Architectural Design Research Articles

Ontology-Based Design Features for Representing Constructability in Architectural Design: Toward BIM in Off-Site Construction

Ontology-Based Design Features for Representing Constructability in Architectural Design: Toward BIM in Off-Site Construction

Multimodal representations of biomedical knowledge from limited training whole slide images and reports using deep learning.

The increasing availability of biomedical data creates valuable resources for developing new deep learning algorithms to support experts, especially in domains where collecting large volumes of annotated data is not trivial. Biomedical data include several modalities containing complementary information, such as medical images and reports: images are often large and encode low-level information, while reports include a summarized high-level description of the findings identified within data and often only concerning a small part of the image. However, only a few methods allow to effectively link the visual content of images with the textual content of reports, preventing medical specialists from properly benefitting from the recent opportunities offered by deep learning models. This paper introduces a multimodal architecture creating a robust biomedical data representation encoding fine-grained text representations within image embeddings. The architecture aims to tackle data scarcity (combining supervised and self-supervised learning) and to create multimodal biomedical ontologies. The architecture is trained on over 6,000 colon whole slide Images (WSI), paired with the corresponding report, collected from two digital pathology workflows. The evaluation of the multimodal architecture involves three tasks: WSI classification (on data from pathology workflow and from public repositories), multimodal data retrieval, and linking between textual and visual concepts. Noticeably, the latter two tasks are available by architectural design without further training, showing that the multimodal architecture that can be adopted as a backbone to solve peculiar tasks. The multimodal data representation outperforms the unimodal one on the classification of colon WSIs and allows to halve the data needed to reach accurate performance, reducing the computational power required and thus the carbon footprint. The combination of images and reports exploiting self-supervised algorithms allows to mine databases without needing new annotations provided by experts, extracting new information. In particular, the multimodal visual ontology, linking semantic concepts to images, may pave the way to advancements in medicine and biomedical analysis domains, not limited to histopathology.

Training-Free Transformer Architecture Search With Zero-Cost Proxy Guided Evolution.

Transformers have shown remarkable performance, however, their architecture design is a time-consuming process that demands expertise and trial-and-error. Thus, it is worthwhile to investigate efficient methods for automatically searching high-performance Transformers via Transformer Architecture Search (TAS). In order to improve the search efficiency, training-free proxy based methods have been widely adopted in Neural Architecture Search (NAS). Whereas, these proxies have been found to be inadequate in generalizing well to Transformer search spaces, as confirmed by several studies and our own experiments. This paper presents an effective scheme for TAS called TRansformer Architecture search with ZerO-cost pRoxy guided evolution (T-Razor) that achieves exceptional efficiency. First, through theoretical analysis, we discover that the synaptic diversity of multi-head self-attention (MSA) and the saliency of multi-layer perceptron (MLP) are correlated with the performance of corresponding Transformers. The properties of synaptic diversity and synaptic saliency motivate us to introduce the ranks of synaptic diversity and saliency that denoted as DSS++ for evaluating and ranking Transformers. DSS++ incorporates correlation information among sampled Transformers to provide unified scores for both synaptic diversity and synaptic saliency. We then propose a block-wise evolution search guided by DSS++ to find optimal Transformers. DSS++ determines the positions for mutation and crossover, enhancing the exploration ability. Experimental results demonstrate that our T-Razor performs competitively against the state-of-the-art manually or automatically designed Transformer architectures across four popular Transformer search spaces. Significantly, T-Razor improves the searching efficiency across different Transformer search spaces, e.g., reducing required GPU days from more than 24 to less than 0.4 and outperforming existing zero-cost approaches. We also apply T-Razor to the BERT search space and find that the searched Transformers achieve competitive GLUE results on several Neural Language Processing (NLP) datasets. This work provides insights into training-free TAS, revealing the usefulness of evaluating Transformers based on the properties of their different blocks.

Introducing circularity in early architectural design education

PurposeWith around 40% of global waste attributed to the built environment sector, architects play a significant role in how resources are consumed, produced and wasted. UK architectural educators have made good progress to embed climate emergency issues in design curricula but the challenges of resource use and waste, and the opportunities afforded by circular economy design principles are less well-considered. The purpose of this paper is to provide new insights on how to introduce circular design principles to early-stage architectural design students within university curricula.Design/methodology/approachThe study described took an experimental approach to designing design project briefs across several interlinked studio design projects for 1st-year Architecture students. Structured as a case study, each section describes the project, learning objectives, teaching methods and project reflections informed by a qualitative assessment of student development, outputs and feedback gathered through a questionnaire and focus group.FindingsIntroducing circular design early is highly beneficial to the development of knowledge on sustainability issues, critical design thinking and creative solution generation. Examples of beneficial teaching approaches include building systems thinking, facilitating collaboration, supporting learning-through-making, using simple analogies and referencing best-practice examples.Originality/valueThis research builds on limited existing circular design literature for built environment fields and through practical insights fills a significant knowledge gap on ways to introduce a complex and dynamic topic such as the circular economy to early-stage architectural design students as they develop fundamental discipline-specific knowledge, skills and competences.

