AI-Driven Digital Twins for Enhancing Indoor Environmental Quality and Energy Efficiency in Smart Building Systems

Abstract

Smart buildings equipped with diverse control systems serve the objectives of gathering data, optimizing energy efficiency (EE), and detecting and diagnosing faults, particularly in the domain of indoor environmental quality (IEQ). Digital twins (DTs) offering an environmentally sustainable solution for managing facilities and incorporated with artificial intelligence (AI) create opportunities for maintaining IEQ and optimizing EE. The purpose of this study is to assess the impact of AI-driven DTs on enhancing IEQ and EE in smart building systems (SBS). A scoping review was performed to establish the theoretical background about DTs, AI, IEQ, and SBS, semi-structured interviews were conducted with the specialists in the industry to obtain qualitative data, and quantitative data were gathered via a computerized self-administered questionnaire (CSAQ) survey, focusing on how DTs can improve IEQ and EE in SBS. The results indicate that the AI-driven DT enhances occupants’ comfort and energy-efficiency performance and enables decision-making on automatic fault detection and maintenance conditioning to improve buildings’ serviceability and IEQ in real time, in response to the key industrial needs in building energy management systems (BEMS) and interrogative and predictive analytics for maintenance. The integration of AI with DT presents a transformative approach to improving IEQ and EE in SBS. The practical implications of this advancement span across design, construction, AI, and policy domains, offering significant opportunities and challenges that need to be carefully considered.

1. Introduction

Buildings represent a critical sector in global greenhouse gas (GHG) mitigation efforts, accounting for approximately 40% of global energy consumption and serving as a major contributor to GHG emissions [1,2]. The United Nations Environment Programme (UNEP) Global Status Report for Buildings and Construction underscores the sector’s pivotal role in achieving sustainability objectives [3]. Enhancing energy management within buildings is essential for addressing climate-related challenges, particularly within the framework of the European Union’s Green Deal, which aims to achieve a 55% reduction in GHG emissions by 2030 [4,5]. These initiatives are closely aligned with the objectives of the Paris Agreement, which emphasizes the necessity of transformative energy management strategies on a global scale [6]. To effectively address these challenges, advanced intelligent systems capable of forecasting, optimizing, and dynamically regulating energy consumption have become indispensable [7,8]. Heating, ventilation, and air conditioning (HVAC) systems, which constitute a substantial proportion of a building’s overall energy demand, exemplify the necessity for such innovations [9,10]. A digital twin (DT) offers a high-fidelity virtual representation of a physical structure, integrating static data derived from building information modeling (BIM)—which provides a structured and detailed representation of a building’s geometry, materials, and systems [7,11,12]—with dynamic, real-time data acquired through internet of things (IoT) sensors monitoring parameters such as temperature, energy consumption, and occupancy patterns [13,14]. By incorporating artificial intelligence (AI) algorithms, DTs facilitate the simulation, analysis, and predictive modeling of a building’s energy performance, offering comprehensive insights into energy flows and potential inefficiencies [15]. These capabilities enable data-driven decision-making and the implementation of energy optimization strategies based on real-time analytics [16].

In modern society, the majority of people spend more than 80% of their daily lives indoors. Therefore, the quality of the indoor environment, known as indoor environmental quality (IEQ), plays a crucial role in influencing the physical and mental well-being of individuals [17]. Smart buildings equipped with diverse control systems serve the objectives of gathering data, optimizing energy efficiency, and detecting and diagnosing faults, particularly in the domain of IEQ. The development of energy-efficient (EE) buildings encompasses a broad spectrum of strategies, including advancements in insulation, the integration of EE technologies, and the optimization of heating, ventilation, and air conditioning (HVAC) systems. A primary focus within smart building systems (SBS) is enhancing indoor environmental quality (IEQ), as it is strongly correlated with occupant comfort and well-being. Optimizing the indoor environment through sustainable energy solutions aligns with the European Union’s energy efficiency objectives. Building energy management systems (BEMS) play a crucial role in optimizing energy consumption within buildings. These systems facilitate the integration of complex building infrastructures with sensor networks, enabling building owners and management teams to make data-driven decisions aimed at improving operational efficiency and reducing costs. Real-time energy monitoring serves as a fundamental component of BEMS, providing the necessary infrastructure for the deployment of advanced technologies such as DT and AI in building management. By enhancing energy efficiency, BEMS serve as a foundational technology for the evolution of SBS, contributing to both sustainability and operational performance improvements [10,18,19].

