IPGRC 2026 Complete Programme

Contingent upon data availability and scope, the paper will either include original modelling results (developing and evaluating a prototype predictive model using DKD datasets) or methodology-centred design do such a model, encompassing feature sets, validation strategies, and clinical integration pathways. In both instances, the principal knowledge addition is a well-articulated translational roadmap for AI-guided regenerative therapy in diabetic kidney disease, advancing from static protocols to dynamic, patient-specific, immune, and injury-aware intervention design.

This study investigates three underexplored Fabaceae species, Thermopsis lanceolata, Genista tinctoria, and Baptisia australis, to identify and characterise their alkaloids, fatty acids, and potential cyclotides. Although Fabaceae plants are widely recognised for their agricultural and bioactive importance, these particular species remain poorly studied, especially regarding their fatty acid profiles and cyclic peptide diversity. Using solvent extraction, TurboStill fractionation, GC-MS, HPLC, and NMR analyses, the study provides new biochemical evidence including the detection of cytisine in T. lanceolata and preliminary signals suggesting the presence of cyclotides.

Objectives

• To generate foundational biochemical data that supports future applications in cosmetics, biolubricants, pharmaceuticals, and agricultural biotechnology.
• To profile alkaloids, fatty acids, and peptides in the aerial parts and seeds of three underreported Fabaceae species.
• To evaluate their potential as renewable feedstocks for high-value bioproducts within the circular bioeconomy.

Innovative Scope and Contribution to Knowledge

• Provides the first integrated phytochemical evaluation of these three Fabaceae species across alkaloids, fatty acids, and cyclotides.
• Offers novel evidence of cytisine and early indications of cyclotides, expanding current knowledge of the chemical diversity of these species.
• Demonstrates their unexplored bioeconomic potential, supporting future valorisation in sustainable bioprocesses.

Airports are essential elements of international transport infrastructure and significantly contribute to global carbon emissions. This investigation examines the potential for implementing green infrastructure at airport estates to enhance sustainability, providing benefits including temperature regulation, energy conservation, improved public health, enhanced ecosystem services, and support for urban biodiversity.

Nevertheless, a primary challenge associated with these initiatives is that vegetation attracts bird species, thereby increasing the risk of bird strikes and associated issues. To optimise environmental benefits while ensuring aviation safety, it is vital to employ effective bird-deterrent strategies alongside airport greening initiatives. The study will also review selected exemplary greening initiatives, identify best practices, and evaluate the potential for their replication across various contexts, particularly within the United Kingdom. 

The paper manufacturing and recycling industry generates substantial quantities of sludge, primarily composed of short cellulose fibers, lignin, organic compounds and inorganic fillers. Managing this waste presents significant environmental and economic challenges, due to the large volumes produced. For instance, in the United Kingdom alone, paper and cardboard waste amounted to 5.4 million tonnes in 2023, with a recycling rate of 73.4%. Biogas production from paper sludge has been proved to be a viable valorisation pathway for this waste stream however, digestion of cellulose is particularly challenging due to the ineffective enzymatic attack by anaerobic microorganisms during the hydrolytic phase.

To address this issue, a two-way ANOVA was used to design the experiment and assess the effectiveness of enhancing hydrolysis through mild hydrothermal pre-treatment (250⁰C, 5MPa) and microbial acclimation. Over a 28-day digestion period, untreated samples yielded approximately 147.8 mL of CH4 per gVS-1.Treated samples produced 15–22% more methane compared to untreated ones, however the net energy balance for embedding the pre-treatment was negative. Acclimatised untreated specimens instead achieved a maximum yield of 206.9 ml, indicating that acclimation is more effective than mild hydrothermal treatment at improving biogas production with no extra energy input.

Salivarian trypanosomes are single-cell protozoan parasites that infect mammals and are transmitted through tsetse fly bites, mainly affecting regions of sub-Saharan Africa. Infections in humans results in the Neglected Tropical Disease known as sleeping sickness which can be fatal if left untreated. In animals, infections result in diseases such as nagana which causes fever, wasting and death. The development of new drugs to treat trypanosomes with minimal toxicity in hosts is needed due to adverse side effects and difficulty in administration associated with existing treatments.

Caenorhabditis elegans is a free-living nematode capable of rapid culture generation via self-fertilisation. Assays using C. elegans have also been reported to show good correlation of toxicity data with mammals due to well-conserved genes and signalling pathways.

The aim of the project was to test molecules for potential anti-trypanosomatid activity then use C. elegans as a 3R-compliant model organism to assess their toxicity.

Initially, all 44 molecules were tested against the S427 wild-type T. brucei with the most effective then being tested against the multidrug resistant B48 strain. The molecules that were more effective than the existing drug diminazene against the resistant strain were then tested at their EC₅₀ for toxicity C. elegans. In these tests, only one molecule resulted in a significant increase in lethality but there were indications of a significant decrease in their fecundity following exposure while the results of the neurotoxicity tests proved to be inconclusive.

Pseudomonas aeruginosa is one of the most important bacterial pathogens that is also considered as one of the main causes of hospital infections. This pathogen is also known as one of the most common causes of infectious agents in pet birds. Pseudomonas aeruginosa fighting nature conteracts with antibiotics leading to emergence of resistant bacteria. The genome of P. aeruginosa is capabe of reading several efflux RND pumps which, from clinical point of view, are very important in multidrug resistance. In this regard, the MexAB-OprM pump is considered as one of the most efficient RND pumps for P. aeruginosa.

The aims of this study were to determine the existence of MexAB-OprM pump genes and antibiotic resistance pattern of P. aeruginosa isolated from birds referred to the Pet Bird Clinic of the University of Tehran. During this study, multiplex PCR was used to detect the P. aeruginosa and MexAB-OprM pumps by using specific primers.

Also using Kirby-Bauer agar disc diffusion method and according to the CLSI recommendation, the resistance pattern of the P. aeruginosa isolates was investigated against 20 antibacterial agents.

The results showed the presence of MexA and OprM in all isolates of this study. However, MexB was not detected in any of isolates. Antibiotic resistance to neomycin, kanamycin, rifampicin and vancomycin was 100%, resistance to the floroquinolones family such as ciprofloxacin, danofloxacin, norfloxacin, ofloxacin and enrofloxacin was 0 %, and nalidixic acid was 14.3%. Multi drug resistance was common among isolates.

The findings of this study showed the prevalence of P. aeruginosa among pet birds. These findings are very important for public health.

Research on wildlife microbiomes is an emerging field, and although non-human primate gut bacteria have been studied for decades, gut bacterial community analysis is a more recent development in primatology. Primates have traditionally been central to research on human evolutionary and biomedical processes; they are also an ecologically important group, which underscores their value for wildlife and conservation research. This study presents a systematic, quantitative, mapping of primate bacterial gut microbiome research to date. Using established review methods and clear eligibility criteria, 261 articles published between 2001 and 2025 were analysed. Data were extracted on primate taxonomy, geographic distribution of sampling, lifestyle type (i.e., wild, captive), research scope, and methodological approaches.

We aimed to: (1) characterise taxonomic, geographic, and lifestyle patterns of research effort; (2) examine temporal trends in publication frequency alongside shifts in research scopes and methodologies; and (3) identify research gaps with direct relevance to primate conservation. To address the lack of standardisation in methodologies and reporting practices across studies, a minimum reporting framework to be used in future studies is included.

The analysis revealed substantial research biases and gaps across primate taxa, geographical distribution, and lifestyle types. Approximately 65.5% of primate species have not yet had their gut bacterial microbiome studied, and around one-third of those that have been studied have only ever been studied in captivity, limiting ecological and conservation implications. Geographic research effort is often biased, leaving megadiverse countries such as Brazil, Indonesia, and the Democratic Republic of the Congo markedly underrepresented.