Turbocharging protein binding site prediction with geometric attention, inter-resolution transfer learning, and homology-based augmentation

BackgroundLocating small molecule binding sites in target proteins, in the resolution of either pocket or residue, is critical in many drug-discovery scenarios. Since it is not always easy to find such binding sites using conventional methods, different deep learning methods to predict binding sites out of protein structures have been developed in recent years. The existing deep learning based methods have several limitations, including (1) the inefficiency of the CNN-only architecture, (2) loss of information due to excessive post-processing, and (3) the under-utilization of available data sources.MethodsWe present a new model architecture and training method that resolves the aforementioned problems. First, by layering geometric self-attention units on top of residue-level 3D CNN outputs, our model overcomes the problems of CNN-only architectures. Second, by configuring the fundamental units of computation as residues and pockets instead of voxels, our method reduced the information loss from post-processing. Lastly, by employing inter-resolution transfer learning and homology-based augmentation, our method maximizes the utilization of available data sources to a significant extent.ResultsThe proposed method significantly outperformed all state-of-the-art baselines regarding both resolutions—pocket and residue. An ablation study demonstrated the indispensability of our proposed architecture, as well as transfer learning and homology-based augmentation, for achieving optimal performance. We further scrutinized our model’s performance through a case study involving human serum albumin, which demonstrated our model’s superior capability in identifying multiple binding sites of the protein, outperforming the existing methods.ConclusionsWe believe that our contribution to the literature is twofold. Firstly, we introduce a novel computational method for binding site prediction with practical applications, substantiated by its strong performance across diverse benchmarks and case studies. Secondly, the innovative aspects in our method— specifically, the design of the model architecture, inter-resolution transfer learning, and homology-based augmentation—would serve as useful components for future work.

Advancing Thermochromic Glass Durability: Reinforcing Thermosensitive Hydrogels with Enhanced Adhesion Techniques

The growing use of glass in architecture has driven research into reducing its energy consumption. Thermochromic (TC) glass technology shows promise for enhancing building energy efficiency by regulating solar heat dynamically. Although TC glass helps reduce heat radiation, additional solutions like Low-E or vacuum glass are needed to control heat convection and conduction. Low-E glass, while effective in lowering heat transfer, may increase surface temperature. Thermo-sensitive hydrogels, known for their light-scattering properties at high temperatures, have been explored to complement TC glass. However, their stability at elevated temperatures remains a challenge, especially for applications requiring durability under varying weather conditions. This study proposes enhancing the adhesion between hydrogel and glass by introducing silica–oxygen bonds. As a result, TC glass demonstrates stable performance over 100 cycles within temperature ranges from 85 °C to 30 °C in summer and 40 °C to −20 °C in winter. Furthermore, by incorporating ethylene glycol, the freezing point of TC glass is reduced to −26 °C, rendering it suitable for use in colder regions. The implementation of TC glass effectively addresses the dual requirements of summer shading and winter heating in areas with both cold winters and hot summers, significantly reducing building energy consumption. This study contributes substantially to developing advanced intelligent building materials, paving the way for more sustainable architectural designs.

Experimental Infrastructure Design for Energy-Independent Car Park Building Based on Parametric Photovoltaic Facade System

The purpose of this study is to develop a new architectural model that responds to environmental pollution. The subject of this study is infrastructure buildings related to automobiles, which cause environmental pollution. Parking facilities accommodate several vehicles, necessitating the design of large-scale parking infrastructure. In this study, the parametric design of an energy-independent building was developed targeting the facade of a large-scale parking facility. As basic research for the development of the parametric design, a parking building was planned toward the optimization of parking space. Based on this basic research, a kinetic photovoltaic facade was developed to achieve optimal renewable energy generation from the perspective of eco-friendly architectural design. Energy simulation using building information modeling (BIM) on the kinetic photovoltaic system developed in this study over a period of one year resulted in the generation of a total of 692,386 kWh·year−1. The novelty of this study is the development of a kinetic photovoltaic facade that is oriented according to the optimal tilt angle every month, focusing on the infrastructure. The significance of the kinetic photovoltaic system lies in the fact that it not only maximizes the efficiency of renewable energy generation but also presents a new architectural design model.