Asset information requirements (AIR) in buildings refer to the specific information needs associated with managing and maintaining building assets throughout their lifecycle [20]. A DT leverages this information by creating virtual replicas of physical assets, which include all relevant data outlined in AIR. This ensures that the DT is accurately representing the real-world asset, enabling better decision-making and performance optimization. The DT integrates various data sources, including those specified in AIR, into a unified digital model. This integration confirms that stakeholders have access to comprehensive information about the building’s assets, enabling them to monitor performance, conduct simulations, and make data-driven decisions based on the most up-to-date information. DTs enable stakeholders to visualize the impact of different scenarios on building performance, helping them identify opportunities for improvement and implement effective strategies to enhance efficiency, sustainability, and occupant comfort [21,22]. Integrating AI and DT offers efficient management of building systems and accurate prediction of outcomes. By analyzing historical and real-time data, AI can anticipate future conditions, enabling it to suggest optimal adjustments for factors like temperature and lighting. This recommendation is closely linked to occupancy patterns and external conditions. AI’s predictive capabilities are key to optimizing IEQ without increasing energy consumption. Emphasizing a proactive approach, the strategy aims to balance occupant comfort with EE. Collaboratively, AI and DT enhance the effectiveness of building operations, making significant contributions to environmentally sustainable practices and resource conservation initiatives [23].

Despite the rapid advancements in smart building technologies, a significant knowledge gap remains in understanding how AI and DT can synergistically improve IEQ while maintaining EE. This gap encompasses the development of comprehensive frameworks of robust integration strategies and scalable solutions that leverage the strengths of both technologies. This study differs from other DT research in several key ways, reflecting its focus, scope, and applications. The study addresses both the comfort of building occupants (IEQ) and the operational efficiency of the building (EE), recognizing the interdependence of these factors. The study also outlines practical implications for various stages of building development, from design and construction to ongoing management and policy implications. The study represents a holistic and multidisciplinary approach, combining insights from AI and data synergy, building design and construction, energy management, and policy studies. Hence, this study explores how AI and DT could work together to improve IEQ in energy-efficient SBS. Using a mixed method (qualitative and quantitative) of approach, the study investigates the fundamental principles supporting AI and DT, analyzing their roles in the SBS. The research intends to study how combining these technologies could create a robust system for detecting faults and predicting maintenance needs, ultimately enhancing SBS. The following research questions are addressed in this study:

i.

What are the AIR needed to create AIM by integrating DT and enabling real-time insights for predictive analytics?

ii.

How can DT enhance IEQ and energy-efficient performance of SBS?

iii.

How can integrating DT with AI be used for decision-making in automatic fault detection, predictive maintenance conditioning, and optimized energy consumption?

The research questions for this study were developed after a thorough examination of current literature in the relevant field to identify common patterns, topics, and gaps. These questions were carefully refined using insights and viewpoints collected from interviews and a survey, which was the foundation of the research approach for this study. This integrated method improved the accuracy and clarity of the study questions and strengthened their importance and detail. The study’s organizational structure is systematically outlined as follows: Section 2 discusses the theoretical basis, Section 3 describes the methodology, Section 4 shows the results, and Section 5 offers a thorough analysis of the implications, limits, and suggestions for further research. The study concludes in Section 6.

