20505 research outputs found
Sort by
Modelling the production impacts of scaling up agroforestry systems in England and Wales
No full textCONTEXT: Land managers and governments face the challenge of using finite land, labour, and financial resources to achieve multiple objectives. Agroforestry, the integration of trees with farming, is promoted as a strategy for achieving multiple policy objectives relating to productivity, climate change and biodiversity. However, regional and national-scale studies validating its effectiveness remain limited.
OBJECTIVE: Our study aimed to model the impacts of scaling up agroforestry on food and fibre production in England and Wales. We developed nine agroforestry scenarios combining three tree types (apple, poplar, and short rotation coppice (SRC) willow) at three planting densities. Each scenario scaled agroforestry to all suitable agricultural land (representing 79% of total agricultural land) as a simple modelling objective, rather than a realistic target.
METHODS: We used the well-established Yield-SAFE model to simulate tree-crop/grass interactions at low, medium, and high tree densities, and inputted the resulting yields into the Optimal Land Use Model (OLUM), a linear programming model with the objective of maximising food energy production under defined constraints. The OLUM was validated using baseline data.
RESULTS AND CONCLUSIONS: Scaling up agroforestry increased domestic supply of tree products, for which the UK is heavily import-dependent. However, this came at the expense of the calorific value of food production, which decreased by 3% to 45%, depending on tree type and density. The largest reductions were observed in arable and vegetable crops, due to reduced area and yields. Ruminant livestock outputs were least affected, supported by increases in grassland area. Timber and apple agroforestry systems were predicted to be more land-efficient than equivalent monocultures (for apples, traditional orchards), based on land equivalent ratios.
SIGNIFICANCE: Upscaling agroforestry could reduce import reliance on tree products while contributing to environmental objectives. To minimise reductions in domestic food supply, policymakers should prioritise agroforestry in pasture-based systems and support wider food system changes. Design improvements could lessen trade-offs associated with tree planting in arable systems. Future research should investigate how scaling up agroforestry systems affects greenhouse gas emissions, biodiversity, soil, water, and the general resilience of the farmed landscape.This research was enabled with funding from UK Research and Innovation (UKRI) under the UK Government's Horizon Europe funding guarantee (grant number 10039008), and the European Union, via the Horizon Europe Research and Innovation Programmes under grant agreements No 101060635 (REFOREST) and 101059794 (DigitAF).Agricultural System
Detection of deoxynivalenol, its modified forms, and zearalenone in individual oat grains using visible-near-infrared spectroscopy and near-infrared hyperspectral imaging
No full textFusarium mycotoxins such as deoxynivalenol (DON), its modified forms, and zearalenone (ZEN) frequently contaminate oats, posing serious health and regulatory concerns. This study assessed the use of visible-near-infrared (Vis-NIR) spectroscopy and near-infrared hyperspectral imaging (NIR-HSI) to classify individual oat grains according to the European legal limits for DON (1750 μg/kg) and ZEN (100 μg/kg). NIR-HSI consistently outperformed Vis-NIR, achieving classification accuracies (CAs) above 91% and F1-scores above 0.65 for DON, ZEN and combined DON + ZEN detection. The most informative spectral regions were in the NIR ranges of 1000-1250 nm and 1300-1500 nm, associated with Fusarium-induced biochemical and structural changes in oat grains. Reducing the spectral input to 20 selected wavelengths preserved NIR-HSI performance, supporting the feasibility of multispectral implementations. These targeted, non-destructive approaches could enable early removal of the few highly contaminated grains responsible for batch rejection, improving food safety, reducing waste, and enhancing the sustainability of oat processing.This work was supported by the Spanish Ministry of Science and Innovation (predoctoral grant FPU21/00073 and Project PID2020-114836RB-I00 funded by MCIN/AEI/10.13039/501100011033) and Cranfield University.Food Contro
National climate change impact assessments underestimate the potential of autonomous adaptation
