Program comprehension is a significant challenge for programmers and has been widely studied in the context of textual programming languages. The growing use of visual programming languages, particularly in architecture, introduces comparable comprehension challenges, yet related research remains limited. In response, this study investigates an approach to support the comprehension of visual programming languages. Prior research highlights the effectiveness of beacons in supporting the understanding of textual programs. Building on this concept, a set of hypotheses was developed to test the presence of beacons and their influence on visual program comprehension in Grasshopper. An experiment was conducted with participants of varying experience levels, involving program comprehension and memorisation-recall tasks. Eye-tracking was used to record participants’ gaze behavior during comprehension tasks. The findings provide insights into the potential implications of beacons for learning and teaching methodologies in visual programming languages, with relevance for both practitioners and academics.
This paper presents a computational design framework for integrating Tension-Activated Kirigami (TAK) into curvilinear architectural surfaces for passive airflow control. Conventional ventilation strategies often rely on fixed geometries or mechanically actuated components, which can limit adaptability and increase system complexity. This study explores an alternative approach in which airflow modulation is embedded within geometric design and material response. Drawing on kirigami principles, TAK patterns introduce strategic cuts that allow flat sheets to transform into strain-responsive surfaces, forming directional apertures under tension. The framework combines parametric modeling, finite element analysis (FEA), and computational fluid dynamics (CFD) within a coupled design process that relates geometric parameters to both structural deformation and airflow behavior. Simulation and prototyping are used to examine how variables such as slit-to-wall ratio and curvature-informed pattern distribution influence aperture formation and flow performance. The results show that TAK surfaces can redirect airflow, promote air mixing, and reduce stagnation, while maintaining controllable deformation under applied strain. These findings suggest that ventilation performance can be achieved through the interaction of geometry and material behavior, offering a low-energy and adaptable strategy for architectural applications such as roofs, façades, and lightweight enclosures.
Accurate estimation of embodied carbon in residential buildings is crucial for minimizing the environmental impacts of the built environment, particularly during the early design phases. However, conventional assessment methods are often time-consuming, data-intensive, and less adaptable in the early design stages, highlighting the need for predictive approaches such as machine learning. In this study, machine learning algorithms were employed to predict the embodied carbon of residential buildings based on process-based life cycle assessment (PB-LCA) data, using internationally recognized emission factors. Seven different models—Random Forest, XGBoost, LightGBM, Support Vector Regression (SVR), K-Nearest Neighbors (KNN), Gradient Boosting, and Multi-layer Perceptron (MLP)—were trained. After hyperparameter tuning, the four best-performing models were selected for detailed analysis. Beyond predictive accuracy, model interpretability was evaluated using SHapley Additive exPlanations (SHAP) to quantify and visualize the contribution of various input features. Results revealed that Gradient Boosting and Random Forest models emphasized infrastructural attributes such as concrete volume and foundation steel weight, while the XGBoost model showed a stronger focus on macro-level features like total floor area. LightGBM exhibited a hybrid behavior between these two groups. Moreover, non-linear patterns were observed in the influence of certain features—such as concrete volume and steel weight—on embodied carbon, particularly as their values increased. Building plan type and structural system also consistently emerged as key determinants across all models. The findings highlight the potential of advanced machine learning models, combined with interpretability tools like SHAP, not only to enhance prediction accuracy but also to support informed decision-making for architectural and structural optimization. This approach is particularly valuable for reducing environmental impacts during the early stages of building design. Finally, the study recommends incorporating local datasets and accounting for broader factors such as construction processes and supply chains to further improve model reliability and applicability.
Refugee camps are temporary places accommodating people who have been displaced from their homes due to an emergency, where this temporary period may extend up to a lifetime. The literature indicates that the quality of urbanism in refugee camps is typically poor due to the limited time allocated to urban planners for design. Refugee camps have an almost unplanned urban development, which leads to a lack of distribution of land uses and of security. This study aims to provide policymakers with a framework for an assistive tool to predict and intervene when necessary in a continuously changing social context of refugee camps’ urban growth. This framework utilises Agent-based Machine Learning (ABML) to predict the urban growth of the refugee camp, focusing on its physical and social aspects, as well as its specific functions.
This work proposes discrete style transfer (DST) as an alternative to conventional neural style transfer methods, reframing the operation from seamless blending to part-based substitution. Neural models such as VGG-based style transfer, GANs, and diffusion models privilege continuity, producing visually coherent but cognitively estranging images that erase seams essential to human perception, critique, and fabrication. Building on theoretical insights, the paper argues that seams and discontinuities should be treated as design data rather than flaws. The method developed here subdivides images into adaptive quadtree tiles, encodes them with an autoencoder into latent embeddings, and substitutes parts via nearest-neighbor search. Results demonstrate significant gains in speed and stylistic fidelity compared to alternative image comparison methods and mosaic baselines, while remaining competitive with neural style transfer in visual similarity. Beyond efficiency, discrete style transfer produces modular, interpretable outputs aligned with architectural logics of repetition, assembly, and material agency.