2. Theoretical Background

2.1. Indoor Environmental Quality (IEQ)

IEQ encompasses factors that enhance the comfort, health, and well-being of individuals residing in a building. These factors include ergonomics, air quality, acoustics, lighting, and thermal comfort, all of which are vital for creating a favorable and conducive living environment [24]. Therefore, ensuring high IEQ levels directly impacts the health, productivity, and satisfaction of building occupants, making it an integral aspect of modern building design practices.

Among the components of IEQ, indoor air quality (IAQ) holds particular significance. Inadequate IAQ poses health risks to occupants, potentially leading to respiratory illnesses and decreased cognitive function [25]. Maintaining IAQ involves various measures such as air filtration, ventilation, and pollutant control [26]. Similarly, thermal comfort is crucial for occupants’ well-being, necessitating proper temperature, humidity regulation, and sustainable HVAC systems for optimal conditions [27]. Adequate lighting also plays a vital role in enhancing productivity and aesthetics within a building [28]. Therefore, an effective IEQ strategy should incorporate energy-efficient systems to improve building aesthetics, occupant productivity, and energy conservation. Despite the critical role of IEQ, there remains a notable gap in understanding and implementing IEQ approaches [29]. This gap stems from factors such as limited research, inconsistent standards, and inadequate awareness regarding IEQ. A significant challenge in addressing IEQ deficiencies lies in the absence of standardized performance metrics. The lack of universally accepted guidelines makes it challenging for stakeholders to effectively implement IEQ measures, resulting in substandard buildings [29]. To address this issue, increased research efforts are needed to identify gaps, develop solutions, and enhance the availability of information for stakeholders such as contractors, architects, facility managers, and engineers [30]. Additionally, global collaboration among stakeholders is essential for establishing standardized protocols and guidelines for building IEQ. IEQ policies should be supported by comprehensive research to inform stakeholders about the critical importance of optimal IEQ in enhancing occupant health, comfort, and productivity. According to Shrubsole et al. [30], it is important to prioritize environmental sustainability and occupant well-being by combining IEQ with BEMS in building settings. Opoku et al. [31] developed a DT platform for a university library representing the “Living Lab” concept for monitoring the IEQ. ElArwady [32] proposed a platform that integrates the DT concept with IoT and BIM technologies for indoor thermal comfort.

2.2. Building Energy Management Systems (BEMS)

Building Energy Management Systems (BEMS) are designed to continuously monitor the state of a building and regulate HVAC systems to ensure optimal energy efficiency while maintaining occupant comfort [27]. These systems collect data from various building components to support the implementation of effective control strategies. A significant challenge associated with BEMS lies in handling the vast volumes of data generated, which must be accurately interpreted by building managers to enable informed decision-making. The ability to process and analyze this continuously expanding dataset is essential for maintaining efficient building operations. This underscores the necessity for advanced methodologies to enhance energy management and IEQ, either through the development of novel solutions or by integrating BEMS with complementary technologies. Ghansah [33] identified that efficient building energy management is one of the significant categories of DT application for smart buildings at the facility management stage. BEMS are designed to regulate, oversee, and enhance a building’s energy consumption through the integration of hardware and software technologies. According to Kozlovska et al. [18], BEMS harness diverse data inputs from HVAC systems, lighting, temperature, and power usage, gathered by sensors, to offer a comprehensive understanding of energy utilization within the building. This data collection facilitates informed decision-making for building management and owners by pinpointing opportunities for energy conservation and addressing inefficiencies. BEMS empower building operators to meticulously monitor energy consumption, establish energy usage benchmarks, and set targets to maximize EE [34]. Moreover, effectively employed, BEMS can avert costly breakdowns and downtimes by analyzing data patterns to identify potential equipment failures in a timely manner.