No full textCentral Europe is projected to lose up to 25% of its crop productivity by 2050 because of climate change, posing significant challenges to agricultural systems and food security. Effective adaptation strategies must consider not only domestic impacts but also global climate effects, including international trade dynamics. We performed a multilevel analysis of climate change impacts on agriculture, using the Czech Republic, a landlocked, crop production-based economy with an open market, as a case study. We integrated the global biosphere management model (GLOBIOM) with the gridded global crop model EPIC-IIASA. Climate impacts were projected with five global circulation models under three climate scenarios, with and without CO2 fertilization, and applied in national, EU-regional, and global productivity change scenarios. The results show that national-only assessments underestimate both risks and opportunities: production is projected to decline by up to 9% when global interactions are excluded but to increase by up to 8% when trade and market effects are included. Autonomous adaptation mechanisms, such as cropland reallocation, shifts in management intensity, and trade adjustments, buffer biophysical yield losses and improve economic outcomes. Neglecting global interactions in national climate change assessments increases the risk of maladaptation and policy inefficiencies. The incorporation of international market linkages enhances the ability to design robust adaptation strategies, enabling countries such as the Czech Republic to maximize resilience while minimizing environmental and socioeconomic trade-offs.This study was financially supported by the Technology Agency of the Czech Republic (grant "Prediction, Evaluation and Research for Understanding National sensitivity and impacts of drought and climate change for Czechia, PERUN"; SS02030040) and by the Ministry of Education, Youth and Sports of the Czech Republic (grant AdAgriF—Advanced methods of greenhouse gas emission reduction and sequestration in agriculture and forest landscapes for climate change mitigation; CZ.02.01.01/00/22_008/0004635).Regional Environmental Chang
Implementation of Direct Numerical Simulation Framework in the Finite Volume Methods-Based Open-Source Code Athena++ for Astrophysical Flows
No full textThis thesis investigates the performance of the open-source code Athena++ for the simulation of viscous incompressible flows across laminar and turbulent regimes, with particular emphasis on adaptive mesh refinement (AMR). Benchmark test cases, including steady-state channel flow, the lid-driven cavity flow, and the Taylor–Green vortex (TGV) in both two and three dimensions, were employed to assess the solver’s accuracy, stability, and computational efficiency. Spatial and temporal discretization schemes were systematically evaluated through grid convergence studies, supported by Richardson extrapolation and Grid Convergence Index (GCI) analysis, to verify the numerical implementation and establish the expected order of accuracy. An AMR strategy based on a local grid Reynolds number (Regrid) was implemented and tested in both laminar and turbulent TGV flows. For two-dimensional cases, computational savings of up to 44% were achieved, while in three dimensions with moderate Reynolds numbers, the savings were up to 35%. The AMR solutions successfully reproduced global flow characteristics such as kinetic energy decay and dissipation rate, but also highlighted the need for higher-order reconstruction and time integration schemes, as well as the limitations of a single refinement criterion. Finally, under-resolved direct numerical simulations (DNS) of the three-dimensional TGV at Re=1600 demonstrated that Athena++, equipped with high-order reconstruction (PPM) and time integration (RK4), can capture turbulence dynamics in close agreement with benchmark solutions when using grid resolutions consistent with Kolmogorov theory. However, AMR revealed an unphysical increase in the spectral energy density at dissipative scales, possibly due to numerically generated instabilities introduced as a result of mesh-size variations. The findings confirm that Athena++ provides accurate and robust solutions for viscous flows and highlight the potential of AMR to reduce computational cost in under-resolved DNS, provided that the mesh refinement and de-refinement criteria are guided by the physical scales of turbulence.MSc in Computational Fluid Dynamic
Robust deepfake detection through the AI-Guard mobile app for Real-Time image identification