SfM Photogrammetry for Cost-Effective 3D Documentation and Rock Art Analysis of the Dombate Dolmen (Spain) and the Megalithic Sites of Chã dos Cabanos and Chã da Escusalha (Portugal)

SfM (structure from motion) photogrammetry is a technique developed in the field of computer vision that enables the generation of three-dimensional (3D) models from a set of overlapping images captured from disparate angles. The application of this technique in the field of cultural heritage, particularly in the context of megalithic monuments, is inherently challenging due to the spatial constraints of these environments and the usual limitations posed by their architectural design, which often results in poor lighting conditions. This article presents an accurate and cost-efficient methodology for the study and documentation of rock art, which has been applied to three megalithic monuments in the Iberian Peninsula: one in Spain and two in Portugal. The three working environments are complex, but the combination of techniques used and improvements such as rendering for the enhancement of engravings and the creation of 3D stop-motion models made it possible to integrate all the information in 3D formats that allow its universal dissemination. This not only preserves the heritage in graphic form but also makes it accessible to the public, both for study and for virtual visits.

Exploring the Facilitating Environment for Mindfulness Meditation: Establishing a Framework through Analysis of Means and Exploratory Factor Analysis

Abstract Objective While mindfulness meditation is an established practice for improving well-being, the physical environment for this activity awaits further exploration. This research aimed to explore how the physical environment facilitates mindfulness meditation and to establish a framework through Analysis of Means (ANOM) and Exploratory Factor Analysis (EFA). The framework was intended to provide guidance for organisations and individuals to better design or modify physical space to support their mindfulness meditations. Method This study used methods to establish an initial research framework from the literature on mindfulness, architectural design, and environmental psychology. A survey of UK mindfulness practitioners was conducted in 2021–2022 to provide insights from a quantitative perspective. The data were analysed using ANOM and EFA to obtain a comprehensive framework based on the responses. Results Through a review of the literature and empirical study, nine physical factors with controllable elements were identified, ranked by their relative importance. Quietness, the use of supportive tools, and natural sounds were found to be the most important. Conclusion This research suggests the priority elements for groups and individuals to consider when adapting the physical environment for mindfulness meditation. The results revealed quietness as the most beneficial controllable element for its facilitation. Additionally, the use of supportive meditation tools was also important. Considering these top elements first could support practitioners' mindfulness meditation more effectively and efficiently. Preregistration This study was not preregistered.

Automatic Architecture Design for Distributed Quantum Computing

Abstract In distributed quantum computing (DQC), quantum hardware design mainly focuses on providing as many as possible high-quality inter-chip connections. Meanwhile, quantum software tries best to reduce the required number of remote quantum gates between chips. However, this "hardware first, software follows" methodology may not fully exploit the potential of DQC. Inspired by classical software-hardware co-design, this paper explores the design space of application-specific DQC architectures. More specifically, we propose AutoArch, an automated quantum chip network (QCN) structure design tool. With qubits grouping followed by a customized QCN design, AutoArch can generate a near-optimal DQC architecture suitable for target quantum algorithms. Experimental results show that the DQC architecture generated by AutoArch can outperform other general QCN architectures when executing target quantum algorithms.

INTEGRATION OF DIGITAL FABRICATION LABORATORIES AND COMPUTATIONAL DESIGN INTO THE ARCHITECTURE CURRICULUM IN TÜRKİYE

The discipline of architecture undergoes transformation practically, cognitively, and pedagogically due to the widespread use of digital design and fabrication technologies in design processes. This paper aims to eluctade the integration of digital fabrication and computational design with architectural design education at undergraduate level. For this purpose, this study examined the undergraduate course contents and spatial conditions of leading universities with digital fabrication laboratories under three headings: digital design, digital fabrication, and spatial reflections. In addition, this research applied semi-structured interviews with the relevant faculty members of three universities that have robotic tools to obtain more detailed information about their pedagogical approaches. According to this study, the curriculum and laboratories are rapidly changing. As a result, architecture departments are expanding their facilities by building laboratories and tools that have the potential for architectural education to fill the gap between practice and theory. However, there are economic limitations to implementing digital fabrication technologies in terms of the tools and materials used by hundreds of students. This research discusses the requirements, potentials, and limitations of theoretical and practical training through digital fabrication tools that should be included in these courses in the curriculum.