This study establishes a foundation for advancing primate gut bacterial microbiome research by identifying the key taxonomic and geographic research gaps where targeted sampling would be most valuable. The proposed minimum reporting framework offers a practical step towards standardising methods and enhancing comparability across future studies.

Chronic wounds are frequently dominated by biofilm‑forming microorganisms, whose recalcitrance drives persistent inflammation, prolonged antimicrobial use and increased risk of recurrent infection. With rising resistance and limited therapeutic options, alternative non‑antimicrobial approaches are urgently needed. Antimicrobial light has emerged as a promising candidate, however, ultraviolet wavelengths pose safety concerns. High‑energy visible (HEV) light at 405nm offers germicidal activity with a more desirable safety profile and has shown potential to support wound healing, yet the therapeutic window remains poorly defined. Current research is limited by a focus on planktonic cells and non‑standardised exposure conditions, and comparative data across multiple wound‑associated pathogens is minimal, where strain‑level variability is largely understood. This study addresses these gaps by evaluating whether 405nm light at 15-16 mW/cm² can reduce the viability of mature single-species biofilms formed by different Staphylococcus aureus, Pseudomonas aeruginosa, and Candida albicans strains, including those notably resistant. Biofilms were established using a standardised in‑vitro model and exposed to 405nm light for 1.5 hours. Viability was quantified using metabolic assays, with live/dead qPCR also utilised.

Across species, 405nm light produced measurable reductions in biofilm viability, with susceptibility varying between bacterial and fungal organisms. Consistent reductions were observed across all strains, with a minimum average metabolic reduction of 11%. Preliminary dose‑response differences in S. aureus for a various exposure times also suggest that optimisation of duration may enhance efficacy. These findings provide foundational evidence supporting the potential of 405nm light an antimicrobial alternative. By generating comparative, strain‑level data across relevant wound pathogens, this contributes to the development of safe, non‑antibiotic phototherapies for chronic wound biofilms. Future work will extend this approach into 3D tissue or organotypic wound models and refine exposure parameters to identify conditions that maximise antimicrobial activity while preserving tissue integrity.

This research develops and evaluates a hybrid text classification architecture designed for interpreting informal, diverse and indirect food descriptions. Existing text classification approaches are either based on neural models that have strong contextual understanding but become unstable when wording changes, or on rule-based systems which provide transparency but cannot handle diverse representations. The objective of this work is to bridge this difference by designing a system that is both reliable in noisy language conditions and interpretable in everyday decision making. 

The work offers a novel contribution by unifying five elements that have not previously been combined in a single classification framework: transformer-based embeddings, prototype driven semantic structures, reinforcement learning refinement, and a rule layer to combine fuzzy cues with symbolic constraints. In this design, all components have distinct role. BERT models meaning, prototypes provide stable reference points, PPO refines uncertain embeddings, and the rule layer applies transparent domain adjustments. The system improves over time through an offline update process that learns from  previously collected user inputs and model decisions. The encoder remains frozen to prevent embedding drift and catastrophic forgetting, while PPO and the rule base evolve by enabling controlled continual learning without disrupting the semantic space.

Evaluation shows that rule-based layer resolves uncertain or indirect descriptions, PPO reduces sensitivity to wording changes, and the integrated architecture delivers more consistent behavior than neural or rule only baselines. This research extends current knowledge by demonstrating a practical method for combining neural models, reinforcement learning and interpretable logic into one coherent decision pipeline. This offers a new direction for text classification in application areas where language is unpredictable, and explanation is essential.

The integration of Artificial Intelligence (AI) into daily life is often framed as a ‘remedy’ for the historical marginalisation of people with disabilities. From automated captioning1 to real-time scene description2 and predictive healthcare3 , the promise of “AI for Social Good” suggests a future where digital and physical barriers are dissolved by intelligent algorithms. 4 However, a systematic examination of AI development reveals a more troubling reality: these systems frequently leave their intended beneficiaries what one may call “disabled by design.” Many AI technologies aimed at addressing disability-related challenges—such as hearing aid quality enhancement5 or alternative text (alt-text) generation6 —are designed primarily to correct perceived deficits, rather than in partnership with disabled individuals. Disabled users are often treated as passive endpoints rather than active participants in the design process. Consequently, when these systems are deployed, they often fail in real-world contexts7 , giving the appearance of accessibility while leaving users functionally disadvantaged. This phenomenon arises from a mismatch between the statistical logic underlying AI and the lived experiences of disabled people rather than technical limitations. By prioritising normative data, medicalised frameworks, and top-down engineering over participatory co-design, AI development produces a generation of systems that exclude, stigmatise, and impose disproportionate verification labour on those they are intended to serve. 

To assess the extent of this exclusion, this paper examines a corpus of 45 recent studies on AI for inclusivity and accessibility. These papers were identified through systematic reviews of work on digital accessibility and inclusive design and were informed by influential research from leading venues such as the ASSETS and CHI conferences (2020–2025). The analysis highlights a “participation deficit”: the gap between research activity in this area and the meaningful involvement of disabled people as co-creators of these technologies.

The foundation of Natural Language Processing (NLP) language models is a mathematical framework that identifies patterns in how language is structured and uses statistical probabilities to generate meaningful responses for users. Due to this the “default assumption for many was that computation, deriving from mathematics, would be pure and neutral” (Caliskan et al). However, relying solely on a probabilistic approach to provide accurate and trustable Language Model (LM) answers for a fast-changing environment —especially using unverified LM training data scraped from the internet—can be highly problematic.

Abstract

This framework introduces a dynamic, metadata-driven and novel certification system for generative AI models deployed in high-risk, real-time environments, called the “Trusted AI Accuracy Labeling Mechanism”. It emphasizes accuracy, provenance, and reproducibility through a multi-layered labeling architecture that integrates data curation, model architecture, training metadata, and empirical evaluation. This mechanism implements the capability to publish domain or task specific Language Model accuracy and transparent accountability through an Trusted Labeling Framework API interaction.

Framework core concept

While research into model performance and hardware footprint reduction is in abundance, significant gaps remain in achieving strict output accuracy—especially in high-risk domains. These challenges are compounded by data curation practices that lack standardization and often operate without transparency. The prevailing momentum toward “bigger, better, faster” models have not been matched by equal rigor in accuracy assurance, leaving both the field of AI and its human consumers exposed to unverified outputs. To address this, this research proposes a Dynamic Certification Engine—a config-aware, metadata-driven labeling pipeline designed to assess and certify AI models for extreme emergency response. This system delivers actionable, real-time, trustable AI solutions tailored for on-campus, multi-building scenarios, where accuracy, latency, and human safety are paramount (Extreme Emergency scenarios).

Vehicle-to-Everything (V2X) communication systems require intelligent task offloading to meet ultra-low latency requirements (20-100 ms) under high vehicular mobility and dynamic channel conditions. A core challenge is that existing approaches, such as deep reinforcement learning and mixed-integer optimization, either incur prohibitive training overhead or exhibit exponential computational complexity, and critically, most do not explicitly account for mobility dynamics, resulting in degraded performance during frequent handovers between roadside units. To address these limitations, this paper proposes a Mobility-Aware Gradient Descent Offloading (MAGDO) algorithm for three-tier V2X edge-cloud systems. MAGDO introduces four key innovations: (1) a continuous and differentiable offloading formulation enabling partial task execution across local, edge, and cloud layers with fine-grained resource control; (2) mobility-aware gradient computation incorporating vehicle velocity, direction, and handover probability;

(3) an adaptive learning rate mechanism responsive to channel variability and mobility dynamics; and (4) a velocity-scaled momentum strategy balancing convergence stability and responsiveness. We formulate a multi-objective optimization problem that considers latency, energy consumption, cost, task success rate, resource utilization, and the handover resilience metric. Simulation results demonstrate a statistically significant 28.8% reduction in latency and 52.4% increase in energy efficiency compared to state-of-the-art deep reinforcement learning baselines (p < 0.001), with 99.8% handover resilience, linear time complexity O(K), and a decision latency of 2.3 ms, making it 63 times faster than mixed-integer programming. MAGDO achieves a 73.2% task success rate with 60 vehicles, versus 36.5% for competing strategies, establishing it as a practical, scalable solution for time-critical autonomous driving applications. 