No full textThe proliferation of deepfake technologies has created an urgent need for robust, real-time detection systems capable of verifying image authenticity, particularly in mobile environments. This paper presents a scalable deepfake image detection framework integrated into the AI-Guard mobile application. Our approach leverages fine-tuned, computationally efficient convolutional neural network (CNN) architectures, including VGG19, InceptionV3, Xception, EfficientNetB0, ResNet50, and MobileNetV3Large. Trained on a large-scale, balanced dataset of over 450,000 real and fake images sourced from six publicly available datasets, the models incorporate advanced preprocessing, adversarial data augmentation, and optimized training pipelines. Among these, VGG19 achieved the highest generalization performance with 98.9% validation accuracy and 93.2% accuracy on previously unseen real-world data. The system supports real-time inference via a REST API, enabling practical mobile deployment with low latency. To support transparency and reproducibility, the curated training dataset has been made publicly available through our institutional repository. Our results demonstrate that AI-Guard offers an effective, deployable solution for forensic image verification, contributing to countering misinformation and enhancing trust in digital media.Human-Intelligent Systems Integratio
Intelligent transformation and green development: a double debiased machine learning evaluation of china’s IIP initiatives
No full textIn promoting the transformation, upgrading, and green development of traditional industries the Chinese government has been introducing Integration of Informatization and Industrialization Pilot (IIP) initiatives based on intelligent manufacturing and the green economy. As such this study examines the impact of China’s IIP policy on the green development of publicly listed manufacturing firms by applying both Difference-in-Differences (DID) and double debiased machine learning (DDML) models to a dataset that spans the period 2007–2022. The evidence suggests that the IIP policy initiative significantly improves firms’ green development via the mediating effect of intelligent transformation. Robustness checks, including DDML model regression and Propensity Score Matching-DID (PSM-DID) with nearest neighbour matching, consistently demonstrate significant improvements in environmental efficiency and productivity due to the IIP. Moreover, these effects are notably pronounced in the Yangtze River Economic Belt, heavily polluting industries, and larger firms. This research addresses a gap in micro-level policy analysis, highlighting the potential of intelligent manufacturing to promote sustainable practices. By offering both theoretical and practical insights, the findings guide policymakers and businesses in leveraging informatization and industrialization for green development.Annals of Operations Researc
Assessment of Multiple Reference Frame (MRF) and Sliding Mesh methods on a brake system and wheel model
No full textSorrell, Matthew - Industrial Supervisor - Red Bull TechnologyThis project investigates the application of the Multiple Reference Frame (MRF) and Sliding Mesh (SM) methods for simulating the aerothermal performance of a rotating wheel and brake disc assembly. The study aims to evaluate the accuracy and computational trade-offs between the steady-state MRF approach and the more computationally intensive, transient SM method, and to determine if a correction factor can be derived to enhance MRF results. The methodology began with the aerodynamic validation of an isolated wheel model against experimental Particle Image Velocimetry (PIV) data, establishing solver settings and mesh independence. Later, a comprehensive model incorporating a 205/55/R16 tyre, rim, and a vented brake disc with pads and caliper was developed. Four steady-state MRF simulations were conducted at varying rotational speeds, and a single transient SM simulation was performed for comparison. Key quantities of interest included the drag coefficient (Cd), heat transfer coefficient (HTC), and temperature distribution. Results indicated that both MRF and SM methods produced aerodynamically valid results for the isolated wheel, with the SM method capturing wake structures with marginally higher fidelity. For the full brake system, significant differences emerged. The SM method predicted a 14.7% higher HTC and a more uniform temperature distribution across the brake disc compared to the MRF result at the same speed, suggesting it more accurately captures the transient convective cooling. A strong linear relationship was observed between rotational speed and HTC in the MRF data. A correction factor was developed and implemented via a User-Defined Function (UDF) to scale the MRF-predicted HTC. While this adjustment successfully brought the maximum temperature prediction to within 2.2% of the SM value, it did not fully replicate the SM results for other thermal variables. The study concludes that while the MRF method offers a computationally efficient approximation, the SM method provides a more physically accurate representation of the convective heat transfer process. A simple single-variable correction factor for MRF is insufficient to fully emulate SM results, as the differences are multi-variate and inherent to the fundamental approaches of each method. Future work should focus on longer-duration SM simulations for robust validation and exploring multi-variable correction methodologies.MSc in Computational Fluid Dynamic
