Development of an energy prediction framework for early-stage design in large-scale apartment complexes combining physical modeling with Gaussian process regression

The reactor vessel auxiliary cooling system (RVACS) of the PRISM (Power Reactor Innovative Small Module) provides a passive means for decay heat removal, primarily through radiative and convective heat transfer. We assess experimental scalability using a reduced-order one-dimensional (1D) nodal heat transfer model and validate it against higher-fidelity two-dimensional (2D) and three-dimensional (3D) simulations. The models solve heat flux by balancing radiative, conductive, and convective heat transfer along the reactor vessel (RV) wall, guard vessel wall, air flow in the riser, and duct wall under steady-state conditions representative of post-scram or reactor shutdown operations. Dimensional analysis and similarity laws, mainly utilizing Richardson number scaling, are applied to ensure consistency between the buoyancy-driven flow behavior between the full-height reactor and that of the experimental scaled-down systems. Using the RV inner wall temperature of 922 K as the primary metric, the 1D model underpredicts the 2D and 3D simulation heat flux values for full-reactor height by 6.05% and 17.36%, respectively. The deviation increases to 17.59% for 2D and 25.29% for 3D at 1/6 height scaling, which is attributable to the inlet and outlet effects and the flow structures of the 2D and 3D models. Both the 2D and 3D scaled models follow the Richardson number scaling trends, with heat flux relative errors reaching maximum values of 14.00% and 10.61% at 1/6 scaling, confirming the strong agreement with the scaling temperature value calculated from the 1D model. The findings demonstrate the ability for a reduced-order 1D nodal framework to sufficiently model the early-stage design and experimental scaling of the RVACS for configurations dominated by radiative heat transfer, shown for geometric scaling of 1/6 or greater, providing an efficient and cost-effective tool for assessing passive decay heat removal performance with a reduction in computational costs. The approach assumes fully developed axial behavior, as it does not explicitly resolve the inlet and outlet effects, which become increasingly pronounced at smaller scale ratios. The methodology is broadly applicable to other passive decay heat removal systems, such as reactor cavity cooling systems and other air-cooled safety designs governed by radiative and natural convection heat transfer.

ABSTRACT Emergency Department (ED) staff must rapidly adapt in response to Mental and Behavioral Health (MBH) patient needs. The purpose of this study was to compare two ED exam room design proposals (parallel and perpendicular bed positions) developed to address the need for flexibility and safety in the ED. Simulation-Based Evaluation (SBE) in Virtual Reality (VR) was conducted to evaluate ED staff perceptions of safety, effectiveness, and patient experience in the room at an early design stage. Twenty-seven ED staff members with experience managing MBH patients performed tasks in VR following a scripted scenario that consisted of a nurse caring for an MBH patient with an arm injury, verbalizing their feedback and experiences for both designs. Surveys and debriefing interviews were conducted to gather further insights. This study found that staff members evaluated the perpendicular bed position as safer for all users and afforded a more efficient workflow; Participants valued several features of the overall room design, including: (i) customizable room controls, (ii) sensory-friendly environment and (iii) flexible features, which allowed for a rapid transition from a medical to an MBH room. SBE confirmed that the perpendicular bed position was preferred by staff, and the findings guided future refinements.

Gearbox housing stiffness strongly influences radiated noise in electric drivetrains, particularly in the absence of engine masking. While high-fidelity vibro-acoustic simulations provide detailed insight, they are computationally demanding for early-stage design screening. This study investigates whether extremely compact spectral descriptors can encode stiffness-related information. The descriptors consist of five 1 kHz band-averaged sound pressure levels between 1 and 6 kHz. These band-averaged quantities are treated as compact spectral descriptors representing the acoustic response of each gearbox housing configuration. The analysis is based on a simulation-derived dataset of twelve spectra representing three ribbing configurations of a single gearbox housing geometry. A Random Forest classifier evaluated using leave-one-out cross-validation (LOOCV) achieved 0.75 accuracy. Confusion matrix analysis indicates clear separation of the flexible concept. Intermediate and rigid configurations show partial spectral overlap. Permutation testing suggests that the observed classification performance exceeds random chance, although uncertainty remains substantial due to the small dataset size. Feature-importance analysis identifies the 2–4 kHz region as the most stiffness-sensitive frequency range, supporting physical interpretations of mid-frequency structural–acoustic coupling. This exploratory study highlights both the potential and the statistical limits of minimal frequency-band descriptors for rapid NVH stiffness screening under small-sample conditions.

