Radiation-induced transient faults pose a growing challenge for safety-critical embedded systems, yet traditional radiation testing and large-scale statistical fault injection (SFI) remain costly and impractical during early design stages. This paper presents a predictive approach for early reliability assessment that replaces handcrafted feature engineering with automatically learned vector representations of source code and execution traces. We derive multiple embeddings for traces and source code, and use them as inputs to a family of regression models, including ensemble methods and linear baselines, to build predictive models for reliability. Experimental evaluation shows that embedding-based models outperform prior approaches, reducing the mean absolute percentage error (MAPE) from 6.24% to 2.14% for correct executions (unACE), from 20.95% to 10.40% for Hangs, and from 49.09% to 37.69% for silent data corruptions (SDC) after excluding benchmarks with SDC below 1%. These results show that source code and trace embeddings can serve as effective estimators for expensive fault injection campaigns, enabling early-stage reliability assessment in radiation-exposed embedded systems without requiring any manual feature engineering. This capability provides a practical foundation for supporting design-space exploration during early development phases.

Humans have an inherent connection to nature, and exposure to natural elements has been shown to reduce stress, improve mood, and support cognitive performance, forming the basis of biophilic design in the built environment. However, existing biophilic design guidance remains largely conceptual and offers limited evidence-based direction on how design properties should be applied. This scoping review addresses this gap by systematically mapping and synthesizing empirical evidence on indoor biophilic design interventions and their properties. Following PRISMA-ScR guidelines, 136 studies published between 2000 and 2025 were reviewed across seven intervention types, including green walls, indoor plants, window views, natural light, natural materials, water features, and nature-inspired visual references. Cross-category analyses identified design properties most consistently associated with restorative outcomes and human cognitive and physiological responses. The findings highlight the importance of moderate greenery levels, high-visibility placement, multi-sensory integration, and the enhanced restorative effects of combining multiple interventions. Contextual factors such as exposure duration and user characteristics were found to influence effectiveness. Based on these findings, the study introduces the Biophilic Intensity Matrix (BIMx), a matrix-based decision-support framework that supports early-stage design by helping designers select biophilic intervention types and compare their relative scale and intensity ranges according to exposure duration.

Abstract The presented study proposes a hybrid multi-resolution simulation framework for predicting and validating the performance of UAV-class micro gas turbine engines in the 4–6 kN thrust range. The approach integrates 1D thermodynamic cycle analysis using the Gas Turbine Simulation Program (GSP), intermediate system-level validation through Flownex, and high-fidelity 2D CFD simulations in ANSYS Fluent for component-level analysis. Parametric studies were conducted for pressure ratios between 1.3 and 1.5 and Mach numbers from 0.1 to 0.6, at a design thrust of 4.56 kN and turbine inlet temperature of 1053 K. CFD results predicted nozzle exit velocities exceeding 750 m/s with minimal flow separation. Temperature predictions across tools showed strong agreement within 5 %, while pressure and mass-flow deviations were observed in turbulence-dominated regions. The framework demonstrates a balanced trade-off between computational efficiency and fidelity, offering a reliable methodology for early-stage design and validation of compact UAV propulsion systems.

The construction sector faces increasing pressure to curtail embodied carbon associated with cement and concrete. This study quantifies the baseline embodied CO? emissions of a multi-storey commercial building and evaluates the reduction achieved through structural optimization. A Building Information Modeling (BIM) workflow provided accurate quantity take-off, which was coupled with Life Cycle Assessment (LCA) across stages. In the baseline design, total concrete volume was 510.038 m³ with associated emissions of 166,100 kg CO? (including transport). The optimized design reduced concrete volume to 435.738 m³ and total emissions to 141,904 kg CO?, a 14.57% reduction, primarily from resizing slabs, beams, and columns while maintaining code compliance and structural integrity. The work demonstrates a practical, replicable pathway for integrating BIM–LCA in early-stage design to deliver tangible carbon savings in resource-constrained contexts.