Public perception of green infrastructure’s role in urban air quality
No full textUrban green infrastructure (GI) is recognised for enhancing environmental quality and wellbeing, yet public perceptions of its air quality benefits remain underexplored. This study investigates how residents across two contrasting urban contexts (Bedford and Milton Keynes) view these benefits, and how socio-demographics, proximity, and use frequency shape perceptions. A semi-structured survey of 51 participants combined closed- and open-ended questions. Descriptive statistics, t-tests, chi-square, and thematic coding identified key patterns. Findings show strong agreement (90.2%) that green spaces improve air quality, supported by a reliable attitudinal scale (Cronbach’s α = 0.771). Mature, evergreen, and mixed vegetation were rated higher than single-species or newly planted areas, indicating preference for diverse, stable planting. Visit frequency strongly correlated with belief in benefits (p < 0.001), and proximity influenced visitation (p = 0.02). Socioeconomic differences appeared: lower- and middle-income groups relied more on public green spaces, while some higher-income respondents were more sceptical. Despite concern over pollution, awareness and use of air quality indices were limited, with reliance on sensory cues. The study highlights the importance of context-sensitive, equitable, and communicatively effective GI planning to maximise ecological and social benefits.MSc in Environmental Engineerin
GROOT: GPT-based human-RObOT interface
No full textLarge Language Models (LLMs) have revolutionized the realm of Natural Language Processing (NLP). Their proficiency in planning and reasoning, combined with code generation capabilities, presents a novel avenue for robotics applications. This work introduces GROOT, a novel speech-based language-agnostic middleware that uses instructions and code examples as grounding principle to leverage Generative Pre-Trained Transformer’s ability to produce code for new and unseen tasks. Unlike methods based on language-conditioned robot policies, GROOT capitalises on auto-regressive code generation inspired by Code-as-Policy (CaP) and ProgPrompt. The aim is to create a human-robot interface using GROOT that enables the embodiment of an LLM to take user instructions like ’move in a square’, ’move 20 cm in front’, ’go to position ’X’ on the grid’ and return policy code based on robot API. In this paper, GROOT was used in a number of experiments to assess its performance to few-shot learning against spatial reasoning, logical reasoning and compound simulated tasks. This work reflects the potential of prompting code-based examples and API-based instructions as a grounding method to integrate large-language models with robotic platforms, envisioning seamless and intuitive human-robot interactions.2025 7th International Conference on Control and Robotics (ICCR
Ammonia partitioning and recovery from industrial wastewater - exploring precipitation, stripping, and sorption technologies
No full textJefferson, Bruce - Associate SupervisorCircular economy in wastewater management is increasingly applied, with
ammonia recovery playing a critical role. Established ammonia partitioning
technologies, being precipitation, typically as struvite, stripping and scrubbing,
and sorption, have been predominantly applied to manure, anaerobic digestate,
urine and municipal wastewater. Industrial effluents also hold potential for
ammonia recovery and have been increasingly targeted by research. These
effluents comprise a wide category of wastewaters with diverse physicochemical
characteristics, generated by different sectors, including food/drink processing,
mining, agro-industrial processes, manufacturing, metallurgy, etc. Some of these
effluents contain high ammonia loads alongside significant concentrations of
ions, metals, and recalcitrant organic compounds, contributing to complex
chemical compositions that can pose challenges for conventional recovery
technologies. Despite the increasing focus on industrial wastewaters, there
remains limited understanding of how to effectively select and operate recovery
technologies, based on the effluent composition and desired recovery outcomes.
