20505 research outputs found
Sort by
A cross-platform study of human situational awareness for heterogeneous low altitude autonomy
No full textHuman–autonomy teaming in the Low Altitude Economy (LAE) requires operators to manage both ground and aerial autonomous agents under time pressure, spatial uncertainty, and cognitive load. This study investigates how visual and haptic feedback affect operator situational awareness (SA) in simulated collision avoidance tasks involving cars and drones. A high-fidelity virtual environment was built using Unreal Engine 4 and AirSim, with haptic cues delivered through a wearable bHaptics vest. Twenty-two participants performed within-subject trials across visual-only and visual–haptic conditions. Results showed that haptic feedback significantly enhanced SA, particularly in dimensions related to information acquisition and spare mental capacity. Improvements were more consistent in car-based tasks, while drone scenarios exhibited greater inter-individual variability. These findings demonstrate the potential of multimodal interfaces to support cognitive performance and reduce platform-related disparities in operator SA. This work provides empirical evidence for designing adaptive, perception-aware interfaces in safety-critical human–autonomy teaming systems.The work is supported by EPSRC CHEDDAR: Communications Hub for Empowering Distributed clouD computing Applications and Research(EP/X040518/1) (EP/Y037421/1). This study was also partially funded by Royal Society Research Grant RGS/R1/231037.2025 IEEE International Conference on Systems, Man, and Cybernetics (SMC
AI-Driven high-throughput document classification and knowledge distillation using hybrid machine learning and large language models through agentic AI
No full textTang, YunEfficient document classification at enterprise scale is increasingly pertinent as organizations must manage millions of unstructured documents per hour while ensuring compliance, high accuracy, operational efficiency, and transparency. This thesis presents a hybrid AI system that combines traditional machine learning (ML) and agentic AI, leveraging large language models (LLMs), to process and classify over one million documents per hour with high precision. This system also enables knowledge distillation for continual learning and cost control. Another important goal of the system is to repurpose high-value talent away from repetitive classification tasks, allowing them to focus on more strategic and impactful work within the organization. The design blends metadata- and content-based ML with content-aware LLM inference, orchestrated by a smart, self-healing, and adaptive routing agent that assigns classification tasks based on complexity and confidence scores. A comprehensive evaluation against established metrics demonstrates the system’s superiority over monolithic approaches, with empirical results underscoring its robustness, adaptability, and auditable intelligence. Traditional machine learning excels in speed and cost, while recent advances in Large Language Models (LLMs) have made nuanced document understanding feasible. This research presents a hybrid document classification system that integrates machine learning and agentic AI reasoning with a knowledge distillation correction loop, achieving scalable, compliant, and adaptive enterprise document governance. Results demonstrate throughput exceeding 1,000,000 documents/hour, robust classification accuracy, cost-effective LLM orchestration, and continuous improvement driven by user feedback. Comparative benchmarking and multi-metric analysis validate the system’s efficacy, generalizability, and auditability.MSc in Applied Artificial Intelligenc
Defining level of service standards for subsidised air transport in Colombia
No full textAerocivil ColombiaAir transport is vital for regional development and social inclusion in Colombia, especially where road and river connectivity are limited or entirely absent. Despite the strategic role of subsidised air routes operated by SATENA, Colombia lacks a structured framework to assess whether these services adequately meet the needs of isolated communities. This thesis addresses that gap by developing a classification framework to evaluate service levels on subsidises air routes, using a combination of operational, geographic, and socioeconomic indicators. A mixed-methods approach was applied across five case study routes selected to reflect Colombia’s geographic and demographic diversity. Quantitative data were gathered form national aviation and statistical agencies, while qualitative insights were obtained through questionnaires directed at both stakeholders and SATENA users. Service quality variables were weighted using the Analytic Hierarchy Process (AHP) and combined into a flexible classification model. The results showed significant variation in service adequacy across the five routes. While some service aligned with regional needs, others fell short, particularly in areas with high poverty and limited ground transport. The findings demonstrate that air connectivity in Colombia remains essential but uneven, and that current resource allocation could be improved. This thesis provides a practical framework for assessing subsidised air services and offers clear implications for policy and planning. By incorporating both operational metrics and user perspectives, the model supports data-driven decisions about air transport provision in remote regions of Colombia.MSc in Air Transport Managemen
Hidden in plain sight – using a scoping review to reveal the neglected sachet economy crisis
Full text linkSingle-portion, multilayer sachets are a popular packaging format for food, home and self-care products in low-income countries. Sachet waste is non-recyclable, and visibly litters land, freshwater and ocean environments, particularly in South East Asian countries described as having a ‘sachet economy’. As annual global sachet sales exceed one trillion units, we ask whether low levels of research on sachets and the sachet economy could contribute to misinformation about environmental and socioeconomic damage and a lack of policy action to reduce sachet use. Based on the PRISMA-ScR method, our scoping review included novel comparative analyses of multiple exclusion screenings and development of a waste hierarchy for literature analysis to reveal how information available to stakeholders changes as criteria are tightened. We found little empirical research, with knowledge gaps often addressed by institutions and NGOs, particularly regarding alternatives to the sachet economy at the top of the waste hierarchy. Policymakers and consumers in the Philippines and Indonesia were the most common target audience and geographical context. A high proportion of grey literature and inconsistent terminology may introduce the potential for bias and reduced impact.