The development of modern vehicles faces an increase in complexity, as well as a need for shorter development cycles and a seamless cross-domain integration. In order to meet these challenges, a graph-based design language which formalizes and automates engineering workflows is presented and applied in a design case study to a Formula Student race car suspension system. The proposed method uses an ontology-based vocabulary definition and executable model transformations to compile design knowledge into a central and consistent design graph. This graph enables the automatic generation of consistent 3D CAD models, domain-specific simulations and suspension kinematic analyses, replacing manual and error-prone tool and data handover processes. The design language captures both the structural and dynamic behavior of the suspension, supports variant exploration and allows for integrated validation, such as 3D collision detection. The study illustrates how graph-based design languages can serve as ‘digital DNA’ for knowledge-based product development, offering a scalable, reusable platform for engineering automation. This approach enhances the digital consistency of data, the digital continuity of processes and the digital interoperability of tools across all relevant engineering disciplines in order to support the validation of early-stage designs and the optimization of complex systems.

Environmental Life Cycle Assessment (eLCA) has become an integral component of building certification systems and regulatory frameworks. However, existing approaches lack robust methodologies for evaluating the environmental implications of socially oriented design intents during early design stages. This gap limits architects and other early-phase stakeholders’ ability to balance social and environmental sustainability effectively when making preliminary decisions. To address this challenge, the present study introduces a systematic methodology for eliciting socially oriented design intents from early design phases and quantifying their environmental impact through eLCA. The proposed framework categorises these design intents into three types: Addition, Removal, and Replacement, based on their impact on building design. Three case studies are presented to illustrate the application of this methodology across these categories. The findings demonstrate that the inclusion of socially oriented design intents can significantly influence a building’s total CO2-equivalent emissions, either positively or negatively. By enabling early-stage assessment of such impacts, this research provides a foundation for more informed decision-making, supporting the integration of social and environmental objectives in building renovation projects.

Participatory Design is one of the traditional human-centred approaches commonly used during the early stages of the design process. Currently, Large Language Models and Generative Artificial Intelligence (GenAI) are increasingly being adopted in the early stage of design, particularly for requirements analysis and conceptual design. To harness the respective strengths of Participatory Design and GenAI, it is essential to approach the problem through the lens of multimodal human–machine perception. In this paper, we propose a new multimodal GenAI-Driven Participatory Design approach and develop the Multi-PDAI platform, which is validated in the context of wheeled humanoid robot design. A crossover experiment (N designers = 8, N users = 16) was conducted with evaluation data on the generated images from design industry experts (N = 10). Platform usability was assessed using the SUS scale. In addition, users’ UEQ data and designers’ IMI data were analyzed, with systematic comparisons of initial and final prompt templates, design completion criteria, and time–process metrics. The results show that the images generated by the Multi-PDAI platform are more innovative and aesthetically appealing, and the platform demonstrates better usability compared to Stable Diffusion (M = 72.5, Sig. = 0.043). These findings further support the hypothesis that participatory design and GenAI are mutually beneficial, and that multimodality in GenAI is essential.

Despite European and international regulatory frameworks promoting circular economy principles, sustainability in the furniture sector is still challenged by the limited access to reliable information about circular materials for designers, manufacturers, and waste managers in the Made-in-Italy furniture ecosystem. This research develops a digital infrastructure to address these information gaps through mixed methodology, combining desk research on regulatory frameworks and existing platforms; field research involving stakeholder mapping and interviews with designers, manufacturers, and waste managers; and the experimental development of AI-enhanced digital tools. The result integrates a web-based platform for circular materials with a CAD plugin supporting real-time sustainability assessment. As AI-assisted data entry showed a reduced form completion time while maintaining accuracy through human verification, testing also revealed how the system effectively bridges knowledge gaps between stakeholders operating in currently siloed value chains. The platform is a critical step in enabling designers to incorporate circular materials during the early design stages, while providing manufacturers access to verified punctual sustainability data compliant with mandatory Green Public Procurement criteria. Beyond the development of an innovative digital tool, the study outlines a corresponding operational model as a practical framework for strengthening the transition toward a circular economy in the Italian furniture industry.