The accelerating growth of criminal activity within high-speed blockchain ecosystems presents a critical investigative challenge. Networks such as Solana enable transactions to be executed and obfuscated within minutes, leaving law enforcement and forensic analysts with an increasingly narrow window to detect, trace, and preserve evidence. By 2026, illicit cryptocurrency flows are estimated to have reached $158 billion, underscoring the scale of the threat. Existing investigative tools are largely designed to monitor wallet balances and surface-level transfers, leaving a significant visibility gap within complex smart-contract interactions. This “tool gap” prevents investigators from observing hidden asset movements embedded in Cross-Program Invocations (CPIs), where sophisticated attackers frequently conceal laundering activity. Simultaneously, the speed of modern blockchains creates acute time pressure, making manual analysis impractical and exposing the limitations of static, rule-based detection systems that adversaries can easily circumvent. To address these challenges, the primary research objective is to develop a specialized forensic engine that automates the real-time reconstruction of nested CPI calls. 

This methodology advances the state of the art by transitioning from reactive address-based monitoring to proactive instruction-level tracing. The system introduces an advanced transaction-tracing methodology that examines internal smart-contract execution paths, enabling investigators to observe hidden asset flows that conventional monitoring fails to detect. This approach replaces rigid threshold-based detection with a machine-learning anomaly-detection model using an Isolation Forest architecture. By learning patterns from large volumes of legitimate transactions, the system autonomously identifies behaviors that deviate from expected norms. The framework is empirically validated using a Solana Local Validator environment, measuring performance through concrete outcomes including detection latency (milliseconds) and F1-score accuracy against simulated exploit scenarios. By prioritizing evidential reliability, the platform ensures analytical outputs are translated into court-ready evidence aligned with NIST and ISO 27001 guidance.

This study introduces a novel hybrid deep learning method for unsupervised anomaly detection of energy consumption time series, which combines a sparse LSTM autoencoder with a GAN. Accurate monitoring of abnormal demand and pricing profiles is of utmost importance in modern power systems to detect potential failures, data integrity issues, non-technical losses, and unrealistic market behavior that can threaten grid stability and restrict energy efficiency.

The proposed model, denoted Sparse-LSTM AE–GAN, adopts a sparsity-regularized LSTM autoencoder to acquire compact and interpretable latent representations of normal energy consumption sequences. A GAN is subsequently trained adversarially between these latent codes and reconstructions to match the distribution of the generated “normal” patterns with that of observed data. To determine anomaly scores, reconstruction errors, and adversarial consistency are simultaneously taken into account, thereby enabling the model to identify small variations in operations that may still be significant at the system level.

We aim to test the method on Italian electricity price data and to compare it with state-of-the-art techniques such as LSTM autoencoders, LSTM-GANs, or classical statistical-based detectors. The relevant performance measures include precision–recall for rare events, robustness to noise, and sensitivity to variations between the various types of anomaly patterns.

The main novelty is a new hybrid architecture that combines a sparse encoding sequence with adversarial learning in the energy domain, which leads to enhanced detection performance as well as enhancing the interpretability of latent features. The goal is to enable the design of more reliable and data-driven monitoring tools for smart grids and energy markets that could potentially impact sustainable energy management and operational decisions.

This dissertation presents the first systematic spatial analysis of honour-based violence in a South Asian city, demonstrating that Karachi’s persistent colonial street grid and domestic compound design actively enable planning, execution, concealment, and prevention of escape in honour killings. Focusing on Lyari and Orangi Town, the two highest-incidence neighbourhoods, the study combines (a) space syntax configurational analysis (global/local integration, choice, and isovist modelling) of colonial-era (1897–1935) and present-day maps with (b) close reading of 28 published survivor testimonies from Aurat Foundation, Panah Shelter, HRCP, and Dawn archives (2018–2025). The research produces the first “escape-failure maps” showing how low-permeability routes to the nearest shelters (Panah, Chhipa, Darul Aman) structurally block women’s flight.

Objectives: 

1.Trace the colonial origins of one-door homes and alley systems in prevalent high-risk areas.  

2. Examine how these features enable honour killings using space syntax.

3. Map failed escape routes to safe houses and identify architectural barriers.  

4. To correlate low integration and low visibility factors with survivor testimonies.

5. To demonstrate how colonial planning logics continue to serve patriarchal surveillance.

Innovative Scope & Original Contribution 

No existing study has ever applied space syntax to honour-based violence, to domestic-scale analysis, or to any post-colonial South Asian context. By treating colonial urban form as an independent variable rather than a neutral backdrop, the dissertation shifts the explanatory frame from “cultural inevitability” to material complicity, providing the first empirical evidence that architecture itself is a perpetrator. The resulting integration maps, annotated with survivor voices, constitute a new methodological tool – “feminist spatial forensics” – transferable to other cities where colonial grids and gendered violence intersect.

The dissertation is therefore an original research contribution that bridges decolonial theory, spatial science, and survivor testimonies.

Nigeria’s heritage buildings reflect a mosaic of cultural identities expressed through indigenous (vernacular), non-indigenous (of foreign origin), and hybrid (combined vernacular and foreign) architectural forms. This paper argues that heritage buildings function as carriers of cultural identity, shaping and sustaining sense of place through their material, spatial, and symbolic qualities. It is grounded on the identity theories developed by  Mary Jane Collier and Milt Thomas (circa 1800’s) and Hall (1990), whose application is under-represented in African-centred heritage conservation scholarship. In addition, Nigerian researchers such as Osasona (2017); Okpalanozie and Adetunji (2021); Adewumi (2022); Folasiji (2022) etc., have consistently emphasized the need for clear documentation of any new interventions made on historic structures, highlighting the need for a centralized digital inventory of heritage buildings.

This study addresses the above gaps by examining how cultural identity is articulated through heritage buildings in Nigeria, and how these expressions inform conservation. The adopted qualitative research methods included a literature review on cultural identity, sense of place, material culture and conservation theory at global and Nigerian scales, and case studies of selected heritage buildings in the South-South and Northern region of the country. The latter entailed archival research, on-site architectural documentation, interviews and community engagement, to assess cultural significance, patterns of use, physical condition, and conservation challenges.

Findings demonstrate that heritage buildings operate as forms of material culture through which collective memory and place-based identity are produced and negotiated. Indigenous buildings reinforce local cultural values and practices, while non-indigenous and hybrid structures function as mnemonic markers of historical epochs that have contributed to the layered identity of place. The paper contributes to heritage and built environment discourse by proposing a palimpsest-informed, context-responsive conservation approach that recognises multiple cultural narratives within heritage buildings. It also offers insights to researchers and practitioners concerned with culturally grounded conservation.

This paper critically analyses how women in the UK contributed towards the turn to participatory design methods in planning, in the timeframe of 1980-1995, during the period commonly described as the Second Wave of feminism.  

A significant push towards user participation in the design process was seen in the UK during the 1970s as part of the community architecture movement – for example, as a response to dissatisfaction with social housing.  It was thought to improve the outcomes of the final design and became mainstreamed by the election of Rod Hackney (a key proponent of the community architecture movement) as President of RIBA in 1986. This push towards user participation can be understood as a precursor to the current incorporation of ‘co-creation’ as a method for improving designs and encouraging strong connections between people and the spaces they occupy.