This research aimed to advance the understanding of how several physicochemical factors impact the mechanisms enabling ammonia partitioning into gas,
liquid and solid phases, in order to establish optimum transfer pathways. The key
knowledge gaps addressed in this research were i) determination of main criteria
for ammonia recovery technology selection for a range of industrial wastewaters,
ii) understanding the feasibility and recovery performance of struvite precipitation
and ammonia stripping at demonstration scale from distillery wastewater, iii)
understanding and quantifying the impact of transition metals and acidic organic
compounds on ammonia stripping, iv) assessment and comparison of ammonia
separation performance via ion and ligand exchange media and influence of
operation parameters (e.g. pH, buffer capacity, metal load, N concentration). The
findings are utilised to generate an informed decision process for
technology/strategy selection and the operational requirements and potential
challenges posed by selected factors, with relevance for industry stakeholders,
technology providers, and consultants. A specific focus was placed on distillery
wastewater as a case study, a sector concerned with ammonia management and
potentially suitable for recovery, particularly in Scotland.
A review of the literature found that struvite precipitation is the most widely
implemented method with industrial effluents, yet stripping and sorption
processes may be preferred for their ability to deliver versatile, ammonia-rich
solutions. The identified technology-selection criteria included the feed
concentration of ammonia and competing cations, and the struvite formation
potential. Based on the practical recommendations developed in this study, an
ammonia recovery strategy for distillery wastewater was established, integrating
anaerobic digestion with chemical precipitation and ammonia stripping coupled
with scrubbing. The performance of this treatment train had never been tested
before for filtered digestate of distillery effluent, addressing a key gap in
understanding for full scale applications. Demonstration scale trials allowed to
understand how the expected performance translated with real digested distillery
wastewater and to validate its feasibility. The results demonstrated its technical
viability, achieving 76% N removal and 80% P removal, while generating high-
quality struvite and ammonia sulphate solution. Moreover, the findings
highlighted the critical impact of pH and addressed operational challenges,
improving readiness for full-scale application.
Beyond distillery effluents, this thesis examined broader challenges in industrial
wastewaters treatment, addressing gaps identified in the literature review,
relevant for a range of industrial wastewaters, including from metallurgy and
agro/food processing. Specifically, the impacts of species found in some of these
effluents, such as transition metals (as Ni, Cu, Zn) and organic, acidic compounds
(as humic acids), on the stripping process were investigated. Results showed that
elevated levels of such species can reduce ammonia availability for stripping, via
complex formation and electrostatic interactions. This highlighted the need for
mitigation strategies to maintain stripping efficiency with these streams.
Additionally, the metal-ammonia bond potential was further explored to assess
ligand exchange (LEX) sorption mechanism as alternative to ion exchange (IEX),
a mechanism often limited by high concentrations of ammonia and competing
cations. Although various media have been tested in literature, comparative
studies on their performance under different conditions are lacking, along with
insights on how factors such as pH, transition metal and cations load can impact
their mechanisms and effectiveness. In this study, two zinc-hybridised sorption
media were tested and benchmarked against IEX media, in synthetic and real
wastewaters (distillery, municipal). The results showed effective removal,
although limited by self-inhibiting pH changes, with a zinc-hybridised media
matching or exceeding IEX resin’s performance only when pH 9-10 was
maintained (75 meq N/g). pH, buffer capacity and Zn/Na loads were
demonstrated to be critical factors to enable or limit IEX and LEX mechanisms.
The findings established operational requirements for hybridized sorption media
and provided research directions for further improvement.
Overall, this work advanced knowledge on the impact of key species on ammonia
recovery technologies, with implications for industrial effluents treatment in
general and distillery wastewater management in particular. The findings
contributed to developing recommendations for selection and operation of
ammonia partitioning strategies, optimizing metal-hybridized sorption media, and
improving process feasibility for full-scale implementationPhD in Wate