Shaping Tomorrow: The Impact of AI on Architectural History and Interior Design Education

Artificial Intelligence (AI) techniques have become popular in architecture and design, and several studies have focused on using these technological advancements to solve various architectural problems. AI is used in various architectural design applications, from intelligent material composition to layout solutions, and it is also vital in supporting the architecture and design education mechanism. A comprehensive understanding of literature is necessary to use these powerful tools in education adequately. This is due to the large volume of research being created and disseminated on this subject as well as the increasing application of AI techniques to address various education design-related questions. This article offers a comprehensive and critical assessment of the study of artificial intelligence applications in architecture and interior design education as a course application in the history of architecture and interior design. The study's conclusions indicate that AI's implications in architecture and design re-imagination are beneficial; however, it was concluded that AI- in its current phase- cannot replace human input and perspective. The students' feedback indicates that critical thinking skills cannot be replaced, and AI complements but doesn't replace their intellect.

Influencing Factors of Spatial Ability for Architecture and Interior Design Students: A Fuzzy DEMATEL and Interpretive Structural Model

Spatial ability is not just a skill but a crucial element for architecture and interior design students, significantly impacting their proficiency in tasks involving 2D drawings, 3D components, and artistic expression. Despite extensive research in this area, a gap remains in the understanding of how to effectively cultivate spatial ability through educational interventions. This study, with its unique approach of identifying key influencing factors and their interrelationships within the fuzzy decision-making laboratory analysis method (Fuzzy-DEMATEL) and the interpretative structural model (ISM), fills this gap. The method visualizes cause-and-effect relationships within a structural model and captures the interdependencies between influencing factors. In a collaborative effort between nine universities in 2023–2024, 17 experts selected through purposeful sampling contributed to the development of a comprehensive list of potential influencing factors. After refinement through filtering, comparison with the existing literature, and expert consensus, seven influencing factors of spatial ability for architecture and interior design students from personal traits and STEAM disciplines were identified, which are sketching and hand drawing skills, mathematical skills, video game practice, descriptive geometry skills, augmented reality practice, spatial talk, and gesturing while talking. Sketching and hand drawing skills, mathematical skills, and video game practice come under cause factors of spatial ability, whereas the rest are effect factors. Proceeding with ISM analysis revealed that sketching and hand drawing skills and mathematical skills are located in the input layer and have a continuous impact on spatial ability. Descriptive geometry skills lie in the transition layer, which are considered as deep influencing factors, the rest of the factors lie in the effect layer. This study delves into the theoretical and practical implications of these findings, offering valuable insights for educational policy and practice.

Single glass and polymer coated phosphate glass microwire photoactuators with instant response times and large actuating angles

Over the years significant scientific attention has been given towards the design of advanced photoactuating architectures as they are important elements for various optical tweezers, grippers, and soft robot applications. Most of the currently available photoactuators, rely on the existence of a free-space illumination pathway from the light sources to the device or on the employment of bendable optical fibers, while consisting of two or more elements, i.e. the photoactuating material and the substrate. Herein, we report on a facile method of preparing single-component pristine microwires (MWs) that exhibit photoactuating features upon utilizing a soft phosphate glass doped with silver nanoparticles (AgNPs), without the need of any additional photoactuating element. The developed photoactuators exhibited bending angles of around 110°, within a couple of seconds. In addition, we have fabricated double-component polymer-coated phosphate glass MWs photoactuators, upon employing PDMS coatings on the surface of the so-formed glass MWs. The introduction of the polymer component boosts the bending angle of the photoactuating device to over 200° within the a few seconds upon modest laser irradiation. The presence of AgNPs within the glass MWs, play a key role on the remarkable performance of the developed photoactuators, both in terms of actuating angles as well as of the respective response times, since they assist on the effective transmission of laser irradiation energy to thermal energy. The fabrication method reported here appears promising for the development of high-performance and low-cost free-space, as well as fiber-based photoactuators.

An analytical model from physical parameters to minimum ranging time for photon-counting LiDARs

Long-distance light detection and ranging (LiDAR) has been highly demanded for applications on unmanned vehicles and drones. CMOS-fabricated single-photon avalanche diodes (SPADs) play a key role in the receiver end due to their high photo-sensitivity and readiness for system-on-chip integration. However, the large amounts of involved components together with the diverse ranging conditions make engineering and optimizing these modules a daunting challenge. In this work, we have developed an analytical model for calculating minimum ranging time from the physical parameters for a photon-counting LiDAR. The experimental verifications of the model have been performed and a good consistency has been obtained. Our work enables architecture design and optimization for making low-cost high-performance SPAD LiDARs.