Background: Extracellular vesicles (EVs) are increasingly explored as nanocarriers in drug delivery; however, selecting an appropriate loading strategy for a given small-molecule cargo still relies largely on empirical, resource-intensive parallel screening within EV formulation workflows. Despite the widespread application of passive incubation, electroporation, saponin-mediated permeabilization, freeze-thaw cycling, and sonication, there is currently no mechanistically grounded, descriptor-informed framework that enables rational prioritization of loading methods during the early design stage of EV-based dosage forms, leading to inefficient trial-and-error experimentation. Methods: We assembled a chemically diverse dataset of 21 compounds with experimentally determined loading efficiencies across five EV loading methods and calculated seven mechanistically motivated physicochemical descriptors (LogP, molecular weight, aqueous solubility, hydrogen bond donors/acceptors, polar surface area, and formal charge) for each drug. Separate Elastic Net regression models were trained for each loading strategy. Model performance was evaluated using leave-one-out cross-validation, a predefined external validation set (n = 4), and 50 repeated random train-test splits. The analysis emphasized decision-level ranking of loading methods rather than the precise prediction of absolute efficiencies. The applicability domain was assessed via leverage analysis to define the supported chemical space for prospective implementation in EV-based formulation development. Results: As anticipated for biologically heterogeneous EV systems, continuous regression performance remained modest (LOOCV R2 = 0.06-0.41). In contrast, decision-level accuracy for identifying the experimentally optimal loading method was consistently high across validation schemes (internal: 76.5%; predefined external: 75%; repeated random validation: 80.5 ± 16.8%). Mechanical disruption methods (freeze-thaw and sonication) demonstrated comparatively greater predictive stability, while misclassification patterns suggested potential nonlinear behavior for highly polar, ionizable cargos. All compounds resided within the leverage-defined applicability domain, confirming adequate descriptor-space representation. Conclusions: This study establishes a mechanistically interpretable, descriptor-based decision-support framework capable of reliably prioritizing EV loading strategies for small-molecule cargos beyond empirical chance without altering standard protocols. By reframing the modeling objective from high-precision efficiency prediction to robust ranking of candidate methods, the approach offers a practical tool to triage between commonly used techniques, thereby reducing experimental burden in early-stage EV formulation development. The framework provides a quantitative basis for integrating molecular-descriptor-guided method selection into rational EV-based drug delivery design and can be expanded with membrane-specific descriptors and larger datasets.

With the advancement of technology nodes and the increasingly stringent requirements for stability, general yield analysis of customized circuits in the early stages of design has become a key bottleneck in manufacturing. In this article, we propose a multi-kernel sparse representation-based classification (MKSRC) method to enhance the efficiency and scalability of failure probability estimation by classifying tail samples. It employs class-balanced sampling to address data imbalance issues and utilizes multi-kernel features with adaptive kernel weights to enhance the accuracy and robustness of the classifier. Experimental results on 32-bit SRAM columns and analog circuits demonstrate that the proposed MKSRC method achieves higher classification accuracy and efficiency compared to other state-of-the-art methods, particularly in scenarios with limited training data. Compared to SOTA yield estimation methods, the MKSRC method achieves an average 2.57–3.29× improvement in both accuracy and efficiency, highlighting its ability to provide efficient and scalable yield analysis solutions for both SRAM and analog circuits.

Abstract The advancement of digital fabrication technologies and computational design has expanded the possibilities in architecture and construction. These technologies enable the creation of innovative structures and the exploration of non-standard materials, resulting in a new understanding of the design process and the development of advanced digital tools. Among these innovations, coreless filament winding, a robotic fabrication process, has facilitated new applications of fiber-polymer composites in architecture over the past decade. Coreless wound structures are characterized by a sequential and emergent nature, where the final geometry is shaped by material behavior. As a result, these structures are inherently difficult to predict, requiring constant feedback between digital models and physical prototypes due to the limited availability of fast, accessible simulation methods. This paper presents a simulation method developed to address this gap by supporting the early design stage of coreless wound fiber structures. The method incorporates relevant design and fabrication parameters while maintaining a necessary level of abstraction to ensure efficiency and accessibility. A case study was conducted to evaluate the simulation’s accuracy and examine the influence of different parameters on the final geometry. The study benchmarks a set of digitally simulated fiber specimens against their physical counterparts. The physical behavior of the fiber elements was captured using a monitoring method that scans the structure after each fiber segment is wound, enabling the observation of material behavior throughout the process. The results demonstrate the method’s efficacy, showing an average displacement deviation of -20 mm and high precision in fiber interaction, with 73% of specimens achieving 100% accuracy in generating intersection points. Furthermore, the study assesses the influence of design and fabrication parameters on fiber behavior, enabling informed decisions in early-stage design. By introducing an accessible and computationally efficient simulation method, the research aims to contribute to the field by providing efficient, early-stage feedback and an initial understanding of material behavior in coreless filament winding, while also identifying current limitations and directions for future research.

