcohol (PVA) hydrogels provide a reusable, high-surface-area platform for catalytic degradation of phenol and 4-nitrophenol (4-NP) under ambient conditions. AuNTs, featuring distinct three-dimensional "heads" and atomically thin quasi-one-dimensional "tails", enhanced catalytic activity in both reduction and oxidation reactions. Compared to spherical gold nanoparticles (AuNPs), AuNTs are nearly twice as catalytically efficient for 4-NP reduction on a per-mass basis, reflecting the influence of anisotropic morphology on surface-sensitive electron transfer. In contrast, phenol oxidation shows weaker morphology dependence, likely proceeding through hydroxyl radical-mediated pathways that are less sensitive to catalyst shape or facet structure. To enable rapid substrate diffusion and facilitate reuse, AuNTs were formulated into PVA inks and inkjet-printed into micrometre-thick hydrogel mesh architectures (8 to 15 µm thick). Although printed meshes show reduced activity relative to free AuNTs in solution, they achieve a nearly fourfold increase in mass-normalised rate constants for 4-NP reduction compared to drop-cast gels (0.24×10⁴ vs. 0.07×10⁴ min⁻¹ g⁻¹) and achieve 26% phenol, a common water pollutant, in 4 hours at room temperature, with consistent performance over multiple cycles. These findings demonstrate the potential of inkjet-printed nanozyme hydrogels as scalable, heterogeneous catalysts. Further improvements may be achieved by optimising catalyst–matrix interactions to reduce diffusion and accessibility barriers. This work addresses a significant challenge in nanozyme catalysis: translating high-performance nanomaterials into practical, reusable formats suitable for environmental remediation
Graph neural networks with hybrid local-global attention for effective prediction of mechanical response in structures
No full textGraph Neural Networks (GNNs) are emerging as a transformative approach for predicting mechanical response in structures by naturally encoding unstructured finite element meshes as graphs. While traditional finite element analysis (FEA) provides trusted solutions for mechanics problems, it encounters significant computational and scalability bottlenecks. Existing machine learning approaches using regular grids fail to capture the irregular nature of real-world meshes, whereas standard GNNs suffer from over-smoothing and limited long-range information
propagation that compromise accuracy. To overcome these limitations, we present a Graph Transformer methodology that implements an encoder-processor-decoder framework augmented with a frequency-controlled hybrid local-global attention mechanism, which systematically bridges local mesh connectivity with global information awareness across the computational domain. Our approach integrates
seamlessly with real-world FEA workflows through an automated FEM-to-GNN pipeline that converts FE simulations to graph representations and generates training datasets directly from commercial solvers, enabling rapid deployment across various engineering applications without manual preprocessing bottlenecks. We validate our approach on open holes Carbon Fibre Reinforced Polymer (CFRP) laminate plates under various loading conditions and geometric configurations, and extend validation to nonlinear woven composites exhibiting plasticity and progressive damage under shear-dominated loadings, employing systematic hyperparameter optimisation with Tree-structured Parzen Estimator (TPE) sampling and mesh convergence studies. The linear case demonstrates excellent predictive accuracy (R2 = 0.98, RMSE = 0.00028) with 70× speedup over equivalent-accuracy FEA and 58 − 64% reduction in training and validation loss compared to standard message-passing GNN architectures without attention mechanisms, while the nonlinear case achieves R2 = 0.97 (RMSE = 0.0013) with 660× speedup
Detection of dengue virus serotype 2 in local Aedes aegypti populations, Madeira Island, Portugal, 2025
No full textPerspectives of young people on the use of neonatal data in the UK National Neonatal Research Database: a patient and public involvement project
No full textBackground
The secondary uses of routinely recorded data have become an important component of health research. The UK National Neonatal Research Database (NNRD) established in 2007 contains routinely recorded data from approximately 1.6 million babies who received care in NHS neonatal units in England, Wales, and Scotland. The earliest included infants are now adults. We aimed to explore their perspectives on the existence and use of the NNRD
Methods
As part of a wider Public and Patient Involvement programme, we established a Young Persons’ Advisory Group: “From Neonate to NOW”. Fifteen young adults with lived experience of neonatal care were recruited to share their views and inform neonatal research priorities. We held online focus groups with 10 members to explore their views on the NNRD. Analysis was thematic and co-produced with two group members.
Results
Young people expressed a strong desire to know how their data have been used. Their wish to understand more about their start in life revealed that the NNRD carries meaning beyond data. Participants described pride in knowing their data contribute to improving neonatal care yet were surprised that they had never heard of the NNRD. They proposed ways to raise awareness of its existence and value. The main themes identified were knowledge and pride, early and clear communication, meaning and identity, and consent and clarity.
Conclusions
As the use of large datasets of routinely recorded health information becomes increasingly common, questions about awareness, consent, and trust will become important, especially as the first waves of children represented in them reach adulthood. This study is among the first to explore how young adults with lived experience of neonatal care view the secondary
uses of their data. Their perspectives highlight areas to address as we move toward a future shaped by big data in healthcare research
Fine-tuning a small vision language model using synthetic data for explaining bacterial skin disease images
No full textVision–language models (VLMs) show strong potential for medical image understanding, but their large scale often limits practical deployment. This study investigates
whether a compact VLM can be effectively adapted for dermatology, with a focus on explaining and diagnosing bacterial skin diseases. We curate a dataset derived from PMC-OA using the BIOMEDICA dataset. We then construct PMC-derma-VQA-bacteria by pairing images with inherited figure captions and synthetically generated question–answer (QA) supervision produced by Google’s Gemini model. We fine-tune SmolVLM under three supervision settings: QA-only, caption-only, and a combined QA+caption strategy. The models trained under these settings are evaluated for both text generation quality and
diagnostic classification performance on a held-out test set. QA supervision yields the best report-generation performance, while the combined QA+caption setting achieves the highest classification accuracy (70.20%). These results suggest that synthetic QA supervision
can meaningfully enhance small VLMs for medical applications
In-situ wind turbine blade inspection using ultrasonic non-destructive testing
No full textOffshore and onshore wind turbine blades present significant inspection, maintenance and repair challenges arising from location, economic drivers, environment and the specific blade architecture concerned. In-situ tasks have traditionally been undertaken by people abseiling from the tower or use of gantries. Harsh conditions associated with windy environs, along with pressures to limit downtime, have led to a range of new technologies becoming available. This paper presents results from the use of ultrasonic nondestructive testing (NDT) measurements of subsurface blade topography arising from in situ and static blade inspection for a range of wind turbine types. The measurements have been enabled using a hexapod robot that can accommodate NDT scanners within its chassis and can, using pneumatic suction for the robot pedipulators, navigate the convex, concave, and flexing form of in situ wind turbine blades. The arising NDT tomographic scans provide detailed information on blade integrity, the presence or otherwise of bonding materials, and local feature condition. Measurements, presented over a 600 mm traverse span, have confirmed the reliability of the robotic platform to deliver high-quality, consistent, and reliable data to be acquired with limited NDT experience and to allow subsurface inspections to be performed and analyzed remotely. In addition to detailed measurement of subsurface blade features, the robot system has also demonstrated the capacity to undertake functions such as lightning protection system verification