This study reviews the current applications of generative artificial intelligence (GenAI) in architectural design education using the PRISMA framework. It compares these applications across the different design stages, namely the pre-design, concept generation, design development, and design production, to identify the current state of evidence and conceptual discussions reported in the literature. The study also discusses the associated opportunities and challenges in this regard. The findings indicate that there is a growing interest in integrating GenAI into architectural design education, especially in the early design stages. However, one of the most significant gaps in this regard lies in the lack of empirical evidence on the long-term impacts of GenAI on students’ critical thinking and problem-solving skills. Future research is needed to explore the integration of GenAI throughout the entire design process, including design development and refinement. There is also a need to incorporate the relevant ethical guidelines for AI-generated content into academic quality assurance systems and to strengthen institutional preparedness through targeted training and policy development.

ABSTRACT Digitally mediated design mediums have revolutionized product design, but their cognitive impact during early concept design stages remains unclear. This study employs cognitive load theory (CLT) and constraint-driven cognition to examine how traditional (TD) and digitally mediated (DM) mediums influence cognitive load, problem-solving, and design outcomes. A quasi-experimental study with 16 design students, divided into TD and DM groups, used two distinct design tasks. Results revealed three key findings. First, the DM medium imposed a significantly higher extraneous cognitive load due to attentional fragmentation and interface management, consuming working memory resources critical for creative synthesis. Second, a fundamental strategic divergence emerged: the TD group engaged in problem-driven cognition through material constraints, yielding higher conceptual novelty (63% vs. 25%) and five times more sustainability considerations. The DM group used solution-driven strategies, leading to more derived outcomes. Third, the cognitive impact was task-dependent; digital tools reduced intrinsic load for well-defined mechanical tasks but offered no advantage for open-ended aesthetic tasks. This study suggests design mediums function as active cognitive environments, not neutral tools. A reevaluation of design education and practice is essential, promoting digital metacognition, retaining tactile skills, and developing hybrid processes that leverage the distinct cognitive benefits of each medium.

Pharmaceutical pollution in aquatic environments poses a significant ecological threat due to the accumulation of bioactive compounds from human and veterinary sources. In support of the EU Green Deal's Chemicals Strategy for Sustainability, this study presents a computational framework for predicting two key environmental risk indicators in fish: bioconcentration and ecotoxicity. Bioconcentration, quantified by the bioconcentration factor (BCF), reflects a chemical's tendency to accumulate in organisms, while ecotoxicity is assessed via the median lethal concentration (LC50) over defined exposure periods. We developed two high-performing machine learning (ML) models, achieving ROC AUC scores of 94.60% for bioconcentration and 96.06% for ecotoxicity, validated across both internal and external data sets. To expand the scope of risk evaluation, we incorporated metabolite prediction using the SyGMa tool, selected after benchmarking multiple alternatives. This enables the assessment of both parent compounds and their potentially toxic metabolites. Model interpretability was enhanced through molecular fingerprint analysis, which identified structural features associated with toxicity and accumulation, informing the early stages of drug design. To support practical implementation, we introduced G.AI.A (https://gaiatox.eu/), an intuitive web platform that allows users to input Simplified Molecular Input Line Entry System (SMILES) strings for rapid prediction of environmental risk end points. The application domain of G.AI.A lies in predictive toxicology, enabling researchers and regulatory bodies to assess the toxicological profiles of small organic compounds, excluding those containing heavy metals, by analyzing their chemical structures. The platform supports batch processing and offers interactive visualizations, facilitating compound screening and early stage environmental risk assessment. By integrating predictive modeling with interpretability and usability, our framework advances green-by-design pharmaceutical development and contributes to sustainable chemical management.

