The integration of Computational Fluid Dynamics (CFD) and Artificial Intelligence (AI) have offered a transformative solution to CFD's prohibitive computational cost in built environment application. While AI surrogates successfully accelerate simulations, a systematic review of 229 studies revealed a fundamental disconnect: studies are overwhelmingly distributed across a single spatial scale, such as a room, a building, or a city. True multi-scale integration, essential for capturing the interconnected physics of the built environment, remains critically under-explored. This core challenge is compounded by persistent hurdles in model generalization, physical consistency, and research reproducibility. This review synthesizes these limitations to propose a comprehensive framework that pivots from isolated models toward a unified, physically consistent approach. We advocate for developing hierarchical, physics-informed learning strategies built upon sophisticated benchmark datasets to ensure reliable and scalable information transfer across room-, building-, and city-scales. This work provides the necessary roadmap to bridge this critical gap, enabling robust, multi-scale simulation for sustainable design and operation.

Heating, ventilation, and air-conditioning (HVAC) systems represent one of the most energy-intensive components in buildings, yet existing control strategies often fail to jointly optimize thermal comfort, indoor air quality (IAQ), and energy efficiency. This study presents a hybrid-action multi-agent deep deterministic policy gradient (MADDPG) framework for intelligent HVAC control, integrating discrete heater actuation and continuous airflow regulation within a unified reinforcement learning environment. The proposed architecture enables cooperative decision-making between specialized agents through centralized training and decentralized execution, thus capturing the mixed discrete-continuous nature of real HVAC operations. The agents interact with a physicsbased single-zone office model that captures coupled temperature-CO2 dynamics at 5-minute timesteps, driven by Ohio 2024-2025 cold-season outdoor temperature profiles and realistic occupancy schedules. MADDPG is trained for 200 episodes using experience replay, with actor-critic learning rates and target-update coefficients selected via a validation study over 12 hyperparameter configurations that maximized average episodic reward on a separate dataset. Final performance is evaluated on unseen November-February 2024-2025 weather and compared against a single-agent deep Q-network (DQN) baseline that controls only the heater under fixed airflow. Relative to DQN, MADDPG reduces total energy consumption by 7-10%, and discomfort hours by 38% on average across all months, while maintaining IAQ close to the threshold at the cost of modestly higher CO2 violations during extreme cold conditions. These results indicate that the hybrid-action multi-agent reinforcement learning is a promising pathway for energy-efficient, comfort-aware HVAC control in intelligent buildings.

Although simulation-based approaches have been widely applied to optimizing operations, existing studies are constrained to specific hierarchical levels or simulation techniques without addressing their interdependencies. This fragmentation has hindered the development of integrated frameworks capable of linking decisions from individual equipment control to site-level strategic planning. To bridge this gap, this review systematically analyzes peer-reviewed articles using a two-dimensional framework that classifies simulation-based approaches by hierarchical level (individual, activity, and process/site) and simulation process component (problem statement, simulation technique, modeling, data management, setup, analysis, and validation). The analysis identifies thirteen major challenges, ranging from computational limitations and coordination issues to integration difficulties across simulation techniques. Building on these findings, the paper proposes eleven research directions to establish a unified multilevel simulation architecture that enables bidirectional information flow and adaptive decision-making. The proposed framework provides theoretical and practical guidance for advancing simulation-based construction resource management that effectively reflects the complexity and dynamics of actual construction environments, contributing to the digital transformation and productivity improvement of the construction industry.

The increasing demand for healthy and comfortable living environments has driven significant advancements in smart home technology. However, current developments often overlook the importance of users’ social-emotional needs and the contextual dynamics within the home environment. This study proposes the integration of social robots as a promising solution to address these gaps. By enhancing smart home functionalities, social robots offer a more holistic approach to smart home design. A comprehensive literature review was conducted using the PRISMA methodology combined with large language model-based topic modeling to identify current research trends in social robotics. The analysis revealed five key research areas: (i) social-emotional intelligence, (ii) physical embodiment, (iii) elderly care, (iv) pediatric care, and (v) therapeutic applications. The study discusses how the core functionalities of social robots can enhance user experience by positively influencing the sensing, perception, and action layers of smart home systems. The findings suggest that the evolution of smart home technology should prioritize not only functional improvements but also the social and emotional well-being of users. Integrating social robots into smart homes will foster more human-centric, interactive, and satisfying living environments.

