ABSTRACT In response to ecological sustainability and resource challenges, this study proposes an Eco-Service Design (ECO-SD) framework for university dormitories, as essential collective living spaces, with the goal of aligning human well-being and ecological transition. The framework consists of four stages: (1) Eco-Service Exploration to identify ecological values and user needs; (2) Eco-Service Visualization using service blueprints to identify barriers and potential impacts; (3) Eco-Service Ideation through KJ clustering, Analytic Hierarchy Process (AHP), Quality Function Deployment (QFD), Function Behavior Structure (FBS) modeling, and co-creation workshops to generate multifunctional solutions; and (4) Eco-Service Assessment using Heart Rate Variability (HRV), Pleasure-Arousal-Dominance (PAD) emotional analysis, Sustainability Design-Orienting (SDO) assessment, and ECO-SD evaluation. The experimental validation, combining HRV measurement and PAD emotional analysis, showed that the proposed Eco-Service-based design increased SDNN and RMSSD while reducing LF/HF, indicating greater emotional stability and relaxation. Participants reported a joyful emotional state, confirming the design’s positive impact on emotional well-being. The SDO and ECO-SD evaluation also demonstrated strong sustainability performance, validating the framework's ecological and service effectiveness. This interdisciplinary approach offers a replicable methodology for sustainable innovation in the design of collective living support systems and contributes to the broader discourse on ecological transition.

ABSTRACT With the rapid development of modern construction industry, intelligent construction has become a key direction to improve engineering efficiency and reduce construction risks. Identifying and predicting potential risks is crucial in the intelligent construction. However, traditional risk identification methods suffer from strong subjectivity and low efficiency. Therefore, an intelligent construction risk prediction model based on Bayesian optimized extreme gradient boosting is proposed. By introducing binary classification and cross entropy functions, the accuracy and stability of the model are optimized. Then, Bayesian optimization is used to select optimal parameters to improve the efficiency of risk prediction. The results showed that the average accuracy of the designed model in the training and testing sets was 99.23% and 98.16%, respectively, which was much higher than models including random forest, light gradient boosting machine, and K-nearest neighbor. Meanwhile, in terms of computational efficiency, the total computation time of the proposed model was only 0.505s. The research has demonstrated the optimization effect of Bayesian on the extreme gradient boosting model and its effectiveness for risk prediction in intelligent construction, aiming to provide guarantees for the development and safety of intelligent construction.

ABSTRACT The building’s exterior envelope serves as the primary barrier against adverse environmental conditions and plays a crucial role in reducing energy consumption. This study evaluates the impact of optimized combinations of façade parameters, including window dimensions and types, thermal insulation thickness, and shading devices, on primary energy consumption, spatial daylight autonomy (sDA), annual sunlight exposure (ASE), and thermal comfort in a residential building located in the cold and dry climate. Variables were generated in Rhino and Grasshopper and assessed using the Honeybee tool. To select parameters rationally and in a short period, a decision matrix and the ROMINA method were employed to guide the design toward optimal conditions. The study was conducted in two steps to reduce complexity. The advantage of this method is its consideration of the interaction effects of parameters, allowing for the simultaneous optimization of all four indicators. This can be considered a rational and rapid method in the design process of buildings for any use and any climate. By examining the best and worst design scenarios in the building under study, energy consumption can be improved by 22.5%, thermal comfort by 37.8%, daylight autonomy by 50.86%, and glare by 70.32%.

ABSTRACT Lack of sufficient experience and easy access to previous historical and technical information on maintenance operations, along with various changes in maintenance staff, increase cost and decrease life cycle of building facilities. This study addresses this problem by presenting a framework to (1) use historical experience for a location-aware maintenance solution recommendation and (2) improve access to building facilities’ technical and historical information on-site. For this purpose, the framework integrates three digital technologies: Augmented Reality (AR), Building Information Model (BIM), and Artificial Intelligence (AI). The framework enables communication with facility managers and staff by natural language with the aid of the AI module. This module also employs case-based reasoning to recommend appropriate solutions using previous successful maintenance operations. The BIM module includes facility information and its maintenance history. The AR module recognizes elements on-site with electronic mobile devices and represents information about facilities. A University building case study is employed to evaluate the practicality and effectiveness of the proposed framework. Facility managers and staff of this building used the implemented framework and indicated that this framework is easy to use and suggests appropriate maintenance solutions.

ABSTRACT Urban environments profoundly influence human well-being and behavior, underscoring the need for design paradigms that seamlessly integrate ecological principles with stakeholder requirements. This study investigates the convergence of affective computing and generative design to develop adaptive, health-promoting urban environments. Initially, a genetic-generative algorithm was created to generate an extensive database of high-rise tower designs optimized for solar exposure and surface-to-volume ratio. Subsequently, an EEG-based Brain–Computer Interface (BCI) system was implemented to capture architects’ subconscious emotional responses to selected designs using Event-Related Potentials (ERPs) and Self-Assessment Manikin (SAM) scales. EEG data from 24 participants were analyzed to extract ERP markers of valence-based preference, revealing significant neural responses in early (250–350 ms) and late (600–800 ms) time windows. A random forest model, complemented by SHAP analysis, demonstrated nonlinear influences of critical design parameters on subjective preference. EEG-derived preference scores were then integrated into a multi-objective optimization workflow, facilitating a data-driven, user-centric design selection process. The findings support the potential for real-time, neuroadaptive architectural design that mitigates decision fatigue while harmonizing objective performance metrics with affective insights. Moreover, this work contributes to research-informed public consultation by providing an evidence-based, inclusive approach for incorporating subconscious user preferences into urban planning and architectural workflows.