Previous literature has explored developments in participatory design methods in terms of major social housing projects such as the Byker housing development (1969 -1983), however, the role of women in the development of participatory design has been sidelined in this history. This paper therefore explores how the groups Matrix, the Women’s Design Collective and the women in planning group at Greater London Council worked with a radically different approach for their time, questioning the traditional hierarchy between architect/planner and user.

This dissertation investigates how colonial architecture in the Belgian Congo functioned as an ideological instrument of power through the design and representation of building facades. Rather than treating architecture as neutral backdrop to colonial administration, the research argues that facades operated as visual and symbolic interfaces through which authority, hierarchy, and European dominance were normalised within the colonial city.

The study focuses on Leopoldville (present-day Kinshasa) during the late colonial period, analysing three institutional buildings: the Palais de la Nation, the Central Post Office, and the Cathedral of Notre-Dame du Congo.

Using a qualitative, interpretive methodology grounded in architectural semiotics, the research draws on the theoretical frameworks of Michel Foucault, Henri Lefebvre, Gottfried Semper, and Roland Barthes.

Each façade is analysed across three semiotic levels – denotation, meaning, and myth – to examine how architectural form, scale, symmetry, and stylistic references contributed to the production of colonial authority as natural, legitimate, and enduring. The analysis demonstrates that classical monumentality, bureaucratic order, and religious symbolism were strategically employed to visually encode sovereignty,

administrative control, and moral superiority, while concealing the coercive and extractive systems underpinning colonial rule.

The dissertation introduces the concept of the “façade of evil” to describe how architectural aesthetics can participate in the normalisation of systemic violence without directly depicting it. By foregrounding the façade as a critical site of ideological production, this research contributes to architectural history and postcolonial scholarship, offering a framework for understanding how colonial power was spatially communicated, and how its material legacies continue to shape urban memory and postcolonial identity in contemporary Kinshasa.

Malware continues to increase in sophistication, leveraging obfuscation, packing, and evasion strategies that undermine the effectiveness of conventional signature-based detection methods. Although machine learning approaches have shown strong capability in static malware detection, their deployment in operational security and digital forensic settings is often limited by a lack of interpretability. This study examines the performance of tree-based machine learning models, specifically Random Forest and Gradient-Boosted Trees, applied to static Windows Portable Executable (PE) features, with an emphasis on explainability and forensic applicability. To address the limited empirical evaluation of explainable tree-ensemble models in PE-based malware detection, SHAP and LIME are used to derive transparent explanations of model decision processes. Experiments are conducted using cleaned benchmark datasets within a constrained cloud-based environment, evaluating classification accuracy, explanation efficiency, and integration into digital forensic workflows.

The results indicate that both models achieve high discriminative performance while consistently linking predictions to forensically relevant indicators, such as PE header irregularities and distributed opcode usage patterns. The findings reveal that interpretability varies across feature representations and demonstrate that explainable tree-based models can effectively support analyst triage, evidential reasoning, and transparent reporting. Overall, this study highlights the practical benefits of integrating robust classification models with complementary explainability techniques to enable scalable, defensible, and operationally viable malware detection. Index Terms – Random Forest, Gradient-Boosted Trees, SHAP, LIME, Portable Executable (PE) Features, Malware Detection, Digital Forensics.

Smart home systems (SHS) face serious cybersecurity and privacy concerns that require a more sustainable solution. These systems are networks of connected devices within the home, including doorbells, smart locks, lighting systems, TVs, smartphones, Bluetooth devices, and CCTV. The growing use of smart home systems has made it one of the fastest‑growing segments of the Internet of Things (IoT), offering enhanced automation, convenience, and user management. However, these benefits come with significant cybersecurity and privacy challenges.

This paper presents a comprehensive examination of cyber threats, vulnerabilities, and privacy risks affecting smart home systems, drawing on an extensive literature review of 20 peer-reviewed studies published between 2022 and 2026. It identifies and classifies the most prevalent cyberattacks and associated access-control challenges in smart home systems, conducts a comparative analysis of the literature, and evaluates existing technologies, including Artificial Intelligence (AI)-integrated solutions. Traditional signature-based security mechanisms fail to detect emerging and sophisticated threats, while privacy protections are insufficient to prevent behavioural inference and misuse of personal data.Key cybersecurity problems identified in the literature include weak password policies, insecure firmware updates, vulnerable communication protocols, insufficient encryption, DDoS attacks that exploit resource‑constrained IoT devices, malware targeting embedded systems.

This paper presents a comprehensive analysis of cybersecurity and privacy challenges in smart homes, reviews recent peer‑reviewed literature from 2022 to 2026, and proposes a research framework for future work, including a research agenda and a conceptual framework to guide the development of secure, privacy‑aware smart home systems.

The objective of this paper includes the following

To conduct a systematic review of the most common cyber attacks and privacy issues in smart home systems.

To present a comparative analysis of 20 Peer‑Reviewed Papers (2022–2026), highlighting the Authors and their Approaches, tools, problems solved, and Identified Gaps.

The rapid growth of the Internet of Things (IoT) has intensified the need for secure communication mechanisms tailored to resource-constrained devices. Although the Advanced Encryption Standard (AES) provides strong cryptographic guarantees, its conventional implementations introduce computational and energy overheads that limit efficiency in low-power IoT environments. While prior research has proposed lightweight block ciphers such as PRESENT and TWINE, these alternatives often trade standardisation, interoperability, or robustness for efficiency.

This study proposes an enhanced AES-based lightweight encryption framework that preserves the security strengths of AES while optimising performance for IoT applications. The framework integrates a Dynamic Encryption Key System (DEKS) that enables periodic symmetric key rotation using cryptographically secure pseudorandom generation and secure Transport Layer Security (TLS)-based distribution. This approach enhances forward secrecy and mitigates key reuse vulnerabilities common in static AES deployments.

A comparative evaluation of AES cipher modes—Cipher Block Chaining (CBC), Counter (CTR), and Galois/Counter Mode (GCM)—was conducted to determine the most suitable configuration for IoT systems. Experimental results demonstrate that CTR mode achieves the lowest execution time and reduced memory consumption while maintaining robust confidentiality.

The framework was validated through a Python-based graphical IoT simulation integrated with MQTT communication, demonstrating secure real-time sensor data transmission with minimal overhead. The proposed approach contributes a scalable, experimentally validated, and standards-compliant lightweight AES solution tailored for secure IoT applications.

This study contributes, a structured Dynamic Encryption Key System enabling forward-secure AES deployments in IoT an empirical performance validation identifying CTR mode as optimal for resource-constrained devices, and  a practical GUI-based implementation demonstrating real-time secure communication.

Mainly because of the Covid -19 Pandemic, Remote and Homeworking (WFH) are now firmly embedded in UK and international working practices, but IT Services down periods due to cyber-attacks create both inefficiencies and is a threat to large scale activity. UK higher education institutions have faced many serious cyber-attacks, the literature demonstrating that UK universities experience incidents far more frequently than the average UK business. Mitigation and management of cybersecurity and cyberthreats is increasingly urgency because universities are currently going through considerable staffing disruption and change due to funding and recruitment crises. Such large central staffing losses have implications for the IT support offered to staff working remotely, making cybersecurity measures a current and critical consideration for universities. But there remains an urgent need to explore and understand how individual staff respond to and interpret cyber threat, in relation to their management of both during periods of remote work and work from home.

This scoping review reveals that literature within UK Higher Education (the context for this study) is relatively scarce in this area and so a scoping review was carried out to examine the current state of the field. Cross-institutional research (Ho & Gross, 2021) reveals that collective responsibility for cyber threat management is the most effective method to embed and sustain risk understanding long term. However, analysis of individual behaviour toward risk and threat shows that there are significant differences between institutional perceived effectiveness of education programmes and the reality of risk evaluation in action, on a day-to-day basis (Madarie et al, 2025). This research will contribute important original knowledge that will help higher education institutions, and their senior leaders in learning, research, and information and learning technology services, to develop effective and workable cybersecurity policies and practices that support and facilitate teaching and learning.

