143174 research outputs found
Sort by
From data to actionable knowledge: AI-AR integration framework for industrial knowledge management
No full textIndustrial knowledge management (KM) remains highly fragmented, with knowledge capture, structuring, and application often treated as isolated activities. Conventional approaches depend on static documentation and experience-based practices, which limit their responsiveness in dynamic, operational environments. In this paper, we design a comprehensive framework for the integration of artificial intelligence (AI) and augmented reality (AR) in industrial maintenance and diagnostics. The framework leverages AI techniques, including natural language processing (NLP) for extracting and structuring domain knowledge and machine learning (ML) models for predictive fault classification. These AI capabilities are seamlessly combined with AR technologies to deliver immersive, context-aware, in-situ task guidance, thereby enhancing decision-making, reducing downtime, and supporting efficient, knowledge-based maintenance processes. A case study is employed to demonstrate the framework’s feasibility by developing a prototype system deployed in a maintenance setting. The AI module clusters maintenance actions and predicts task categories, while the AR component renders step-by-step maintenance instructions anchored in the physical workspace. This integration enables the transition from unstructured, digital maintenance records to step-by-step in-situ repair guidance, reducing reliance on expert memory or static manuals. The results show that combining AI-driven text analytics with AR-based visualisation creates a cohesive knowledge workflow that improves operational efficiency. The proposed framework offers a scalable approach to embedding KM into frontline industrial routines, laying the foundation for more adaptive, technician-centred knowledge systems
Advancing global climate-resilient, digitized sanitation in cities
No full textAround 55% of the world’s population lives in urban areas, a proportion that is expected to increase to 68% by 2050.
New challenges for sanitation and wastewater and fecal sludge management place growing pressure on policymakers and industry professionals, and investment and innovation in this sector has become critical.
The need for climate adaptation interventions, alongside digital technology use for the improvement of service management, are two key trends that have emerged as solutions to address these growing challenges in global sanitation and wastewater management
Transfer learning of data-driven crystallisation processes via constrained neural ordinary differential equations
No full textModelling complex crystallisation processes remains challenging due to limited ex perimental datasets, high measurement noise, and the need for generalisability across varying operating conditions. Neural Ordinary Differential Equations (NODEs) and transfer learning (TL) offer promising tools to overcome these limitations by provid ing data-efficient, flexible, and transferable modelling frameworks. This work investigates the use of NODEs to model protein crystallisation dynamics under data-scarce conditions. A NODE trained on a data-rich source system successfully captures solute consumption and particle size dynamics, but when applied to data-sparse target sys tems, scratch-trained NODEs exhibit limited generalisation and unphysical behaviours. To address this, several TL strategies are evaluated, including layer freezing, parame ter deviation penalisation, and system-embedding within the neural architecture. Results show that layer freezing and deviation penalty consistently improve knowledge transfer, while system-embedding offers robustness in noisy or undersampled datasets. In addition, physics-informed NODEs, constrained to enforce monotonic concentra tion decay and crystal growth, demonstrate greater stability under high noise and
sparse measurement regimes, ensuring physically consistent predictions. Overall, the combination of constrained NODEs with appropriate TL strategies provides a robust framework for accurate, transferable modelling of crystallisation systems in low-data
regimes
Self-powered real-time monitoring of electrochemical pollutantdegradation using a spiral-tube triboelectric nanogenerator
No full textReal-time monitoring of electrochemical pollutant degradation is essential for intelligent water treatment but remains challenging in power-limited or remote scenarios. Here, we develop a fully self-powered sensing platform based on a vertical contact–separation triboelectric nanogenerator (CP-TENG) incorporating a spiral PTFE microtube and copper electrodes. The
device harvests fluid-induced mechanical energy while simultaneously transducing variations in liquid conductivity, flow rate, and pollutant composition into distinct electrical signatures. The output signals exhibit strong quantitative correlations with the degradation kinetics of ammonia nitrogen and Reactive Blue 19, enabling continuous tracking of both inorganic and organic pollutants without external power or optical instrumentation. The CP-TENG further demonstrates high sensitivity in distinguishing conductive and non-conductive liquids and maintains excellent operational stability over prolonged cycling. This work establishes TENG-based architectures as dual functional platforms for sustainable energy harvesting and intelligent
environmental sensing, offering a promising route toward autonomous wastewater treatment and next-generation smart monitoring infrastructures
Advancing large language models for comprehensive scientific assistance
No full textDespite their transformative potential, large language models remain ill-suited for scientific inquiry. They hallucinate citations, miss quantitative evidence in tables, construct arguments without attribution, and amplify biases. These failures undermine trust where researchers need it most. This thesis reimagines large language models as accountable companions that augment scientific judgment through grounding, argumentation, and responsible generation.