Design is the most fundamental element in the architecture, engineering, and construction industry, with building information modeling (BIM) and generative design gaining prominence for managing complex design tasks. However, design goals often overlook diverse environmental considerations such as human health and biodiversity, focusing mainly on carbon or greenhouse gas emissions. To address these issues, this study proposes a BIM-based framework for automated conceptual design integrating generative design and life cycle assessment. A case study on a single-detached house was conducted to validate the proposed framework. The total of fourteen cases were generated based on spatial connectivity and building material quantity through generative design, with case 12 selected as the optimal design using a 1:1 Pareto optimal solution. For case 12, the total cost was calculated at US$ 1.42 x 109, with economic costs comprising 1.76 % and environmental and human health costs accounting for 97.89 % and 0.72 %, respectively. The analyzed impacts from these categories are automatically visualized within the BIM model through the developed parametric algorithm, supporting architectural decision-making processes. The BIM-IGL framework developed in this study integrates BIM, generative design, and life cycle assessment to automate the evaluation of economic, environmental, and human health impacts during the conceptual design phase. By enabling comprehensive analysis up to the endpoint level in LCA, the framework supports designers in selecting optimized design solutions early in the design process.

Energy benchmarks play a critical role in reducing energy consumption in the building sector. To derive more specific and appropriate energy-related targets, room-level energy benchmarks are required. However, simply comparing the energy consumption of similar rooms does not account for the occupants and the characteristics of each individual room. Therefore, an analysis of individual room must be conducted beforehand. To this end, this study proposed a real-time individual room-level energy benchmark that reflects the heating and cooling characteristics of each room. Real-time individual room-level energy benchmark is calculated based on occupant related thermal characteristics and demand related thermal characteristics. Occupant related thermal characteristics pertain to the range of thermal environments that occupants in the room can accept and are analyzed through thermal preference votes. Demand related thermal characteristics refer to the possible energy consumption of the room based on heating and cooling demands and are analyzed using kernel density estimation on actual energy consumption data. A case study conducted on an office room confirmed its unique occupant related and demand related thermal characteristics. The study also confirmed that meeting the ideal thermal environment and energy consumption requirements proposed by the real-time individual room-level energy benchmark in this study could improve thermal comfort while saving energy by approximately 40 %. The proposed energy benchmark enables immediate responses to real-time occupants and environment changes and contributes to deriving specific and appropriate energyrelated targets for energy conservation.

Real-time monitoring of metabolic rate (MET) and clothing insulation (CLO) is essential to ensure effective occupant-centric control (OCC) for thermal comfort. This study aims to develop multi-task model using semisupervised learning to enhance occupant activities and clothes detection performance by utilizing both labeled and unlabeled data. The convolutional neural network-based model and training approach with pseudo labels to update all parameters comprehensively were proposed. The developed model is validated by conducting comparative analysis with state-of-the-art models and applying it in a real-world environment. The results demonstrate that the developed model, employing semi-supervised learning and the dual-phase training method (DPTM), achieves superior performance in activity and clothes detection outperforming previous studies with a 15.8 % higher mean Average Precision (mAP) for activity detection and a 25 % improvement for clothes detection. The findings highlight the potential of this multi-task model using semi-supervised learning to automate data collection improving the accuracy of estimating occupant thermal comfort. This approach can dynamically optimize indoor environments tailored to individual needs within the OCC framework, enhancing thermal comfort and energy efficiency through precise monitoring of occupant information.