The need for inclusive design in architecture is essential in making the world more accessible for disabled people. However, it can be difficult for architects to understand the needs of disabled people. Existing research suggests one way to mitigate this is through experiential simulation, where participants experience a real time simulation of disabilities. Virtual reality (VR) is gaining popularity and may be able to simulate visual impairments more realistically than other simulation methods. There is a gap in research that use an authentic context in VR; existing simulations have been performed in a fictitious VR environment, and thus none have centred around heritage, where accessibility can be deemed more difficult to implement, or less important, than in modern architecture.

This paper closes this gap by providing a cohort of Master of Architecture students and staff at University of Salford with an experiential simulation of several visual impairments via VR. The simulation is set in Portmeirion- a village in Wales with high cultural importance which has been previously visited by all participants. In the simulation, participants first moved around the Portmeirion VR model with no impairments, then looked around the model overlaid with filters simulating symptoms of cataracts, colour-blindness, loss of visual field, and loss of central vision.

Pre-and post-simulation questionnaires focussed on participants empathy towards visually impaired people; it also questions how important participants think accessibility is for cultural heritage sites. Descriptive statistics of the Likert-scale responses measure the impact of the simulation on participants empathy and perceived importance of inclusive heritage environments.

Overall, the research provides a methodology for experiential simulation of visual impairments in a university environment, offering a contribution to inclusive design pedagogy. It answers whether an experiential VR simulation of visual impairment measurably improves architecture students’ empathy and perceived importance of accessibility in heritage environments.

The UK housing association sector is under growing pressure to modernise asset management in response to evolving safety legislation, sustainability targets, and digital transformation agendas. Despite national standards such as BS EN ISO 19650 and sector-specific guidance like the BIM4HAs Toolkit, many housing associations struggle with data management, limited digital skills and training, inadequate organisational readiness, and cultural resistance to change. Providers are required to comply with new regulatory expectations under the Building Safety Act and the Golden Thread of information.

This paper investigates the barriers, enablers, and readiness factors shaping BIM adoption within housing associations. Using the People–Organisation–Process–Information–Technology (POPIT) framework as a conceptual lens, the study examines how organisational structures, leadership behaviours, information workflows, digital capability, and technological infrastructure interact to influence BIM adoption in a sector dominated by retrofit activity, asset management, and resource constraints.

The research highlights several recurring challenges, including financial limitations, digital skills gaps, complexity of standards focused on new-build contexts, fragmented data systems, and organisational resistance to ISO 19650. However, it also reveals emerging opportunities, such as Golden Thread & Safety Compliance, lifecycle asset management, Energy Efficiency & Sustainability, Cost & Lifecycle Efficiency, Stakeholder Collaboration, and Resident Experience.

Building on these insights, the paper proposes the development of a BIM Adoption Framework tailored to housing associations. The framework provides a structured pathway for assessing digital maturity, identifying capability gaps, and taking targeted steps toward BIM adoption. It serves as both a diagnostic tool and a strategic guide for improving information management.

The study contributes to ongoing digital transformation by addressing a gap in the research relating to socially oriented, low-adopting sectors. It supports practitioners seeking realistic, context-sensitive approaches to digital adoption that enhance long-term value, safety, and efficiency within the UK housing association sector.

Keywords: BIM adoption, housing associations, ISO 19650, POPIT, BIM4HAs

Small and medium-sized enterprises (SMEs) play a critical role in the construction industry; however, they often operate under conditions of limited resources, constrained digital capability, and high exposure to project-related risks. In comparison to large organisations, SMEs typically rely on informal processes, experience-based decision-making, and fragmented information management practices. These conditions increase vulnerability to cost overruns, scheduling delays, coordination failures, and quality-related risks. This paper examines the potential of BIM to support more effective risk management within construction projects delivered by SMEs.

The study focuses on the specific challenges faced by SMEs in adopting BIM, as identified in existing literature and reflected in practical industry conditions. Rather than addressing BIM as a fully integrated enterprise-wide system, the research explores how selective and task-oriented use of BIM tools can contribute to improved risk identification, coordination, and decision-making in SME-led projects. A quantitative research approach was adopted, drawing on data collected from construction professionals working within small and medium-sized firms involved in building projects where BIM adoption remains limited or partial.

The findings indicate that even modest levels of BIM implementation can provide tangible benefits for SMEs. BIM-supported visualisation and coordination tools were found to improve understanding of project scope, reduce design-related uncertainties, and enhance communication among project participants. These improvements contribute to more proactive management of technical, time-related, and cost-related risks, particularly during early project stages. the study highlights persistent barriers to BIM uptake in SMEs, including financial constraints, lack of specialised skills, and uncertainty regarding return on investment.

The paper concludes that BIM can function as a practical risk management support tool for SMEs when implemented in a context-sensitive manner. By aligning BIM use with the operational realities of small and medium-sized firms, digital adoption can be positioned as an achievable and valuable step toward improved project outcomes rather.

The use of uncrewed aerial systems (UAS), or drones, has rapidly expanded across scientific, commercial, and medical domains. In ecology, drones are valuable for data collection and monitoring, yet their noise may affect wildlife. Birds, with over 11,000 species and global distribution, are ideal model organisms due to their sensitivity to environmental change. This research investigates whether evolutionary history and morphology predict avian sensitivity to drone noise. Using secondary data, we compiled vocal and hearing traits, morphology, and ecological attributes for over 70 species and analysed these using phylogenetic and allometric models. Initial results suggest that vocal learners (e.g., Psittaciformes and oscines—or songbirds), due to their ability to modulate vocalizations, and species with lower dominant vocal frequencies are less sensitive to drone noise because of reduced masking effects.

We predict body mass, cranial morphometrics, inner ear structures (columella, cochlea, basilar papilla), and vocal apparatus morphology (syrinx, trachea) to be key indicators of vulnerability. Next, models will be refined using additional secondary data to identify a best-fit “total evidence” approach combining phylogenetic and allometric methods, validated against road traffic noise as an acoustic analogue. Finally, these models will be tested using primary data from simulated drone noise playback experiments, recording responses of native UK bird species across two contrasting habitats with different ambient noise profiles. This research aims to improve understanding of species-specific responses to anthropogenic noise and inform guidelines for drone use in sensitive areas.

This research aims to investigate how circular-economy principles can guide procurement strategies implemented within the UK’s public sector. It examines the feasibility of Circular Public Procurement (CPP) by analysing the perspectives, attitudes, and influence of key stakeholders across the procurement supply chain. Since the public sector is an important consumer in the economic system, it has a huge potential to promote the use of sustainable products and lifestyles. To achieve the 2050 net-zero target and 2030 Sustainable Development Goals, strengthening the application of circular procurement in the UK public sector will help to accelerate the transition to a circular economy in the UK and achieve the GHG reduction target. Current legislation, including the Procurement Act 2023, provides limited guidance on circular economy. This gap underscores the need for clearer definitions, frameworks, and evidence.

The study has two main objectives:

Develop a clear and academically grounded definition of CPP, addressing fragmented terminology and conceptual ambiguity in existing literature. This contributes directly to theory-building and policy clarity.
Conduct sector-specific empirical evidence by investigating CPP practices, as well as the barriers and challenges behind them, to figure out the potential opportunities of transition to a more circular way of the UK public procurement. Investigate the perceptions and attitudes of stakeholders towards circular public procurement, and map out their influence on the decision-making and implementation of circular public procurement policies.