Under the dual pressures of global climate change and accelerating urbanization, landscape design has been tasked with the critical mission of enhancing urban environmental resilience and ecological livability. However, conventional design practices often struggle to efficiently integrate complex sustainability norms with aesthetic creativity, leading to a disconnect between form and function. To address this issue, this study proposes and validates an AI-enabled sustainability decision-support framework. The framework is based on a “Generative-Critical” multi-agent workflow that enables “Self-Correcting” iterative optimization of design schemes through a built-in expert knowledge base and a quantitative scorecard. The framework’s effectiveness was validated through a cultural park case study and a blind evaluation by 10 experts. It guided a design from an initial concept with only aesthetic forms and lacking effective stormwater management, to an ecologically integrated scheme that strategically incorporated bioretention ponds at key nodes and converted hard plazas into permeable pavements. This transformation significantly elevated the scheme’s sustainability score from 59.3 to 88.0 (p < 0.001), while the framework itself achieved a high system usability scale (SUS) score of 85.5. These results confirm that the proposed “Generative-Critical” mechanism can effectively guide AIGC to adhere to ecological-technical norms and constraints while pursuing aesthetic innovation, thereby achieving a scientific integration of aesthetic form and ecological function at the early conceptual design stage. This study offers a scalable methodology for AI-assisted sustainable design and provides a novel intelligent tool for creating resilient urban landscapes that possess both environmental performance and aesthetic value.

The design of aircraft components is a complex process that must simultaneously account for environmental impact, manufacturability, cost and structural performance to meet modern regulatory requirements and sustainability objectives. When these factors are integrated from the early design stages, the approach transcends traditional eco-design and becomes a genuinely sustainability-oriented design methodology. This study proposes a sustainability-driven design framework for aircraft components and demonstrates its application to a fuselage panel consisting of a curved skin, four frames, seven stringers, and twenty-four clips. The design variables investigated include the material selection, joining methods, and subcomponent thicknesses. The design space is constructed through a combinatorial generation process coupled with compatibility and feasibility constraints. Sustainability criteria are evaluated using a combination of parametric Life Cycle Assessment (LCA) and Life Cycle Costing (LCC) regression models, parametric Finite Element Analysis (FEA), and Random Forest surrogate modeling trained on a stratified set of simulation results. Two methodological pathways are introduced: 1. Cluster-based optimization, involving customized clustering followed by multi-criteria decision-making (MCDM) within each cluster. 2. Global optimization, performed across the full decision matrix using Pareto front analysis and MCDM techniques. A stability analysis of five objective-weighting methods and four normalization techniques is conducted to identify the most robust methodological configuration. The results—based on a full cradle-to-grave assessment that includes the use phase over a 30-year A319 aircraft operational lifetime—show that the thermoplastic CFRP panel joined by welding emerges as the most sustainable design alternative.

Accurate prediction of subway-induced environmental vibrations for unopened lines remains a significant challenge due to the difficulty in determining appropriate excitation inputs. To address this issue, this study proposes an excitation modification method based on field measurements and numerical simulations. First, field measurements were conducted on a subway line in Shanghai to analyze vibration propagation characteristics and validate a two-dimensional finite element model (FEM). Subsequently, based on the validated model, frequency-band excitation modification formulas were derived. Distinct from existing empirical approaches that often rely on simple statistical scaling, the proposed method utilizes parametric numerical analyses to determine frequency-dependent correction coefficients for four key parameters: tunnel burial depth, tunnel diameter, soil properties, and train speed. The reliability of the proposed method was verified through theoretical analysis and an engineering application. The results demonstrate that the proposed method improves prediction accuracy for tunnels in similar soft soil regions, reducing the prediction error from 10.1% to 5.2% in the engineering case study. Furthermore, parametric sensitivity analysis reveals that ground vibration levels generally decrease with increases in burial depth, tunnel diameter, and soil stiffness, while exhibiting an increase with train speed. This study improves the reliability of vibration prediction in the absence of direct measurements and provides a practical tool for early-stage design and vibration mitigation for unopened lines.