Comparative analysis of neural network architectures for short-term FOREX forecasting

This document outlines the analysis, design, implementation, and benchmarking of various neural network architectures in a short-term frequency prediction system for the FOREX market. Our objective is to emulate the decision-making process of a human expert (technical analyst) through a system that swiftly adapts to market condition changes, thereby optimizing short-term trading strategies. We have designed and implemented a series of LSTM neural network architectures that take exchange rate values as input to generate short-term market trend forecasts. Additionally, we developed a custom ANN architecture based on simulators for technical analysis indicators. We performed a comparative analysis of the results and came to useful conclusions regarding the suitability of each architecture and the cost in terms of time and computational power to implement them. The ANN custom architecture produces better prediction quality with higher sensitivity using fewer resources and spending less time than LSTM architectures. The ANN custom architecture appears to be ideal for use in low-power computing systems and for use cases that need fast decisions with the least possible computational cost.

ARCHITECTURE DESIGN OF HEALTH ASSET DETECTION SYSTEM IN HOSPITAL

Efficient management of hospital assets is essential to ensure that operations can run optimally and the quality of health services is good. However, the recording and management of assets in hospitals carried out manually often causes data errors, information mismatches, and also assets are only known by the manager without being explicitly recorded. In overcoming this problem, the researcher aims to develop a hospital asset detection system architecture using an iterative and incremental methodology approach. The stages of this system development include identification of needs and conceptual models, logical architecture design, conceptual architecture design, logical architecture, physical architecture, technology selection, and evaluation. This system utilizes YOLO model reading technology for asset detection and identification, storing detection results into a local database using SQLite3, sending data to a central server via API, and post-processing data by selecting the highest confidence score stored in a MySQL database and then using the data to manage asset management and asset visualization. The implementation of this system successfully reduces manual recording time, improves asset visibility, and optimizes resource usage, thus contributing to the improvement of efficiency and quality of health services.

Needle-Plug/Piston-Based Modular Mesoscopic Design Paradigm Coupled With Microfluidic Device for Point-of-Care Pooled Testing.

Emerging diagnostic scenarios, such as population surveillance by pooled testing and on-site rapid diagnosis, highlight the importance of advanced microfluidic systems for in vitro diagnostics. However, the widespread adoption of microfluidic technology faces challenges due to the lack of standardized design paradigms, posing difficulties in managing macro-micro fluidic interfaces, reagent storage, and complex macrofluidic operations. This paper introduces a novel modular-based mesoscopic design paradigm, featuring a core "needle-plug/piston" structure with versatile variants for complex fluidic operations. These structures can be easily coupled with various microfluidic platforms to achieve truly self-contained microsystems. Incorporated into a "3D extensible" design architecture, the mesoscopic design meets the demands of function integration, macrofluid manipulations, and flexible throughputs for point-of-care nucleic acid testing. Using this approach, an ultra-sensitive nucleic acid detection system is developed with a limit of detection of ten copies of SARS-CoV-2 per mL. This system efficiently conducts large-scale pooled testing from 50 pharyngeal swabs in a tube with an uncompromised sensitivity, enabling a truly "sample-in-answer-out" microsystem with exceptional performance.

Research on the Design of Medical Imaging Assisted Diagnosis System Based on Machine Learning

The design and application of medical imaging diagnostic assistance systems have become an indispensable facet of modern medicine. Technologies based on machine learning can handle vast and intricate medical imaging data, offering precise diagnostic support and significantly enhancing the efficiency and accuracy of clinical diagnoses. This study focuses on the development and implementation of an efficient medical imaging diagnostic assistance system utilizing machine learning algorithms such as Principal Component Analysis (PCA), Support Vector Machines (SVM), and Multi-Layer Perceptrons (MLP). Through the architectural design of intelligent diagnostic assistance systems, data preprocessing, and feature vectorization, combined with the application of classification algorithms, a robust diagnostic support system capable of addressing diverse medical scenarios has been constructed. Performance evaluation experiments indicate that this system demonstrates high accuracy and robustness in processing medical imaging data, holding potential for clinical implementation. The findings of this study provide new insights for the advancement of intelligent diagnostic systems in medical imaging and lay a solid foundation for the development of future clinical diagnostic tools.

Lightweight CNN architecture design for rolling bearing fault diagnosis

Lightweight CNN architecture design for rolling bearing fault diagnosis