ABSTRACT With the rapid development of intelligent buildings, scientific and effective quality management is of great significance. Therefore, the study adopts drone oblique photography technology to achieve 3D modeling of intelligent construction sites, and effectively obtains ground point clouds using fabric simulation filtering algorithms. Meanwhile, using Fast RCNN (Fast Region Convolutional Neural Network) to achieve object detection during the construction process. In addition, the Plan Do Check Act (PDCA) cycle will be introduced into overall construction quality management. The research results show that the average total error rate, first level error rate, and second level error rate of the fabric simulation filtering algorithm are 6.14%, 2.01%, and 9.35%, respectively. In the safety helmet wearing detection, the detection accuracy and recall of the fast area convolutional neural network reached 89.6% and 91.8%, respectively, with a frame rate of 39.8 frames per second. In addition, in practical applications, the algorithm has a pass rate of over 88% in various engineering projects. This management strategy combines intelligent construction sites with the planning execution inspection action cycle, which has significant performance advantages and application effects, providing reliable technical support for intelligent remote building quality management.

ABSTRACT Indoor Environmental Quality (IEQ) in school classrooms is essential as it directly influences students’ well-being and academic performance. This research aims to develop the IEQ Point-in-Time Method to categorize existing classrooms based on a quick estimation of their indoor environmental potential. The method is designed to be cost-effective, applicable in various urban and rural educational settings. It enables the comparison of classrooms regarding their design features by categorizing them into defined compliance levels. What sets this method apart is its integration of LiDAR-based digital twins with a checklist of design factors influencing IEQ, providing a streamlined assessment process. Validated in 30 primary school classrooms, it effectively evaluates compliance with established parameters and allows straightforward visualization of results. The findings show that the IEQ Point-in-Time Method effectively ranks classrooms that need intervention to ensure good indoor environmental quality for students. This method allows for efficient categorization and comparison, helping school management or those responsible for school infrastructure to prioritize necessary improvements. Additionally, it is time-efficient, can be conducted by non-specialist staff, and allows for repeated verification of classroom characteristics after the initial assessment.

ABSTRACT This research is conducted in the context of current efforts for net zero delivery in Scotland. Such a target is widely acknowledged as challenging by the industry necessitating some aspects of building design to change for higher performance. This study looks at ‘what is optimal’ in terms of carbon, both in operation and embodied in materials for dwellings by using the Pareto Optimal solution in a simulated case study in Edinburgh. At first, the potential of available glazing systems to find an optimum in operational and embodied carbon is investigated. The study then calculated the embodied carbon and operational carbon for added insulation and compared them against each other to find the most energy and environmentally-efficient option. An analysis to investigate the LCA of the 52 HVAC systems for the second phase of the building's life span is then conducted and finally a parametric analysis of using shading devices to observe if they can offer operational savings for the case study concludes the paper. This study demonstrates some recommended design solutions by globally used standards e.g. Passivhaus and common practices may not lead to carbon emissions minimisation in the refurbishment process.

ABSTRACT Feedback from completed buildings to architects is essential to improve the environmental and social sustainability of future designs. The literature discusses the lack of uptake of traditional Post-Occupancy Evaluations or Building Performance Evaluations due to various issues and barriers. In addition, existing feedback methods and formats may not provide architects with information on aspects they believe are necessary to improve future designs. Furthermore, existing post-occupancy feedback methods were developed before current, broader conceptions of sustainability became accepted and have concentrated on some aspects of sustainability with the omission of others. In the context of this paper, school buildings are selected as an initial typology to examine architect’s perceptions of desired feedback. This paper presents a survey of primary and secondary school architects in Australia, asking on which aspects of school architecture they desire feedback. In summary, they desire feedback from users regarding the usefulness, flexibility and variety of the spaces, different aspects of comfort including felt responses, building support for community engagement, sustainability education, and current and novel pedagogy, in addition to traditional energy and thermal performance measures.

ABSTRACT In recent years, the building industry, as the main source of carbon emissions, has received extensive attention from various sectors of society in context of national carbon peaking and carbon neutrality goals. As a result, prefabricated buildings supported by Building Information Modeling (BIM) technology have been guiding the development of the industry towards new technologies and processes. Due to the gradual increase in building energy consumption and increasing severity of material waste, the government has begun to vigorously advocate and support prefabricated buildings. This paper takes prefabricated reinforced concrete structure buildings as the research object, starting from various stages of the building's entire life cycle, and explores the modern construction technology that combines BIM and prefabricated buildings. It analyzes the planning and analysis design, scheme coordination design, and deepening design of prefabricated components in design stage of prefabricated buildings based on BIM technology. Furthermore, it explores the application of intelligent technology and information technology in the entire life cycle of prefabricated buildings. Based on a deep understanding of theoretical knowledge and practical investigation of prefabricated buildings, this paper explores the application of BIM technology, proposes the feasibility and applicability of applying BIM technology to actual projects, and provides targeted recommendations.