Achieving optimal daylighting is a cornerstone of sustainable architectural design, impacting energy efficiency and occupant well-being. Fast and accurate prediction during the conceptual phase is crucial but challenging. While physics-based simulations are accurate but slow, existing machine learning methods often rely on restrictive parametric inputs, limiting their application across free-form designs. This study presents a novel, geometry-agnostic framework that uses only building facade unfolding diagrams as input to a Generative Adversarial Network (GAN). Our core innovation is a 2D representation that preserves 3D facade geometry and orientation by “unfolding” it onto the floor plan, eliminating the need for predefined parameters or intermediate features during prediction. A Pix2pixHD model was trained, validated, and tested on a total of 720 paired diagram-simulation images (split 80:10:10). The model achieves high-fidelity visual predictions, with a mean Structural Similarity Index (SSIM) of 0.93 against RADIANCE/Daysim benchmarks. When accounting for the practical time of diagram drafting, the complete workflow offers a speedup of approximately 1.5 to 52 times compared to conventional simulation. This work provides architects with an intuitive, low-threshold tool for rapid daylight performance feedback in early-stage design exploration.

BackgroundPeople with a learning disability are frequently excluded from clinical trials, with around two thirds of trials either directly or indirectly excluding this group. This contributes to the shocking health inequalities they experience, with people with a learning disability having higher rates of long-term health conditions and dying on average 20 years younger than the general population. Improving inclusion of under-served groups in trials is a priority area for research funders and regulators. A UK-wide collaboration, ‘No Research About Us, Without Us’, was formed to explore and address the barriers to engaging and involving people with a learning disability in research. The project consisted of a number of intersecting work streams. This paper reports the findings from Working Group 3 which aimed to produce practical examples about how a trial could be redesigned to ensure it is more inclusive of people with a learning disability.MethodsThe redesign process consisted of three steps: (1) identifying an appropriate trial using predefined criteria, (2) selecting a tool to systematically review the trial, and (3) identifying barriers to inclusion of people with a learning disability and proposing alternative design approaches that could have widened access to the trial.ResultsFollowing review of a funder’s portfolio, we selected a platform trial (PANORAMIC) which had sought to include people with a learning disability as a high-risk group for COVID-19 and yet had only made up 0.01% of those recruited. Using the INCLUDE Impaired Capacity to Consent Framework, our co-produced analysis identified practical strategies that could have ensured greater inclusion of people with a learning disability. This included involving people with a learning disability at the earliest design stage, revisiting eligibility criteria, making reasonable adjustments (e.g. high-quality easy read versions of all documents), and simplifying overly complex study processes.ConclusionTo achieve better health equity and improve the quality of clinical trials, researchers must pay greater attention to accessible study design and ensure appropriate accommodations are in place to enable inclusion of people with a learning disability. We outline some practical strategies that can inform the design and conduct of future trials to improve inclusion.

This paper presents a unified aerodynamic design and optimization framework for morphing Blended-Wing-Body (BWB) Unmanned Aerial Vehicles (UAVs) operating in subsonic and near-transonic regimes. The proposed framework integrates parametric CAD modeling, Computational Fluid Dynamics (CFD), and surrogate-based optimization using Response Surface Methodology (RSM) to establish a generalized approach for geometry-driven aerodynamic design under multi-Mach conditions. The study integrates classical aerodynamic principles with modern surrogate-based optimization to show that adaptive morphing geometries can maintain efficiency across varied flight conditions, establishing a scalable and physically grounded framework that advances real-time, high-performance aerodynamic adaptation for next-generation BWB UAVs. The methodology formulates the optimization problem as drag minimization under constant lift and wetted-area constraints, enabling systematic sensitivity analysis of key geometric parameters, including sweep, taper, and twist across varying flow regimes. Theoretical trends are established, showing that geometric twist and taper dominate lift variations at low Mach numbers, whereas sweep angle becomes increasingly significant as compressibility effects intensify. To validate the framework, a representative BWB UAV was optimized at Mach 0.2, 0.4, and 0.8 using a parametric ANSYS Workbench environment. Results demonstrated up to a 56% improvement in lift-to-drag ratio relative to an equivalent conventional UAV and confirmed the theoretical predictions regarding the Mach-dependent aerodynamic sensitivities. The framework provides a reusable foundation for conceptual design and optimization of morphing aircraft, offering practical guidelines for multi-regime performance enhancement and early-stage design integration.