By exploring stakeholder attitudes, challenges, and drivers, the research identifies what enables or hinders CPP adoption in real organisational contexts. The findings aim to support government efforts to strengthen circular economy policy, fill the current gap in procurement-related legislation, and provide key empirical evidence from the public sector perspective. Ultimately, the study will contribute a new conceptual clarity, practical insights, and policy-relevant evidence to advance sustainable and circular public procurement in the UK.

This study proposes a hybrid Artificial Intelligence-Building Information Modelling (AI-BIM) framework to support multi-objective decision-making in construction project management, with a focus on cost (5D), time (4D), and sustainability (6D). Motivated by the increasing complexity of project delivery and the aims of Construction 4.0, the framework integrates machine learning, predictive analytics, and optimisation within BIM environments to simulate, forecast, and improve project outcomes in near real time. The approach links design, construction, and early operation, enabling proactive management of schedules, budgets, and environmental performance indicators.

Methodologically, the research adopts a mixed-methods design: a systematic literature review and case studies underpin the conceptual architecture, while computational modelling implements data pipelines and algorithms within Autodesk Revit and Navisworks. The expected contributions are i) a unified AI-BIM architecture for 4D/5D/6D integration; ii) a configurable optimisation engine for simultaneous schedule, cost, and sustainability targets; and iii) empirical evidence of performance gains on real-world projects. The work aims to provide practitioners with an interpretable, data-driven decision support tool aligned with industry digital transformation goals.

This paper presents a conceptual AI Framework for Real-Time Project Performance (AiN-RTP) management, specifically engineered for the unique contractual dynamics of the NEC ECC Option A (Activity Schedule) and Option B (BOQ). The objective of this paper is to move project control from retrospective reporting to proactive, data-informed intervention, aligning technological capability with the principles of NEC contractual governance. The framework is a multi-layered architecture that integrates diverse data streams like Digital Twin models, real-time IoT sensing, and financial data with a core AI Analytical Engine. The innovative scope and contribution to knowledge lie in two areas – a) contractual NLP integration and b) differentiated contractual utility. The framework incorporates a Natural Language Processing (NLP) subsystem to analyse past data of historical projects and unstructured text from Early Warning (EW) notices and Compensation Event (CE) proposals, providing a probabilistic risk severity score that forecasts delay with potential time and cost impacts before formal assessment or actual occurrence.

This framework would provide distinct value streams for NEC Options A and B. For both the options, live project data is verified by AI to deliver a transparent project performance with an objective Estimated Final Cost (EFC) forecast, margin visibility, and deviation control against the lump sum scope or as per the quantity. In either of the scenarios, the AI breaks down the cost and timeline. This conceptual AI framework is validated through a synthesis of peer-reviewed literature and a qualitative industry expert workshop. This paper’s original research contribution is the theoretical foundation for implementing a cohesive, AI-driven governance system that enforces the collaborative and forward-looking spirit of the NEC contract, Option A and Option B, respectively, drastically reducing reactive disputes and improving project predictability.

This paper presents findings from a questionnaire study examining the professional implementation of thermal imaging (TI) within the UK retrofit industry. While TI is widely recognised as a valuable diagnostic tool for identifying heat loss and performance defects, its use in retrofit practice remains inconsistent and largely informal. Existing research has focused predominantly on the technical capabilities of thermography, offering limited insight into how TI is understood, applied, and integrated by practitioners in real-world retrofit contexts. This study addresses this gap by investigating practitioners’ familiarity, confidence, training, access to equipment, communication practices, and perceived barriers affecting the adoption of TI.

Data were collected from 31 professionals across the UK, including surveyors, contractors, retrofit coordinators, policy specialists, and researchers. The findings reveal a clear disparity between awareness of TI and its routine application in retrofit workflows. Key barriers include high equipment costs, limited access to certified training, inconsistent interpretation skills, and weak integration of TI within established standards and assessment frameworks. Although participants highlighted the communicative and trust-building value of thermal imagery, results are often poorly embedded in decision-making and quality assurance processes.

The significance of this study lies in providing empirical evidence of the systemic and institutional factors constraining the effective use of TI in retrofit practice. By foregrounding professional realities rather than technical performance alone, the paper informs the development of more robust training pathways, standardised reporting approaches, and regulatory integration strategies. These insights support efforts to enhance retrofit quality, evidence-based decision-making, and the wider adoption of thermal imaging within the UK housing sector.

The construction industry continues to grapple with challenges relating to cost and time overruns. Meanwhile, delivering construction projects within budget and the expected timeframe with specified quality are some of the hallmarks of a successful project delivery. However, this seems practically impossible or rather difficult to achieve. Unsurprisingly, industry stakeholders are increasingly demanding project cost efficiency, sustainability, and improved project delivery; therefore, there is an urgent need for the industry to meet these demands. In response, emerging digital technologies, such as building information modelling, big data analytics, machine learning, and artificial intelligence, have been widely suggested as transformative tools capable of enhancing construction cost management practices.

This study investigates the role of digital technologies in improving construction cost management through a comprehensive review of existing literature. Employing a secondary data approach, the research will utilise statistical methods to analyse and synthesise findings from previous studies. The anticipated outcomes aim to offer actionable insights for reducing cost overruns, fostering sustainable cost management, and enhancing overall project delivery. Also, the study’s findings are expected to contribute to the existing body of knowledge by enhancing and transforming construction cost management practices, thereby enabling professionals to deliver more efficient, cost-effective, and successful projects.

Cost overrun is a longstanding worldwide issue that affects both developed and developing countries. Its comprehension requires a vivid appreciation of how various stakeholders perceive the root causes as this shape how they approach and mitigate it. Despite the numerous studies on cost overrun, there is limited attention to how client, contractors, subcontractors, consultants and project managers perceive its root causes particularly in the downstream oil and gas sector. Hence, this study examines the perception of cost overrun amongst key project stakeholders using eight categorised causal factors obtained from 32 causes of cost overrun established in literature. A quantitative research method was adopted using a structured questionnaire survey administered purposively to stakeholders from three downstream oil and gas megaprojects. Out of 286 administered questionnaire, 213 responded (206 valid) with a 74.4% response rate. Relative Importance Index ranked the severity of each causal category. Levene and Kruskal-Wallis tests assessed perception differences across stakeholders. Post-hoc analysis (Dunn BH and Tukey HSD) established specific group differences.

Findings reveal that market and external factors ranked 1st (RII=0.78), design and scope changes ranked 2nd (RII=0.75) and execution and implementation challenges ranked 3rd (RII=0.70). Four causal categories showed significant perception amongst the stakeholders: preliminary stage challenges (p=0.0176), tendering and contract challenges (p=0.0394), bureaucratic & regulatory challenges (p=0.0095) and cultural & trust issues (p=0.0227). Post-hoc comparisons revealed that the subcontractor consistently perceived greater severity in the four significant causal categories as compared to other stakeholders. This implies the subcontractor suffers more from the impact of cost overrun due to his role as the ground man who implements distinct aspects of the project and faces the dynamism of the market. Therefore, the study recommends early involvement of all stakeholders particularly the subcontractors in front end planning of the project.

Noise is highlighted by the World Health Organisation (WHO) as the second most harmful environmental pollutant with a growing body of evidence demonstrating its negative impacts on human health. UK policy, guidance and standards prescribe acceptable noise levels, manage noise as a statutory nuisance and regulate noise-intensive activities via licensing. Human-centred noise monitoring is therefore tightly controlled, with measurement standards governing the precision of instrumentation. However, there is limited consideration to the effects of noise on wildlife. Non-human animal species produce, perceive and use acoustic signals differently to humans, across a much broader spectrum meaning the protections in place may not sufficiently safeguard animals. With human hearing being limited to what we term the audible range and being reflected in our regulations and instrumentation, infrasonic and ultrasonic sounds may be vastly underestimated.