Additionally, BEMS play a crucial role in regulating various building systems to optimize energy-related functions automatically, such as adjusting temperatures, lighting levels, and HVAC settings while ensuring occupant comfort [18,19]. By automating numerous building processes, BEMS enhance operational efficiency and reduce the need for constant manual oversight. Through strategic scheduling of equipment activation based on predicted occupancy patterns and external weather conditions, these systems significantly contribute to energy-saving initiatives.

BEMS incorporate advanced real-time monitoring and reporting mechanisms that deliver timely and actionable feedback to building operators and management. This sophisticated feature fosters a proactive approach to energy management, ensuring that energy consumption is optimized and aligned with sustainability objectives. These reports may include information on energy usage, realized savings, system performance, and recommendations for enhancements. The effectiveness of BEMS has resulted in their widespread adoption across various sectors, including residential, industrial, and commercial buildings [18]. Focusing on the building’s air conditioning systems due to their significant impact on overall energy consumption, Elnour et al. [35] adopted a broad approach that also included other energy systems. They provide a comparative evaluation of simulation-based and data-driven modeling in a case study (QU Soprts and event building) while exploring various ML algorithms. Additionally, they examine key smart applications of the building’s data-driven DT. Renganayagalu [36] explored the development and implementation of a data-driven digital twin for an office building in Norway, with a primary focus on predicting energy performance. Cespedes-Cubides [37] investigated the implementation of DT technology in building operations and maintenance (O&M), with a particular focus on enhancing energy efficiency throughout the building’s lifecycle. Karatzas [38] proposed a DT-enabled framework that incorporates social dynamics into heritage building management, promoting sustainable practices that balance conservation requirements with energy efficiency in a comprehensive manner.

2.3. Asset Information Requirements (AIR)

Asset information requirements (AIR) are crucial for overseeing and maintaining physical assets across various business sectors. These requirements outline the essential data and details necessary to uphold an asset’s peak performance and extend its lifespan. By providing stakeholders with comprehensive insights, AIR enable informed decision-making regarding assets, operations, and maintenance strategies, utilizing a structured framework [20]. They offer a comprehensive overview of an asset, encompassing its specifications, standards, design, condition, operations, and regulatory compliance. This systematic approach enhances asset management (AM) procedures, ensuring efficient resource utilization and fostering operational excellence and strategic planning. AIR serve as a foundational framework for the data, information, and documentation needed to support assets throughout their lifecycle. Desbalo et al. [39] proposed a conceptual framework aimed at enhancing the performance of built assets while supporting owners, end-users, and managers in defining data and information requirements for BIM-enabled asset information management. In AM, DTs offer asset managers reliable, real-time records of real estate data [40].

DTs play a transformative role in asset lifecycle management (ALM), as noted by [41,42]. Asset information modeling (AIM) is indispensable for asset owners, containing all the necessary data and documentation to sustain an asset. AIM goes beyond static data by integrating DT, dynamic virtual representations of physical assets, facilitating real-time insights and predictive analytics [22]. Through AIM, DT provides a comprehensive view of assets by amalgamating current conditions with historical operational data, thereby enhancing decision-making capabilities and streamlining cost-effective AM [41].

Efficient data management in asset operations relies on AIR, enabling the DT to generate interactive virtual models. These models offer predictive analytics and real-time monitoring, thereby enhancing asset performance and decision-making effectiveness. Yeom et al. [43] proposed integrated DT and extended reality (XR) solutions for building energy management and occupant comfort, stressing the importance of fidelity, interoperability, and data security. The subsequent section will delve into a detailed analysis of DT, exploring its utilization and benefits within AM frameworks.

2.4. Digital Twins (DTs)

The creation of a DT commences with meticulous data collection from various sources, including sensors, internet of things (IoT) devices, computer-aided design (CAD) models, historical archives, and manual inputs. These data are then amalgamated into a unified platform or system. Subsequently, a virtual representation is fashioned to accurately mirror the physical entity or system. Advanced modeling techniques such as three-dimensional modeling, physics-based modeling, and machine learning (ML) algorithms are employed to construct a lifelike DT. The simulation capabilities of DTs enable the replication of real-world behaviors under different scenarios, enhancing predictive accuracy and operational efficiency [44,45].