We develop capability-specific enhancements targeting three critical bottlenecks. First, we demonstrate that existing models fundamentally lack scientific table understanding, a capability essential for evidence-based reasoning. Through curriculum training that progresses from general table reasoning to domain-specific scientific tables, we show that models which truly comprehend tabular evidence, rather than superficially processing table captions, develop representations that substantially improve peer review score prediction. This finding establishes that deep engagement with quantitative data is necessary for models to assess scientific claims as researchers do.
Second, we treat scientific argumentation as fundamentally distinct from generic reasoning. Through a pipeline combining reasoning-aware distillation, multi-task learning, and reinforcement learning with composite rewards, we teach models to locate claims, attribute evidence, and draft targeted critiques with measurable gains across argument mining, generation, and discourse evaluation. Third, we address the alignment tax, the trade-off between safety and performance, by introducing methods that preserve factual faithfulness while reducing toxicity and bias without degrading language quality.
These capabilities converge in a retrieval-augmented workflow designed for transparency. Outputs expose sources, reveal reasoning, and acknowledge limitations. Ablation studies confirm that modules contribute additively and interact constructively, delivering grounded, auditable, and bias-aware assistance. The collective findings reframe large language models from error-prone editorial aids to trustworthy instruments for evidence-based inquiry that scientists can genuinely rely on.Open Acces
Sex and age specific genetic risk across the dilated and arrhythmogenic cardiomyopathy spectrum: insights from the SHaRe registry
No full textDeep learning tools for structural and functional analysis of cardiac MRI
No full textCardiovascular diseases are the leading cause of death worldwide, underscoring the need for earlier detection and improved risk stratification. Cardiac magnetic resonance imaging (MRI) is a key modality for assessing myocardial structure and function. Regional functional analysis has the potential to enhance diagnostic and prognostic accuracy by revealing subtle early-stage abnormalities. However, cardiac MRI is constrained by trade-offs in acquisition time, spatial resolution, and coverage. Standard protocols typically rely on thick-sliced short-axis stacks and complementary long-axis views, which only partially compensate for limited through-plane detail.
This thesis investigates deep learning approaches to overcome these limitations and enable more comprehensive, patient-specific assessments of cardiac function. First, we propose a multi-view super-resolution framework that combines standard 2D and 3D cardiac MRI views to reconstruct high-resolution volumetric representations. While promising in design, this method did not demonstrate substantial improvements in practice.
We then extend the multi-view concept to segmentation of the right ventricle, a structurally complex and clinically significant chamber. By leveraging spatial relationships across views, the proposed model improves segmentation accuracy over single-view approaches.
Next, we present a regional motion analysis framework for the left atrium, incorporating segmentation, motion estimation, strain computation, and atlas-based analysis to detect functional impairments. To address data scarcity, we integrate test-time optimisation, data augmentation, and regularisation.
Finally, we develop spatial-aware physics-informed neural networks aimed at eventual integration as a motion regulariser in cardiac analysis. By embedding physical constraints into the learning process, we have shown in an example application that the model can yield more consistent and physiologically plausible estimations.