Our research highlights the urgency of further research into social and environmental impacts of the sachet lifecycle and alternatives such as refill. As sachet use rises and spreads globally, we assert that without robust research and practical guidance for decision makers, environmental and socioeconomic consequences of sachet use will continue to grow in South East Asia and spread quickly to other low-income economies, hidden in plain sight.The Lead Author is in the Water-WISER (Waste Infrastructure and Services Engineered for Resilience) Centre for Doctoral Training funded by UKRI through the Engineering and Physical Sciences Research Council. Grant number EP/S022066/1.Cleaner and Responsible Consumptio
Social life cycle assessment: a systematic review from the engineering perspective
Full text linkSocial Life Cycle Assessment (SLCA) has emerged as a crucial element in sustainability discussions, gaining traction in academic research and industry practices. In response to increasing consumer awareness, governmental regulations, and corporate responsibility, sustainable production models have been widely adopted, influencing engineering projects to address not only environmental but also social impacts. This systematic review aims to explore the application of SLCA within the engineering field, focusing on key trends, methodologies, and challenges engineers face when implementing SLCA. To achieve this purpose, 196 peer-reviewed studies published between 2010 and 2024 were analysed through bibliometric and content analysis. The results reveal an increasing trend in SLCA research, with key contributions from energy, mechanical engineering, and chemical engineering fields. Despite its growing relevance, several methodological challenges persist, including a lack of standardisation, data availability issues, and difficulties in integrating SLCA with environmental and economic assessments. The findings of this research can facilitate academic researchers, industrial managers, and policymakers in implementing SLCA practices in engineering applications for a comprehensive sustainability assessment.This work was funded through Innovate UK program grant agreement with grant number [103040]International Journal of Sustainable Engineerin
Body force modelling for axial and centrifugal compressor pre and post-stall aerodynamics
Full text linkTunstall, Richard - Industrial Supervisor
Mazur, Steven - Industrial Supervisor
Harvell, John - Industrial SupervisorThe operation of a jet engine is limited in a way to ensure that the manifestation,
propagation and growth of local flow compressor instabilities is prevented. This inevitably
leads to more conservative compressor designs with increased surge margins
and reduced operating range to avoid the occurrence of unstable compressor phenomena.
These instabilities are known as rotating stall and surge and their implications
on the structural integrity and operability of the engine can be catastrophic.
These phenomena and the inception mechanisms that trigger their occurrence are
not yet fully understood, while the knowledge gap is exacerbated when considering
centrifugal or axi-centrifugal configurations. The cost of experimental campaigns
for the investigation of the post-stall response of aero-engine compression systems is
excessive, while the compressor is usually restricted to low rotational speeds to prevent
severe structural damage. The experiments can only be performed at a more advanced,
mature phase of the compressor design process. Sophisticated transient simulations
using commercial CFD software offer an alternative approach but the excessive
associated computational cost requirements make their real-life usage challenging.
Through-flow codes combining reduced and higher-order modelling methods are a
computationally efficient alternative, however very few validated implementations are
reported in the literature and their capabilities are strictly limited to axial compressor
configurations.
A lower-order modelling approach is developed, whereby the compression system is
solved as an empty duct with body force-fields imparting turning and losses to the flow.
A new body force model, applicable to all types of blades is developed. The blade
curvature is fully defined in three dimensions accounting for axial, circumferential,
radial forces and blade leaning. The flow-field solution is obtained transiently by
solving the 2D axisymmetric Euler equations on a body force-relevant grid which
ensures that the grid lines are aligned with bladed domains. The governing equations
with blockage in the relative frame of reference are derived, thus replicating both
metal and aerodynamic blockage effects, along with a method for the precise definition
of the additional blockage terms in convoluted ducts. The Godunov scheme coupled
with 3 Riemann solvers is used to obtain the fluxes. New analytical, simplified
models are derived to estimate aerodynamic blockage and mixing losses at reverse
flow conditions in impeller and axial blade passages. Loss correlations are used to
estimate losses at forward flow conditions, while mixed and reverse flow are treated
with a separate model.