DTs continuously receive real-time data from their physical counterparts through sensors and monitoring tools, evaluating the performance, status, and condition of the physical asset. Advanced analytics methods such as machine learning and AI algorithms may be utilized to extract insights and detect anomalies. By analyzing both historical and current data, DTs can anticipate potential malfunctions or performance deterioration, thereby reducing downtime and enhancing efficiency through proactive maintenance and optimization strategies. By providing a unified platform for data exchange and analysis, DTs facilitate collaboration among engineers, operators, and managers, enabling effective decision-making and problem-solving throughout the asset or system’s lifecycle [46,47,48,49,50,51,52]. Bortolini et al. [53] conducted a thorough analysis of the applications of DTs in enhancing building energy efficiency.

The adoption of DT technology has significantly enhanced the execution of construction projects, resulting in cost savings and reduced completion times. DT technology is increasingly recognized for its transformative potential, particularly in cutting-edge domains like smart building design and construction. This innovative approach offers numerous benefits, including accelerated construction schedules, substantial cost reductions, enhanced productivity, facilitated collaboration, optimized architectural designs, and advancements in safety and sustainability practices [23,44,53].

The applications of DTs encompass a diverse range of building types, including both historical and modern structures [54,55]. Tahmasebinia [56] examines the application of DT technology and ML modules across various building types and energy challenges by analyzing multiple real-world examples. Ni et al. [57,58] conducted two studies on the utilization of DTs for managing historical buildings in Sweden. The first study integrated sensor data, historical records, and machine learning algorithms to enhance energy efficiency, building preservation, and occupant comfort in three historical buildings [57]. The second study introduced an IoT-based DT designed to improve the preventive conservation of a theater in Norrköping, Sweden. Wang et al. [59] proposed a DT framework aimed at optimizing green building maintenance and automating management processes.

Exploring the capabilities of DTs reveals the potential integration with AI, ushering in a new era of synergy that enhances AM effectiveness. The subsequent section will delve into the pivotal role of AI in AM, analyzing its impact on operational efficiency and decision-making processes.

2.5. Artificial Intelligence (AI) in Smart Building Systems (SBS)

AI is driving substantial transformation within the construction industry, significantly boosting operational efficiency and streamlining processes. Utilizing AI technologies like data analytics, machine learning, and predictive modeling, several facets of building maintenance and management are being enhanced. These advancements notably elevate fault detection mechanisms in energy-efficient building systems, employing a range of technologies and methodologies to elevate operational standards in the sector [16,60].

By analyzing large volumes of data generated by sensors and IoT devices installed in smart buildings, AI algorithms can identify abnormalities or deviations from typical patterns, indicating flaws or inefficiencies in building systems in real time. Abdelalim et al. [61] investigated the transformative integration of AI and DT technologies within BIM frameworks, leveraging IoT sensors for real-time data collection and predictive analytics. AI-driven predictive maintenance systems utilize sensor readings, historical data, and environmental factors to anticipate potential faults or breakdowns in building equipment, thereby averting costly malfunctions and enhancing the overall reliability of building systems by preempting issues before they arise. AI algorithms swiftly identify patterns associated with common faults or inefficiencies in construction systems, learning to detect errors more promptly and accurately by recognizing subtle symptoms of a problem that human operators may overlook through training on labeled data. AI-based fault detection systems analyze sensor data to pinpoint the underlying causes of faults or flaws in building systems, offering insights into the root issues impacting system performance by comparing sensor readings with established failure patterns. AI optimization algorithms continuously adjust building systems to enhance EE while reducing flaws and inefficiencies, dynamically regulating HVAC, lighting, and other systems based on real-time data and environmental conditions [62,63].