Changes in the energy sector due to the COVID-19 pandemic provided insight into the future energy transition accompanied by technological advancements and social changes. Building sectors also analyzed the impact of schemes such as non-face to face education on building energy consumption, yet the effectiveness of the schemes themselves was omitted.Therefore, this study aims to determine the impact of non-face to face education, i.e. scheme, on building energy consumption and education in educational facilities. In this regard, non-face to face education ratio, energy consumption percentage change, and academic achievement were presented as indicators. Six years of information on non-face to face education and energy consumption for 30 high schools in Seoul, South Korea, as well as national test results of students nationwide are collected, respectively. Partial correlation analysis between scheme and energy consumption, scheme and academic achievement confirmed that scheme significantly reduced energy consumption and partially maintained academic achievement. Conversely, it is possible that energy consumption actually increased after scheme ended. Procedures and results contribute to the analysis and evaluation of the effectiveness of energy transition that can be expected from future technological progress and social change, and probable measures for the success of energy transition can be considered.

Recent studies have reported that occupants' physiological responses can be indicators of indoor environmental quality (IEQ). As a result, there is an emerging demand for devices to measure physiological responses, especially in wearable form. Previous reviews suggested and analyzed physiological responses that are affected by IEQ and corresponding measuring devices, but a full-scale review on the proper measuring method to give directions for future research and development is still necessary. In this regard, this study reviewed physiological response measuring methods to give directions to the improvement of existing wearable devices and the development of future wearable devices. Physiological responses related to IEQ were identified from review papers published over the last 10 years, and their measuring methods were classified based on locations of measurement, availability of wearables and existence of reference. Then each classification was analyzed and evaluated by four kinds of requirements for wearables in order to propose directions and guidelines for developing future wearable devices. This review is expected to be the guidelines to measure physiological responses for the future IEQ research and contribute to the development of the wearable devices that can be applied to monitoring or controlling IEQ in the future.

Previous studies have proven that it is hard for occupants to perceive concentration of indoor air pollution (IAP) and resulting indoor air quality (IAQ) on their own. Therefore, a method is needed to encourage them to turn their attention to actual IAP, in this context, alerting is thus suggested. However, previous studies pose limitations in that they failed to analyze the effects of alerting concentration of IAP on occupants' IAQ perception. To fill the research gap, this study sought to explore a proper strategy to help occupants have a clearer perception of IAQ. A one-month observational experiment was conducted on nine subjects under three scenarios with different alerting strategies. In addition, the visual distance estimation method was used to quantitatively analyze similar tendencies between the subject's perceived IAQ and concentration of IAP for each scenario. The experimental results confirmed that when an alerting notification was not sent, the occupants could not clearly perceive IAQ as the visual distance was the highest at 0.332. On the other hand, when the alerting notification whether the concentration of IAP exceeded the standard or not was sent, the occupants could perceive IAQ relatively clearly as the visual distance was reduced to 0.291 and 0.236. In conclusion, not only installing a monitoring device but also establishing proper alerting strategies on the concentration of IAP is essential to facilitate occupants' IAQ perception and protect occupants' health.

Despite a growing interest in indoor air quality (IAQ), a scientific basis for human self-awareness and understanding of IAQ has yet to be provided. To fill this research gap, this study examined the effect of numerical data through monitoring devices on the occupants' perception of IAQ. Therefore, an experiment was planned to evaluate the perceived IAQ based on a self-report questionnaire among ten subjects in an office environment. The experiment was divided into three sessions, with a classification model built for each session using automated machine learning and then derived variable importance for predictor variables. Results showed that when occupants did not check the numerical data of IAQ factors, they tended to perceive IAQ based on the perceived thermal comfort. In addition, occupants made a biased judgment of the current IAQ condition based on the past perceived IAQ checked 30 min ago. When occupants checked the numerical data of IAQ factors, they responded more sensitively to IAQ factors than to thermal environment factors when perceiving IAQ conditions. The results of this study confirmed that real-time numerical data of IAQ factors measured using monitoring devices are essential for occupants to perceive and judge the current IAQ clearly.