The validity and limitations of the different modelling approaches are investigated
extensively using test-case scenarios and CFD data. The validation of the overall
through-flow framework is carried out on axial compressor and centrifugal compressor
stages using CFD and if available, experimental data. Steady-state, forward and
reverse flow characteristics are in good agreement with CFD and experimental
data, while the flow-field is reproduced with reasonable accuracy. Transient, post-stall
simulations in centrifugal compressors are carried out and validated against
experimental data. The code constitutes the first successful, validated, through-flow
approach, capable of predicting the post-stall, steady-state and transient aerodynamic
performance in centrifugal compressors.PhD in Aerospac
Architecting ATA 28 with model-based systems engineering: center of gravity balance and system performance
Full text linkThe fuel management system for a fixed-wing aircraft has been developed and explored with the model-based systems engineering (MBSE) methodology for maintaining the center of gravity (CoG) and analyzing flight safety. The system incorporates high-level modeling abstractions that exploit a mix of behaviors and physical detail resembling real-world components. This approach enables analysis for a multitude of system requirements, verification, and failure scenarios at high simulation speed, which is necessary during system definition. Initially, the CoG is maintained by directly accessing the flight deck valves and pumps in both wings and controlling them through the bang-bang control law. In the refinement phase of the fuel system controller, the manual and individual controls of the valves and pumps are replaced with an autonomous fuel transfer scheme. The autonomous scheme achieves no more than a 20 kg difference in fuel between the wings during normal conditions. In the event of failures, the controller achieves no more than a 100 kg difference in fuel between the wings. The difference returns to 20 kg within a settling time of 5 sec and a maximum allowable overshoot safety margin of 10% of the 20 kg difference in normal conditions (±2 kg). The specification 20 kg/5 sec band varies with pump and valve parameters. Although this specification is sufficient for a system-level model, it can be refined with pump and valve parameters and nonlinear effects in the network. The system identification method is also trialed to control an individual engine by estimating a proportional integrator derivative (PID) controller of the engine plant. The safety tests are initiated in a user interface enabling error detection and injection. The fuel system model is used for analyzing refueling, defueling, and jettison scenarios with appropriate flow rates.
Besides the CoG maintenance, several aspects of configurations of the system’s functional and logical architecture, considering increasing component redundancy and activities for MBSE framework, have been conducted. The logical and temporal verification of system requirements is performed in simulation. To ensure traceability and coverage, the requirements and the associated verification artifacts are digitally linked to the implementing blocks. Test scenarios are implemented for investigating resultant and emergent behaviors at various levels of system hierarchy by isolating either the subsystem or the components that have been performed. To further check out the MBSE workflow, the fuel system controller code has been directly emitted from the controller model for DO-178C objectives. At the mission-level validation, a jettison scenario is developed for a mission and flight plan in the digital mission engineering and systems analysis environment of Systems Tool Kit (STK) Aviator. The aircraft fuel system configuration is set using the fuel system model.
The power of MBSE methodology supported by a modeling and simulation framework provides plenty of opportunities for through-life analysis in the early design lifecycle phase.SAE International Journal of Aerospac
Integrated cost-efficient itaconic acid production from waste potatoes via a holistic upstream-to-downstream approach
Full text linkItaconic acid (IA) holds enormous potential in the polymer industry as a green substitute for several valuable chemicals derived from fossil fuels. However, its full potential is limited by high production costs, resulting from the use of expensive feedstocks and downstream processing approaches. In this context, this study strategizes an inexpensive approach for the efficient production and recovery of IA using damaged/defective waste potatoes, a major food waste across the globe. Initially, the optimized acidic saccharification process (1.5 % HCl, 20 % w/v solid loading) yielded 170 g glucose/kg potato. The fermentation of activated charcoal-treated potato hydrolysate by Aspergillus terreus DSMZ 23081 resulted in 32 g/L IA with a yield and productivity of 0.21 gIA/gGlu (36.2 gIA/kgWPB) and 0.17 g/L/h, respectively. The downstream processing was performed by implementing a salting-out extraction approach. Among various solvent-salt combinations, maximum IA extraction efficiency was observed when 25 % (w/v) Na2SO4 was used as a salting-out agent with sec-butanol as the extractant. Optimizing the physicochemical parameters (pH, solvent volume, shaking time, etc.) led to 92 % recovery from pure IA solution. Finally, when the process was applied to potato-based fermentation broth, 80.5 % IA was recovered with 85 % purity. By integrating valorisation of a low-cost agro-industrial waste with an optimized salting-out extraction strategy that overcomes key limitations of conventional IA purification, the developed bioprocess demonstrates a simple and economically feasible end-to-end strategy for sustainable IA production.Separation and Purification Technolog
An analysis and projection of diabetes prevalence in East England region
No full textBackground:
Diabetes prevalence continues to rise in England, placing increasing pressure on primary and specialist healthcare services. This study examined how demographic, socioeconomic, and healthcare access factors influence Diabetes Mellitus register size across six Integrated Care Systems (ICSs) in the East of England and generated scenario based projections of future diabetes burden and specialist workforce requirements.