AI’s intelligent automation plays a pivotal role in promptly adjusting HVAC systems, lighting, and other building functions, optimizing performance and minimizing energy consumption through real-time data analysis and environmental considerations [62,63]. AI’s incorporation with DT represents a significant advancement, where AI algorithms autonomously analyze extensive datasets, predict asset performance, and enhance overall system effectiveness. The ensuing discussion will elucidate the synergistic relationship between AI and DT, underscoring their combined efficacy in improving operational efficiency and fostering innovation in the architectural domain.

3. Materials and Methods

Mixed methods offer a valuable approach to expanding the scope of research by incorporating multiple methodologies for different research aspects [64]. This study will utilize a scoping review to provide a comprehensive overview of existing research literature, while data pertinent to the research context will be gathered through questionnaire surveys and interviews. This approach enables a thorough investigation into the application of AI-driven DTs for fault detection in energy-efficient smart buildings. An outline of the research methodology is depicted in Figure 1.

3.1. Scoping Review

A scoping review involves analyzing existing literature to identify and map out published work related to a specific topic or question. It is a valuable method for exploring the breadth of a subject, identifying gaps in the literature, and determining the feasibility of conducting a systematic review [65]. The literature search was conducted using Scopus and Web of Science databases, with searches performed using keywords, abstracts, and titles. Data collection followed a systematic approach using predefined keywords, with publications from 2019 to 2025 and language restricted to English. The literature search process is illustrated in Figure 2, following the preferred reporting items for systematic reviews and meta-analysis (PRISMA) flow diagram. This flowchart outlines the systematic steps for identifying and retrieving relevant studies, ensuring transparency and adherence to established research standards. Using this approach ensures systematic reporting in the selection and inclusion of relevant data for this investigation. The search terms utilized were as follows:

-

Code 1# = TITLE-ABS-KEY (“digital twin*” OR “virtual twin*”) AND (“AI” OR “artificial intelligence”) AND energy AND building*)

-

Code 2# = TITLE-ABS-KEY (“digital twin*” OR “virtual twin*”) AND “energy efficien*” AND “smart buil*”)

During the initial literature search, a total of 151 publications were identified from the Scopus (n = 85) and Web of Science (n = 66) databases. Following a pre-screening process, 46 duplicate records were excluded. Subsequently, upon closer examination of titles and abstracts, 24 publications unrelated to the specified sector were removed. Five additional articles were discarded after a thorough evaluation. Furthermore, 28 articles were eliminated, as they did not focus on smart building systems. Consequently, the review comprised 48 pertinent papers, systematically shown in Figure 2.

3.2. Sampling and Data Collection

3.2.1. Interviews

Participant selection in the inquiry was carefully chosen to match with the project’s objectives, focusing on professional jobs relevant to the research issues. The criteria for selection were expanded to consider more than just professional relevance, now including a thorough assessment of the interviewees’ competence and competency. This was carried out to make sure that people with the required level of knowledge and expertise were included to give meaningful replies to the question. All participants were required to give consent before digital recordings and transcriptions of the conversations were made using video-mediated communication. To guide the discourse, an interview protocol was meticulously developed, comprising a series of semi-structured questions designed to canvass a spectrum of relevant topics. The semi-structured interview questions as listed in Appendix A.1 included how AI-driven DT enhances IEQ, EE, and asset management but faces challenges in adoption, data integration, operational efficiency, and sustainability. This method also allowed the freedom to investigate emerging issues throughout the discussions. The qualitative data obtained from the interviews were analyzed systematically using NVivo, a specialist program for qualitative data analysis, enabling a detailed study of the compiled material. The demographic features of the participants that are pertinent to the analysis are detailed in

Table 1. Demographic and professional profile of interviewees.