Together, these contributions aim at advancing the capabilities of cardiac MRI analysis by enabling a more detailed and personalised assessment of cardiac function. The methods developed in this thesis lay the groundwork for integrating regional functional analysis into clinical workflows, ultimately supporting earlier diagnosis.Open Acces
Diagnostic accuracy, fairness and clinical implementation of AI for breast cancer screening: results of multicenter retrospective and prospective technical feasibility studies
No full textArtificial intelligence (AI) promises to enhance breast cancer screening. Here we evaluated Google's mammography AI system (version 1.2) across two phases: a retrospective study using 115,973 mammograms from five National Health Service screening services with 39-month follow-up and prospective noninterventional feasibility deployment at 12 sites (9,266 cases). The primary endpoint was AI sensitivity and specificity versus first reader using a 5% noninferiority margin. The secondary endpoints were performance versus second or consensus readers and breast-level analyses. Retrospectively, AI achieved superior sensitivity (0.541 versus 0.437 for first reader, P < 0.001) and noninferior specificity (0.943 versus 0.952, P < 0.001). Cancer detection rate increased from 7.54 to 9.33 per 1,000 women, with AI detecting 25.0% of interval cancers. Performance was particularly strong for first screens (39.3% fewer recalls, 8.8% higher detection) and invasive cancers. No systematic demographic disparities were observed. Simulated second-reader replacement reduced reading time by 32% while increasing detection by 17.7%. Prospective deployment confirmed technical feasibility but revealed a distribution shift requiring threshold recalibration. Implementation requires adaptive calibration and continuous monitoring to ensure safety and equity
Innovation, mergers, and the limits of the competitiveness narrative
No full textThe Draghi report on European competitiveness has reignited a familiar debate. Its diagnosis – that Europe suffers from fragmented markets, underinvestment, and an inability to scale – has been widely accepted. One of the policy conclusions drawn from it, not always by Draghi himself but certainly by those invoking his authority, is that merger control should be relaxed to allow European firms to grow larger and compete globally. The implicit promise is one of efficiencies and innovation: bigger firms, so the argument goes, invest more, innovate more, and ultimately
serve consumers better. This is a seductive narrative. It is also an empirically contested one. And it is precisely at this juncture, when political momentum risks outpacing analytical rigour, that clarity from economics is most needed
Silicon-based hexacarboxylic acids and their application in metal-organic frameworks
No full textMetal-Organic Frameworks (MOFs) represent a rapidly expanding class of crystalline materials characterised by their tuneable porosity and modular design. These materials offer exceptional versatility in both structure and function, making them attractive candidates for a wide range of applications. The development of new MOFs depends greatly on the design of novel organic linkers, as they play a crucial role in determining both the structure and the properties of the resulting frameworks. Chapter 1 introduces MOF chemistry, covering its historical background, key design and synthesis strategies, activation and characterisation methods, and the use of organosilicon multicarboxylic acids (nCOOH ≥ 3) in MOF synthesis. Chapter 2 outlines the synthetic approaches toward organosilicon multicarboxylic acids and details the preparation of organosilicon hexacarboxylic acids featuring different cores and functional groups. Ten hexacarboxylic acid compounds were synthesised and characterised using various techniques. Chapter 3 introduces hydrogen-bonded organic frameworks (HOFs), presenting notable examples based on multicarboxylic acids and discussing the X-ray crystal structures of hexacarboxylic acid compounds prepared in this study that form HOFs. Chapter 4 explores the use of L1-H6 and L4-H6 in 3D MOF synthesis. Eleven novel 3D MOFs were synthesised by treating L1-H6 with a variety of metals, including trivalent, divalent and monovalent metals. These reactions yielded MOFs with a range of topologies, including rare nia and fsy nets. In addition, the reaction of L4-H6 with In(NO3)3·5H2O afforded L4-In with nia topology, isoreticular to L1-In. The use of water stable L1-Fe and L1-Fe-Cl MOFs as adsorbents for phosphate and perfluorooctanoic acid (PFOA) removal from aqueous solutions is described in Chapter 5. L1-Fe-Cl showed higher adsorption capacities for phosphate (27.18 mg/g) than L1-Fe, and an outstanding capacity for PFOA (795 mg/g), and excellent recyclability. These findings confirm the potential of L1-Fe-based MOFs as efficient adsorbents for the removal of phosphate and PFOA from contaminated water sources.Open Acces