Methods:
A longitudinal panel design was applied using annual data (2012–2021) for six ICSs. Descriptive trend analysis summarised changes in diabetes registers, GP practice numbers, GP list size, deprivation (IMD scores), and population density. The association between these factors and Diabetes Mellitus register size was quantified using a fixed effects panel regression model, selected through F-tests, Breusch–Pagan LM tests, and Hausman specification testing. Future diabetes registers (2023–2027) were estimated using a regression based deterministic projection framework integrating: (i) model based forecasting, (ii) four scenario models based on plausible changes in population growth, deprivation, and GP capacity, and (iii) linear trend extrapolation of endocrinology consultant workforce numbers.
Results:
Diabetes registers increased across all ICSs, with the region experiencing a 13% rise between 2012 and 2021. Regression findings showed that higher deprivation strongly predicted larger diabetes registers (β=2824.24, p<0.001), while increases in GP list size and GP practice numbers were also significant predictors. Under Scenario I, projected diabetes registers for 2023 ranged from 60,603 (Cambridgeshire and Peterborough) to 85,574 (Hertfordshire and West Essex). Scenario II, which incorporated greater increases in deprivation, produced larger projected registers across all ICSs, including 75,463 in Bedfordshire and 88,399 in Hertfordshire. Patient to consultant ratios were projected to increase in Bedfordshire and Mid and South Essex, suggesting potential specialist workforce shortages.
Conclusion:
Demographic growth, rising deprivation, and pressures in primary care are key drivers of the increasing diabetes burden in the East of England. Projection results indicate that several ICSs may face widening gaps between patient demand and specialist capacity. Strengthening consultant staffing, directing resources toward more deprived areas, and supporting primary care resilience will be essential to maintain equitable diabetes care in future years.Funding was received from Health Education England.BMC Public Healt
Plant-ash glass in Natronland: a review of the use of plant-based fluxes in glassmaking from the 1st to the 9th century CE in Europe and the Mediterranean basin
No full textThis article examines available evidence for the use of plant ashes in glassmaking in the Mediterranean basin and Europe from the 1st to the 8th-9th century CE. During this period, glassmaking in these regions was overwhelmingly dominated by mineral fluxes, primarily natron, but the chemical composition of some glasses found in secure archaeological contexts suggests the use of plant ashes in these glasses, which are, on the other hand, compositionally distinct from glass made in regions where plant ash was used as the standard flux, for instance east of the Euphrates. The chemical characteristics of these glasses are examined, and different criteria are presented to aid in their identification. The chemical features are also interrogated to suggest possible silica and alkali sources and to determine possible similarities and differences in raw materials supply vis-à-vis natron glass. The archaeological and chemical evidence is brought together to try to define the technological, economic, and institutional implications of a possible glassmaking industry using plant ash, whose production parameters differ from those mobilised by the natron glass industry, in the periphery of the Roman world.This article was conceived during the international conference A flash in the pan or early steps of a glass? Late Antique (6th-7th century) plant ash-glass in the Mediterranean and continental Europe, held in Córdoba on 2–4 April 2024, with the financial support of project PID2019-108736 GB-I00. PROVERUM. Pruévalo e verás ques cierto. Recetas conocimientos de la sociedad medieval para el siglo XXI, funded by the Spanish Agencia Estatal de Investigación (AEI), Ministerio de Ciencia, Innovación y Universidades. The conference was held at the Archaeological Museum of Córdoba, which was kindly ceded by the Consejería de Turismo, Cultura y Deporte. Junta de Andalucía. The first author's work in Late Antique glass (including the analysis of the samples from Malaga) was undertaken within the framework of project The production, circulation, and consumption of glass in the Iberian Peninsula during the Visigothic period (fifth to seventh centuries), which was awarded a Rakow Grant (2018) by the Corning Museum of Glass. Thilo Rehren and Marcel Franken gratefully acknowledge support from the A. G. Leventis Foundation towards their research, including the participation in this conference. The earlier joint work by Daniela Rosenow and Thilo Rehren on Egyptian glass has benefitted from financial support by the Thyssen Foundation and UCL Qatar, which is gratefully acknowledged. Caroline Jackson gratefully acknowledges support from NERC (NERC OSS/340/0207) for funding trace element analysis (Beniot Disch/Kingston), and EPSRC (EP/F019750/1) for EPMA analysis (Eddy Faber/Nottingham).Journal of Archaeological Scienc