ABSTRACT Sustainable architecture faces significant challenges, particularly during the early design stages where critical decisions often lack sufficient detail for traditional environmental analysis. This paper presents the first development and use of an artificial intelligence-based tool designed to predict embodied carbon emissions from high-level natural human language descriptions of buildings. The new approach combines a Histogram-based Gradient Boosting regression model with a multi-step system of Natural Language Processing techniques to convert complex, unstructured text into structured features and quantities suitable for predictive modelling. The work rests on a foundation of 150,000 new synthetic training samples, generated by systematically randomising building specifications. Evaluation of the method’s performance was based on four strands: extraction sensitivity, relative accuracy, linguistic robustness, and usability. In tests of extraction sensitivity, the method successfully identified core structural and external elements over 80% of the time. Relative accuracy assessments with seven real-world buildings revealed a Spearman’s rank correlation of 0.71, confirming the system’s ability to identify differences in carbon-intensity. Linguistic robustness was proven by describing identical buildings in multiple ways, with predicted values differing by only 10%. A user study of 43 industry professionals produced a System Usability Scale score of 84.74. This reflects a strong acceptance of the method and the potential for its integration into existing workflows, emphasising its promise as an approach for advancing architectural practice. Collectively, these outcomes highlight the success of this new approach for embodied-carbon assessment and underscore the potential of AI-enabled insight in sustainable design.

The early stages of structural design increasingly make use of computational tools that support rapid exploration, performance-informed decision-making, and closer interaction between design and engineering. This systematic mapping study examines how Algorithm-Aided Design (AAD) and the Finite Element Method (FEM) are applied and combined in conceptual design workflows. Based on a structured search across three academic databases and a coding scheme applied to 87 publications, the literature is mapped according to algorithmic strategies, FEM applications, element types, disciplinary domains, and levels of integration. The results show that algorithmic and predictive approaches are reported with increasing frequency after 2020, alongside growing use of surrogate models and optimisation routines. Linear-elastic analyses and shell- or beam-based models are frequently reported, particularly in civil engineering contexts, while nonlinear, dynamic, and solid-element analyses appear more prominently in mechanical domains. More tightly coupled AAD–FEM workflows become increasingly visible after 2021, reflecting a growing interest in real-time or near-real-time simulation feedback during early design exploration. At the same time, the literature highlights persistent challenges related to computational cost, fragmented toolchains, limited interoperability, and the relatively limited use of multiscale or advanced material models in conceptual design. Taken together, the findings suggest that continued progress toward more integrated AAD–FEM workflows is closely tied to advances in computational efficiency, improved data exchange and interoperability, and the development of more accessible design–analysis environments across disciplinary boundaries.

In order to comply with the principle of sustainable development in product design, in addition to the mechanical properties of products, recycling properties should also be taken into account at the early stages of design. This paper explores the interplay between mechanical and recycling properties in product design in order to achieve a compromise between these design aspects. The research included typical metrics used to evaluate a product for its mechanical and recycling properties. The tests were carried out on a lap connection made in four variants: as a two-bolt, three-bolt, two-rivet and three-rivet connection. It was demonstrated that the stiffness of bolted connections is significantly lower compared to equivalent riveted connections. On the other hand, using three rivets instead of two in a connection yields better results in terms of load-bearing capacity compared to a similar increase in the number of fasteners in a bolted connection. The results demonstrate the impact of material structure of components and dismantling operations on the financial performance of the recycling process in relation to the assessment of recycling aspects in product design.