This study was conducted to ascertain how the use of recycled materials instead of virgin materials affect the installation of renewable energy systems for the energy transition of buildings. LCA was used to estimate the life cycle energy use and GHG emissions of building and renewable energy systems with and without using recycled materials. For the case study, a public building and four representative photovoltaic (PV) systems were selected. National recycling standards and processes were considered for suggested recycled materials. In general, replacing virgin material with recycled counterpart reduced embodied GHG emissions more than embodied energy. However, due to high carbon intensity of operation energy and building materials without available counterparts, using recycled materials reduced the life cycle energy use and GHG emissions of the case building by 4.9% and 3.3%. Among four PV systems, using recycled materials was most significant in the single crystalline-silicon (sC-Si) PV system, 44.5% reduction of energy use and 41.3% reduction of GHG emissions. Therefore, when recycled materials are used for achieving the energy transition of the case building with the sCSi PV system, the PV system requirements could be reduced by a maximum of 9.6% for energy payback and 4.9% for GHG payback. This study demonstrated that the use of recycled materials is effective at reducing the embodied energy and GHG emissions of PV systems as well as buildings. To achieve carbon-neutral buildings, particularly, the use of recycled materials based on a thorough LCA should be considered in addition to renewable energy system.

The emergence of the coronavirus disease 2019 (COVID-19) pandemic has led to a cost and time crisis for most construction projects around the world. Adherence to COVID-19 response guidelines for construction sites may prevent the occurrence of COVID-19 cases, which eliminates the risk of site closure. However, adding a disinfection process to the construction process, as mandated in COVID-19 response guidelines, increases overall construction cost and time. Conversely, however, not adding a disinfection process may result in COVID-19 cases among construction workers, which would delay construction and perhaps even cause closure of the construction site. Therefore, this study analyzed the feasibility of COVID-19 response guidelines for construction sites, especially the addition of a disinfection process, in terms of cost and time. To this end, CYClic Operations NEtwork (CYCLONE) models were developed to simulate the construction process, and a case study was conducted to validate the applicability of the suggested approach. The results showed that compliance with COVID-19 response guidelines increased the number of working days and the construction costs of the subject construction project, but because there was no more risk of construction site closure, the construction delays were short, and the liquidated damages were minimized. Through the method proposed in this study, it is possible to estimate construction cost and time before and after the COVID-19 pandemic; this method could be used to provide data for both owners and contractors to pro-actively recognize and respond to situations or damage caused by the COVID-19 pandemic. The data could also be used as evidence in case of future damages or disputes.

This paper aims to analyze the impact of recycled materials on energy reduction and clean energy transition of buildings.The life cycle energy of building using virgin materials is compared with that of building using recycled materials, and the energy savings due to recycled materials is calculated.Also, the impact of recycled materials on clean energy transition is verified by calculating the module area of the photovoltaic panel achieving the same amount of energy savings.As results of a case study, replacing virgin to recycled materials reduced 10% of life cycle energy.It means that the use of recycled materials can reduce 10% of module area of PV system for achieving clean energy transition.Recycled materials should be considered for fundamental energy reduction and effective clean energy transition.

Over the past decades, various tools that can perform building life cycle assessment (LCA) as well as life cycle cost (LCC) or CO2 analysis, have been developed. Even though these developed tools should be effectively evaluated and improved to encourage the continuous use of such tools, no research has been conducted on this matter. In this regard, this study sought to propose a framework for evaluating a building LCA tool from both the developer's and user's perspectives. In the developer evaluation process, experts evaluate if the design and implementation status (i.e., content status) are appropriate, and determine the design and implementation problems (i.e., content problems) based on six evaluation criteria, through content evaluation. In the user evaluation process, the users and evaluators determine the usability problems based on six usability attributes, through usability evaluation. The developer and user evaluation results are then interpreted through Satisfaction-Importance (S–I) analysis and Severity-Priority (S–P) analysis to prioritize the area of improvement and to determine the improvement strategy. To verify the proposed framework, a case study was conducted on an actual building LCA tool. The evaluation results showed that the problems corresponding to the assessment method and result should be preferentially improved in terms of content, while those corresponding to learnability, efficiency, and errors should be preferentially improved in terms of usability. Therefore, it is expected that the utilization of the proposed framework can effectively evaluate and improve various conventional building LCA tools in a reasonable way.