New, highly integrated, disruptive aircraft concepts are being devised to reduce aviation’s environmental footprint, but their performance is oftentimes challenging for the aircraft designer to assess. Furthermore, these novel aircraft often introduce new risks, such as noise, that cannot be addressed quickly by available methods. Overall, in the pursuit of more environmental friendly aircraft configurations and the lack of methods to design such aircraft, aircraft-level trade-offs between noise and performance are challenging. The present study aims to close this gap by using a machine learning-based approach for one unconventional aircraft to investigate usability in the early stages of aircraft design. Based on overflight noise measurements, noise models for this aircraft are created with different approaches and base models. The single-output models show good performance, with mean absolute errors around 1 dB, good rank correlations and R2 scores above 0.9. Support vector regression provides reasonably good agreement from experiments requiring only a small effort to set up; Neural Networks achieve better performance, but increased effort is required to obtain the model.

Assessing environmental impacts across the full life cycle of buildings is essential for advancing toward a net-zero and regenerative built environment. However, life cycle inventory generation and impact assessment remain methodologically complex and time-intensive, limiting their integration into early design decision-making. This study aims to quantify and reduce the embodied carbon of a regenerated building while optimizing material selection based on environmental performance and circularity potential. An integrated Building Information Modeling–Life Cycle Assessment (BIM–LCA) framework combined with Sensitivity Analysis (SA) was applied within a circular economy perspective. A regenerative building was modeled using BIM, and Industry Foundation Classes (IFC) data were employed to conduct a detailed life cycle assessment to quantify embodied carbon and identify emission hotspots across life cycle stages. The results indicate that material extraction, processing, and manufacturing dominate environmental impacts, contributing more than 85% of total CO2 emissions. Sensitivity analysis further demonstrates the influence of material choices on overall carbon performance. The findings underscore the importance of evaluating embodied carbon at early design stages to support informed decisions regarding material efficiency, renewability, and recyclability. The proposed BIM–LCA framework provides a scalable, data-driven approach to support early-stage decarbonization strategies and contributes to reducing the carbon footprint of buildings in alignment with net-zero and regenerative design objectives.

This study addresses the energy significance of the coefficient of performance (COP) in vapor compression systems and the practical need to forecast COP quickly and reliably. Because COP directly reflects the amount of cooling delivered per unit of input power, accurate prediction supports energy savings, refrigerant selection, and early stage design decisions, especially for low Global Warming Potential (GWP) refrigerants. Authors develop data-driven models to estimate COP without full thermodynamic calculations. A synthetic dataset of 2,000 samples is generated in Engineering Equation Solver (EES) for four refrigerants (R1234yf, R134a, R290, R600a) by using five inputs: refrigerant type, evaporation temperature, condensing temperature, subcooling, and superheat. Five supervised learning algorithms are trained and compared: Linear Regression, Polynomial Regression, Random Forest, Decision Tree, and Support Vector Machine. The study evaluates model performance using the Coefficient of Determination (R²), Root Mean Squared Error (RMSE), and Mean Absolute Error (MAE) based on an 80/20 train/test split. Results show Polynomial Regression (degree 3) delivers the highest accuracy (R² ≈ 0.9999; RMSE ≈ 0.0071; MAE ≈ 0.0053), with Random Forest as the next strongest baseline. The findings suggest that lightweight, well-tuned regressors can provide fast, precise COP predictions, reducing analysis time while guiding system design and parameter optimization. The approach offers an accessible tool for engineers seeking efficient, low-carbon refrigeration solutions.