Long term passive acoustic monitoring (PAM) and direct sound measurements of environmental soundscapes with handheld devices to record audio are an essential component of wildlife and biodiversity monitoring. With the application of artificial intelligence, with applications such as automated identification of species and habitat health indicators, the use of digital audio recordings has accelerated. Devices previously designed for presence/absence surveys are becoming progressively more relied upon for other uses such as noise monitoring to inform wildlife protection. This paper reports the results of standard test procedures used for human-focused equipment, applied to wildlife audio recorders to evaluate their suitability for monitoring noise affecting wildlife within an existing regulatory framework.

The increasing effects of Urban Heat Island (UHI) in high-density and historical neighborhoods have been a growing concern recently. Public squares play a vital role in supporting social interaction and entertainment; however, these spaces are under heat stress. In Seville’s Casco Antiguo district, such conditions especially exist in large historic squares. In this study, Alameda Square in northern Casco Antiguo was evaluated to improve thermal comfort. This square is the largest and most important social gathering place for locals and tourists, with a long bike path and streets that face discomfort. Few studies assess mitigation strategies in historic Mediterranean squares. In this study, thermal conditions in the square are examined using both on-site measurements and ENVI-met simulations. Key thermal parameters such as air temperature (AT), globe temperature (GT), surface temperature (ST) and mean radiant temperature (MRT) were monitored, and the Universal Thermal Climate Index (UTCI) was selected as the most widely used in outdoor thermal comfort studies.

Three pavement strategies and one shading solution will be modeled: 1. Reflective pavements, 2. Photovoltaic-integrated pavements with vegetation, and 3. Paved grass and canopies in the gathering places of the Alameda Square. These strategies were selected based on a balance between intervention limitations in the historic district and effectiveness of the strategies used based on existing literature. Comparative simulations will explore potential improvements in thermal comfort before and after implementing mitigation strategies. Preliminary results show that interventions can reduce Land Surface Temperature (LST) above 4.30°C, UTCI between 2.20°C–3°C, and decrease MRT by 7.8°C–10.6°C. This combined approach suggests a sustainable route for restoring historic Mediterranean urban areas, contributing positively to sustainability and social inclusion.

Despite Green Infrastructure (GI) presenting significant potential for providing diverse human and ecological benefits, realisation repeatedly reinforces existing (ecological) injustices. Not only does this materialise through uneven benefit distribution, but also from decision-making processes that typically disproportionately favour already privileged groups at the expense of marginalised human and ecological communities. Concerningly, limited studies to date have explored the extent of such challenges, including for the city of Coventry, UK. This study aims to bridge this research gap, utilising semi-structured interviews with urban practitioners (n = 14) working on GI in Coventry. Reflexive Thematic Analysis includes theoretical engagement with Ecological Justice, while remaining open to alternative understandings and interpretations of justice in the GI context. This provides greater analytical flexibility to engage with diverse perspectives of practitioners and increases compatibility with existing socio-political contexts.

Thematic analysis identifies two themes that collectively raise how existing GI realisation leaves certain human and ecological communities overlooked and undervalued.

Firstly, contested and contradictory community roles highlights how structural challenges limit depth of authentic, inclusive community engagement. This includes short-term development agendas being prioritised over long-term community need, and a growing reliance on communities during maintenance stages potentially exacerbating existing injustices as engagement becomes tied to capacity. The second theme, ecology as silenced and sometimes catalysers, raises how anthropocentric agendas result in ecology being instrumentalised and largely overlooked throughout realisation. While promising that there are cases where ecology can disrupt dominant approaches, there remains a need to question the underlying value judgements that shape this. Three recommendations are outlined for transitions towards ecologically just GI. Community-derived knowledge should be integrated into decision-making utilising participatory methodologies to co-produce GI. Such involvement should occur from the outset, countering current late-stage reliance during maintenance. Finally, ecological visibility should increase in policy and practice, aided by enhanced enforcement.

Urban food-growing is increasingly promoted as a strategy for enhancing urban resilience, biodiversity, and food security, yet its integration into everyday public green spaces remains limited and underexamined. In the UK, food-growing is still largely framed as temporary, community-led, or peripheral, rather than as a legitimate component of mainstream urban green infrastructure. This research offers an original contribution by positioning public perception as a central determinant of the acceptability and feasibility of integrating food-growing interventions into public green spaces, using Sheffield as a case study.

Employing a mixed-methods approach, the study combines a structured questionnaire with closed- and open-ended questions alongside a visual survey component. Data were collected from 412 Sheffield residents aged 18 and above through online and in-person surveys. Statistical analyses, including correlation and regression models, were used to examine how sustainability awareness, land-use priorities, intervention preferences, and policy constraints shape public attitudes, while qualitative responses provided interpretive depth.

The findings reveal strong public support for integrating food-growing into public green spaces, with over 90% of respondents recognising its value for placemaking, social cohesion, education, and community wellbeing. Crucially, the study demonstrates that social and civic benefits outweigh environmental motivations in shaping public acceptance, challenging dominant policy narratives that frame urban food-growing primarily through ecological or productivity-based lenses. Nonetheless, concerns related to governance, safety, land-use conflict, and long-term economic viability highlight persistent structural barriers to implementation.

By empirically linking public perception, policy feasibility, and landscape design, this study advances urban food scholarship and reframes food-growing as a permanent, publicly negotiated landscape intervention rather than a marginal or temporary practice. The research offers actionable insights for urban geographers, planners, landscape architects, and policymakers seeking to develop inclusive, resilient, and publicly supported urban food-growing strategies.

The petroleum industry is key in the global economy; it also represents a significant aspect of Nigeria’s economic productivity and development plans. More than 70% of energy demand in the country is derived from petroleum product and its derivatives. The national distribution of petroleum products has since tilted towards road tanker trucks, pipelines and waterways transport in Nigeria since the abandonment of rail lines in the 1990’s which was a more sustainable distributing means and has witnessed environmental and social burden such as oil spillages incidences, property loss, air pollution, and associated emissions. In ensuring a sustainable distribution of petroleum products, it involves a strategic commitment to uninterrupted consumer supply at competitive prices and uncompromised quality, achieved through logistics processes characterized by efficiency, flexibility, adaptability, and resilience.

This research is a study of the downstream marketing and distribution operation of refined petroleum products in Nigeria and to answer specific research questions, a comprehensive review of related literatures based on the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses was carried out using web of science, google scholar and science direct database, and after the screening process a total of 81 papers was considered for eligibility for a full text review. To get a balanced perception that meets the objectives of this study, a structured study survey and expert interviews with stakeholders in the downstream oil supply chain will be carried out and an integrative framework using a systems thinking approach will be developed.

This study situates logistics agility and resilience within broader sustainable development goals (SDGs) and nationally determined contributions, particularly SDG 7 (affordable and clean energy), SDG 9 (Industry, innovation, and infrastructure) and emission reduction. It recommends effective supply chain coordination among stakeholders, policy reforms, infrastructure revitalization, and digital integration to optimize Nigeria’s petroleum logistics chain.

The construction industry continues to face persistent challenges including low productivity growth, labour shortages, high accident rates, cost overruns, and increasing sustainability pressures. While robotics has been proposed as a means to address these issues, current deployments are often fragmented, task-specific, and poorly integrated with broader construction workflows. This limits their effectiveness in the dynamic and uncertain environments typical of real construction sites.

This paper presents a structured review and synthesis of robotics in construction through a role-based functional perspective. Rather than categorising systems by individual tasks or robotic platforms, the paper defines roles as the functional responsibilities robotic systems fulfil within construction processes. This framing aligns more closely with construction practice and enables comparison across heterogeneous robotic technologies.