Indicators capable of reflecting the life cycle energy consumption and generation for both the building and renewable energy system are necessary to quantitatively evaluate substantive energy transition of the building since most of research or policies only focused on the operation energy of the building. In this regard, this study applied the energy payback lime to the building installed with the renewable energy system, and proposed "building driven-energy payback lime (BD-EPBT)." To evaluate the BD-EPBT of the building, a building applied with South Korea's energy transition policy was decided as a case study. With the annual energy generation (19.2 million kWh/yr) based on the current policy, the BD-EPBT was calculated at 53.9 years, which exceeded the building's lifetime (40 years), and the actual energy transition ratio was 22.2%, far less than the national objective (30%). In addition, the annual energy generation required to achieve a 30% energy transition target within the building's lifetime was estimated as 26.4 million kWh/yr, which is more than 137% of the current policy-based annual energy generation. Consequently, the energy transition of the building sector cannot be achieved by the current policies. Therefore, the BD-EPBT should be introduced for the substantial energy transition of the building with the renewable energy systems. This reduces the life cycle energy consumption of buildings and renewable energy systems and helps to effectively install the renewable energy systems.

This study proposes building driven-energy payback time (BD-EPBT) as a new life cycle assessment indicator.The BD-EPBT is the time required to outrun the total life cycle energy used in a power system-integrated building with annual surplus energy.As a result of a case study, it was shown that the BD-EPBT is highly associated with the embodied energy of construction materials.Also, by increasing the power system capacity (i.e., PV system) from 3 kW to 5 kW, BD-EPBT can be reduced by one fifth.The BD-EPBT is a novel concept capable of providing comprehensive analytics solutions to energy decision problems for building owners.

As people spend more time indoors, it is important to identify the relationship between indoor environmental quality (IEQ) and building occupants' health. This study aims to analyze building occupants' psychophysiological response to the indoor climate and CO2 concentration changes. While the indoor climate and CO2 concentration change, the study measured verbal scales (Indoor air quality (IAQ) satisfaction, thermal comfort vote (TCV), thermal satisfaction (TS), thermal sensation vote (TSV) and thermal preference (TP)) from 22 healthy subjects as psychological responses and monitored blood pressure (BP) at the seated state as the physiological response. The results are as follows: (i) IAQ satisfaction, TCV and TS had a negative correlation between -0.558 and -0.789 with BP; and (ii) if IAQ satisfaction, TCV and TS were below the neutral level, the systolic BP of some subjects was shown to exceed 140 mmHg, the hypertension warning state. The study is differentiated from previous studies as follows: (i) the verbal scale of the previously-used psychological responses was validated from the perspective of the psychophysiological approach (TCV approximate to TS not equal TSV); and (ii) for states with a high CO2 concentration, even if the operative temperature decreases from warm to neutral, the TCV and TS increase and then BP decreases as opposed to in previous studies. Namely, the unhealthier the IEQ condition is, the more the BP increased regardless of the change in the operative temperature. Through the results of this study, it is possible to implement a healthy indoor environment in both a psychological and physiological approach during the building operation phase.

Experimental analysis was conducted on the indoor air pollutant. concentration using natural ventilation and fillers. The study targeted two office rooms each of which was occupied by four people, and with the same outdoor environments. A non-woven fabric filter (room A) and an electrostatic filter (room B) were installed on the window frame, and the indoor air pollutant concentration and indoor climate factors were monitored based on the number of occupants and the occupants activities. The results are as follows: (i) when the number of occupants in each room increased from 0.03-006 to 1.53-1.63, room A showed a 60% average PM10 concentration increase while room B showed an opposite result (10% average PM10 concentration decrease), meaning the electrostatic filter's lower resistance to flow contributed to better ventilation and also decreased the influence of the occupants on the indoor air pollutant concentration. A low correlation (0.323-0.350) between the CO2 concentration and the occupants in room B also proved these results; (ii) while the average PM10 concentration in room A was 9 mu g/m(3) higher than that in room B, the average PM2.5 concentration in room A was higher by only 0.2 mu g/m(3), which showing that much of the generated or resuspended indoor particulate matter was PM10; and (iii) clue to the more frequent heal transfer from outdoors to indoors, room B consumed 23% more healing energy. The results of this study are expected to be used as bases for the establishment of an appropriate management strategy that:considers the indoor air pollutant concentration caused by the number of occupants and occupants' activities by combining natural ventilation and fillers. (C) 2018 Elsevier B.V. All rights reserved.