Graphical assets play an important role in design and development of games. There is potential in the use of AI-driven generative tools to aid in creation of such assets, improving pipelines. However, there is little research to address how generative methods can fit into the wider pipeline, and no guidelines or heuristics for creating such tools. Hence, we conducted a user study with 16 game designers and developers to examine their behaviour and interaction with such tools. Findings highlight that early design stage is preferred by all participants. Designers and developers prioritise rapid variations over initial quality of assets. Results also strongly raised the need for better integration of such tools in existing design/development environments and pipelines, specifically regarding common data formats and output manipulability. Informed by these results, we provide a set of heuristics for creating tools that meet the expectations and needs of game designers and developers.

Domain-specific languages for hardware can significantly enhance designer productivity, but sometimes at the cost of ease of verification. On the other hand, ISA specification languages are too static to be used during early stage design space exploration. We present PEak, an open-source hardware design and specification language, which aims at improving both design productivity and verification capability. PEak does this by providing a single source of truth for functional models, formal specifications, and RTL. PEak has been used in several academic projects, and PEak-generated RTL has been included in three fabricated hardware accelerators. In these projects, the formal capabilities of PEak were crucial for enabling both novel design space exploration techniques and automated compiler synthesis.

Generative AI (GenAI) is increasingly applied in urban planning for text production, visualization, analytics, stakeholder communication, and participatory engagement. Large language models (LLMs) enable the creation of synthetic participants to support the early-stage design, analysis, and testing of participatory tools. This article demonstrates an innovative use of GenAI through synthetic inhabitants and experts in an immersive digital urban planning environment. DigitalTurku serves as a proof-of-concept for an immersive planning tool within an urban digital twin. The case relies on synthetic personas—residents and expert stakeholders—to evaluate how a GenAI-assisted urban platform may shape participation experiences and trust in local urban planning. The findings indicate that synthetic experts expressed a reduced bureaucratic distance, enhanced transparency, and more meaningful participation. However, assessments of tools and digital environment usability varied according to digital skills and demographic characteristics embedded in the personas. The use of synthetic personas helps identify opportunities and challenges in immersive urban planning environments and supports the design of digital tools in smart cities to strengthen human residents’ spatial understanding and experiential engagement in planning processes. The creation of synthetic data and participants is convenient with LLMs. Despite these tools’ limitations, they can play a valuable role in piloting participatory planning processes to support and complement human-based participation.

Reducing the environmental impact of logistics warehouses is a critical challenge, particularly during the early design phase when limited data is available to guide decision-making. This study aims to establish carbon footprint targets for logistics warehouses in alignment with climate neutrality objectives. Using Life Cycle Assessment (LCA) methodologies, the environmental impacts of 16 Lidl Company logistics warehouses in France were evaluated. A correlation analysis revealed that warehouse size and cold storage capacity are the strongest predictors of carbon footprint (Pearson coefficients of 0.78 and 0.68, respectively). Based on these relationships, a carbon footprint threshold function was developed using a linear regression model optimized by a Non-dominated Sorting Genetic Algorithm II (NSGA-II), achieving an error margin below 7%. The resulting model quantifies emissions per pallet space according to storage type, ranging from 1.11 TCO 2 eq/pallet for dry goods (high shelf) to 4.96 TCO 2 eq/pallet for fresh-produce block storage. These findings demonstrate that achieving carbon neutrality for logistics warehouses requires not only energy-efficient operations but also substantial reductions in embodied emissions through low-carbon materials and optimized design strategies. The predictive carbon footprint threshold function proposed here provides a robust, data-driven tool to guide the design of future industrial buildings aligned with national and international sustainability goals.

ABSTRACT This study presents a clear and analytically explicit framework for exploring hybrid hydrogen–electric propulsion in micro-to-small fixed-wing unmanned aerial vehicles (UAVs). By combining classical lift–drag relationships with a first-principles energy balance, the model expresses flight endurance and range through simple, closed-form relations involving hydrogen mass, battery capacity, and cruise speed. Unlike approaches that rely on computationally intensive CFD simulations or extensive hardware testing, the framework is entirely physics-based, enabling fast, transparent, and reproducible performance estimation during early design stages. A parametric design-space analysis examines how mass allocation between hydrogen and battery storage influences endurance and range. Under idealized, constant-mass and constant-efficiency assumptions, the analytical model predicts that balanced hybrid configurations can yield theoretical endurance values approaching 70 hours and ranges on the order of 3700 km, representing a substantial improvement relative to battery-only operation. Contour-based visualization illustrates trade–offs associated with hybridization ratios and cruise conditions, providing intuitive insight into system-level behaviour. Owing to its simplicity, interpretability, and open analytical structure, the framework serves as a practical conceptual-design tool for UAV sizing, propulsion planning, and sustainable long-endurance flight studies. The model supports early-stage decision-making by offering a physics-grounded means of exploring hybrid UAV performance trends prior to more detailed analyses.