Five primary roles are identified and analysed: digital construction and precision building, production and heavy operations, inspection and monitoring, human–robot collaboration, and sustainable and resilient construction. For each role, the paper examines representative applications, enabling technologies, and common limitations, particularly with respect to adaptability, environmental variability, and system integration. Enabling technologies such as BIM, digital twins, computer vision, artificial intelligence, and robotic localisation and mapping are discussed as cross-cutting requirements for effective deployment.

The paper’s contribution lies in providing a unifying conceptual structure that clarifies how robotics currently supports construction activities, where limitations persist, and how future research can move beyond isolated solutions. The analysis motivates further investigation into coordinated multi-robot systems, cooperative perception, and adaptive task allocation, which form the foundation of the author’s ongoing doctoral research.

Seagrass meadows are known to harbour a highly diverse and productive ecosystem. As they function as foundational species, they can transform low-biodiverse sand flat regions into highly diverse ecosystems, encouraging the occupation of other marine plants and animals. However, anthropogenic pressures are intensifying the stress seagrass are under and decreasing the overall resilience of the plant. One of the major contributing factors to the loss of seagrass ranges is the influx of nitrogen and phosphate, which triggers coastal blooms. Under ambient conditions, seagrass and epiphytes have a symbiotic relationship, and epiphytes can often increase the abundance of grazing invertebrates. However, during algal bloom events caused by coastal eutrophication, fast-growing opportunistic epiphytes belonging to the families Phaeophyceae, Rhodophyta, and Chlorophyta occupy seagrass strands, reducing light availability, decreasing oxygen in the water column, and increasing physical drag.

This investigation aimed to measure the biodiversity of the meadow with emphasis on epiphyte and invertebrate interactions. This was done by identifying species on seagrass strands of Zostera marina and Zostera noltei at eight quadrat points within an intertidal seagrass meadow. It was found that the species richness was 14 and the Shannon diversity was 2.18. The most abundant species on the seagrass strands were Phaeophyceae, Rhodophyta, and Chlorophyta, and the most common invertebrates were Amphipods, Isopods, Gastropods, and species 2 (unclassified egg), with no significant difference observed between quadrats. Biota was found within. The average percentage cover of seagrass strands was 18.3%; however, significant differences were observed between quadrat samples. As this is the first time this meadow has been sampled and had the biodiversity investigated, this study should serve as a guide for future management of the site.

Seagrasses are indicators of healthy marine ecosystems, providing an array of ecosystem services including coastal protection, nutrient cycling, improving water quality and carbon sequestration. They are also essential nurseries and habitats for the juvenile and larval stages of commercially and recreationally important fish species. Despite their ecological and socio-economic importance, one third of global seagrass beds have declined in the past 100 years, largely due to human activity. Threats to seagrass ecosystems include fluctuations in temperature and salinity, however, it is eutrophication that is widely considered the primary driver of seagrass loss. This process occurs due to the nutrient enrichment of marine systems, often due to anthropogenic activity including urban and agricultural development. In particular, ammonia toxicity often arises from nutrient enrichment and has serious deleterious effects on seagrass photosynthesis and pH regulation.

This study investigates how nutrient enrichment and ammonia toxicity affects photosynthetic efficiency and leaf structure of eelgrass (Zostera marina). Plants were collected from two sites; one unpolluted and minimally impacted by human activity (Porthdinllaen on the Llŷn Peninsula, Wales), and the other identified as having poor water and sediment quality (Walney Channel, Barrow-in-Furness, England). Leaf anatomy and stable carbon-isotope analysis will be used to quantify plant stress and assess overall seagrass health in these contrasting environments. This study aims to determine stress responses at the leaf-level to inform conservation strategies in mitigating seagrass decline due to eutrophication.

Seagrass meadows are significant global carbon sinks that can mitigate the impacts of climate change, yet the fundamental relationship with microbial communities in their sediments remains somewhat unclear. This research investigates the effect seagrass (Zostera marina) has on sediment particle size, carbon concentrations and bacterial microbiome in comparison to adjacent bare sediments, in a temperate seagrass meadow in Porthdinllaen, North Wales. Contradictory to existing literature, bare sediments had slightly higher organic carbon concentrations (mean= 2.13%) than in seagrass sediments (mean= 2.04%). Similarly, particle size analysis revealed a lower median particle size within bare sediments (112.95µm) compared to seagrass sediments (129.57µm).

There was also significantly more variation across seagrass sediments within both findings compared to the greater homogeneity in bare sediments. Microbiome analysis conducted via DNA extraction from sediment samples revealed distinct phylogenetic communities in seagrass and bare sediments, with minimal overlap. Seagrass sediments showed higher variability but lower overall alpha diversity than bare sediments, signifying the role of fewer, specialised bacteria in seagrasses versus abundant, generalist bacteria in bare sediments. This research shows the complexity of microbial interactions in sediments, and how this can alter not only with vegetation presence and absence, but also in relation to specific local conditions (including nutrient enrichment and hydrodynamics).

Seagrasses are the only flowering plants that grow in marine environments, and they provide a multitude of ecosystem services, including carbon sequestration, sediment stabilisation, increased biodiversity, and nutrient filtering. But despite these benefits, seagrass has been declining worldwide due to anthropogenic causes such as pollution and urbanisation. Zostera noltei is a species of seagrass native to the UK and has suffered a similar decline as around 44 – 85% of seagrass in the UK has been lost since the 1920s. This project aims to identify suitable sites for the restoration of Z. noltei in the Upper Mersey Estuary as Natural England has identified potentially suitable seagrass habitat in this area. While the Mersey estuary has historically been extremely polluted, cleanup efforts over the last 40 years have made huge improvements to the environmental conditions and water quality to the point that it is healthy enough to support a seagrass population.

Introduction of Z. noltei to the Mersey would benefit the estuary through increased biodiversity, water quality monitoring, and nutrient filtering. To achieve this goal, a systematic review of the environmental parameters that Z. noltei requires to thrive will be conducted. Secondly, environmental conditions of the Mersey estuary will be collected from both existing sources and through field work where necessary. These two datasets will be used to create a map in Geographic Information Systems (GIS) that will compare the conditions in the Mersey estuary to the conditions that are best for Z. noltei survival and used to determine where the most optimal sites for a Z. noltei planting project would be.

Seagrass meadows play a crucial role in carbon sequestration and long-term carbon storage, and recent years have seen a significant increase in studies regarding the carbon stocks within seagrass sediment. A key methodological challenge arises from the varying “depth of refusal” (the maximum depth coring can reach) among different seagrass meadows due to sedimentation differences. This variability usually causes surveyors to use different core samplers—potentially introducing methodological bias into the results like organic carbon density.

To examine and minimise this potential inconsistency, we specifically focused on comparing the differences in organic carbon density acquired using two standard methods: a Russian corer and a piston corer.

We selected six UK seagrass meadow study sites (Spurn, Copperas Bay, Goldhanger, Islay, Porthdinllaen, and Carlingford) to conduct sediment coring, adapting the tool based on the “depth of refusal” limitation (Russian corer for > 20 cm depth; piston corer for < 20 cm depth). Among the sites, Spurn was used as a critical case study for this direct methodological comparison. Subsequently, the sediment cores were analysed for the vertical changes in sediment bulk density, organic carbon content (%), and organic carbon density using a detailed layering strategy (1-cm intervals for the top 6 cm, and 2-cm intervals from 6 cm to 50 cm or refusal depth).

Russian coring is suitable for collecting the sediment cores of intertidal seagrass meadows, while in the sites with a shallow depth of refusal, like Porthdinllaen and Carlingford, piston coring may be a proper replacement to collect sediment cores.

Overall, this examination, particularly using the Spurn case study, provides a direct reference for further studies, offering a standardised approach for the sediment-core collections from different seagrass habitat types, thereby minimising methodological inconsistency between different coring techniques in future seagrass carbon stock assessments.