Coordinated robust optimization of building and surrounding microclimate in early-stage design under uncertainty

Frequency-domain tonal noise prediction models provide a rapid means of assessing rotational noise, which is particularly useful during the early stages of rotorcraft and propeller design and optimization. This paper presents a comprehensive review of the most widely used frequency-domain models, including four steady loading noise models, three unsteady loading noise models, and three thickness noise models. Unified and consistent formulations are provided for all models, and the contributions of individual terms are grouped and analyzed. The models are then compared under various operating conditions, including steady hover, unsteady hover, axial forward flight, and edgewise forward flight. Overall, the models show good agreement with measurement data, although certain deficiencies have also been identified. Finally, recommendations for model usage under different flight conditions are provided.

• Lattice body compliance is incorporated into analytical gear mesh stiffness • FEM results agree with analytical predictions within 5% deviation • Unit-cell size accounts for approximately 74–75% of stiffness variation • Pareto optimization enables about 52% mass reduction • Tooth-tip deflection is limited to 0.349 mm at the optimal design This study proposes an analytical–numerical framework for evaluating mesh stiffness in polymer lattice-core gears manufactured by additive manufacturing, in which lattice body compliance is explicitly integrated into analytical gear mesh-stiffness modeling. A three-dimensional mesh-stiffness model augments classical tooth compliance components, with a gear-body compliance term K body derived from the Gibson–Ashby scaling law for bending-dominated lattices. A response surface methodology based on a face-centered central composite design (FCCCD) is employed to quantify the influence of unit-cell size and strut diameter on translational mesh stiffness, tooth-tip deflection, and mass reduction. Unit-cell size is the dominant factor, explaining approximately 74–75% of the variance, while strut diameter provides secondary stiffening. A balanced design (a ≈ 10 mm, t ≈ 1 mm) achieves approximately 52% mass reduction while maintaining deformation within acceptable limits, offering practical guidance for early stage design. The proposed framework establishes a direct analytical link between lattice topology and gear mesh stiffness, providing a reliable, design-oriented tool for lightweight polymer gears in mechatronic and automotive applications. Experimental validation and extension to TPMS and graded lattices are identified as directions for future work.

An analytical framework for quantifying recoil and foreshortening in ultra-thin cobalt–chromium coronary stents

RNA-based technologies have demonstrated significant potential for diverse applications, ranging from vaccination to gene editing. However, their widespread adoption is limited by the critical challenge of efficient delivery. Lipid nanoparticles (LNPs) have emerged as a widely utilized RNA delivery system, yet their formulation design and optimization primarily rely on empirical trial-and-error, which is labor-intensive, time-consuming, and cost-prohibitive, thus hindering the rapid development of RNA therapeutics. To facilitate the early-stage design and optimization of LNPs for enhanced delivery efficiency, in this study, we construct LNPs-TE, a benchmark dataset comprising over 10 000 experimentally measured transfection efficiency (TE) values, and introduce LNPs integrated feature fusion Transformer (LIFT), a deep learning framework for LNPs TE prediction. Comprehensive experiments demonstrate that LIFT effectively integrates multidimensional molecular representations of ionizable lipids, the key component in LNPs formulation, achieving superior predictive performance, with an average Pearson correlation coefficient of 0.845 for regression and an area under the receiver operating characteristic curve (AUC-ROC) of 0.818 for multi-class classification across multiple datasets. Through scaffold-based splitting and activity cliff tasks, we further validated the exceptional generalization ability and robustness of LIFT, which achieved over a 10% improvement in the coefficient of determination (R2) compared with state-of-the-art baseline models, highlighting its potential as a practical and stable approach for the virtual screening of efficient LNPs formulation. The relevant data, model and code are made publicly available at https://github.com/U12458/LIFT.