No	Position (Role)	Type of Organization	Years of Experience	Company Size *	Operating Region
1	Application Manager	IT Company	4	Large	France
2	Process Manager	Consultancy Firm	4	Medium	Sweden
3	BIM Manager	Consultancy Firm	6	Large	Finland
4	Valuer	Consultancy Firm	3	Large	Sweden
5	Consultant	Consultancy Firm	8	Large	Sweden
6	Digitalization Lead	Energy Company	14	Large	Netherlands
7	Digital Twin	Manufacturing Company	14	Large	United States
8	Digital Twin Consultant	Consultancy Firm	17	Small	India
9	Manufacturing Engineer	IT Company	3	Small	India
10	BIM Modeler	Consultancy Firm	4	Medium	United States

Note: * Small: 0–50 employees; medium: 50–250 employees; large: 250+ employees.

of the supplementary documentation.

3.2.2. Questionnaire Survey

This study has implemented a systematic collection of primary data utilizing a computerized self-administered questionnaire (CSAQ). This survey was designed based on the results of a scoping review to enhance the strength of the study findings. Participants were intentionally chosen from a variety of functions and positions in different businesses to thoroughly investigate the issue. The study specifically targeted people working in the industry sector, while intentionally omitting those in academic professions. The survey questionnaire was distributed strategically to specialists in the building sector from several countries, specializing in BIM, AI, or DT technologies. This method aimed to gather a diverse range of viewpoints and understandings from various geographical locations. The sample’s representativeness may be limited despite efforts to include specialists with varying degrees of experience and skill. The questionnaire was divided into sections, each concentrating on a particular issue pertinent to the study’s goals. Among the sections was the first part, which covered demographic information where the participants’ educational background, the organization’s size, and the location in which the building industry operates were recorded. In the second section, the participants were asked about the advantages, challenges, and potential impacts of incorporating AI and DT technologies into their professional practice. The questionnaire results were analyzed using Statistical Package for the Social Sciences (SPSS), a statistical analysis software. Descriptive statistics involving mean, median, standard deviation, and associational statistics, including Spearman’s rank correlation, and reliability (Cronbach α), were measured. Due to the Likert-type data’s ordinal nature and the absence of normality assumption, Spearman’s rank correlation coefficient was employed to compute the correlation [66].

The survey was distributed to a specific group of 431 people using professional networking channels. 85 people engaged in the outreach, resulting in a response percentage of 19.7%. As the study aimed at exploratory research, and the target population was relatively small and homogeneous (e.g., professionals in a specialized field), a power analysis (typically aiming for 80% power and α = 0.05) was conducted to determine whether the sample size of n = 85 was sufficient to detect a statistically significant effect in the study. Power analysis results show that sample size n = 85 is acceptable, as seen in Appendix A.3.

Table 2. Demographic distribution of survey respondents by company type, size, and region.

Company Type	Engineering Consultancy	Real Estate Agency	Construction Contractor	Facility Management	BIM Solutions	Energy
Role—%	BIM Coordinator—16%	BIM Modeler—10%	Project Manager—8%	Facility Manager—10%	BIM Developer—8%	Industrial Engineer—4%
	Building Engineer—8%	Architect—8%	BIM Manager—12%	BIM Specialist—10%	Digital Development Lead—6%
Company Size and %
Small (<50)	9%	8%	6%	3%	3%	1%
Medium (50–250)	8%	9%	7%	3%	2%	2%
Large (>250)	11%	8%	7%	6%	5%	2%
Operating Region and %
Scandinavia	3%	2%	2%	2%	1%	1%
Europe	7%	6%	5%	5%	3%	2%
N. America	3%	3%	2%	2%	1%	1%
S. America	1%	1%	1%	1%	0%	0%
Australia	1%	1%	1%	1%	0%	0%
Asia	6%	5%	5%	5%	3%	1%
Middle East	3%	3%	3%	2%	1%	1%
Africa	1%	1%	1%	0%	0%	0%

outlines the demographic features of the respondents.

4. Results

4.1. Thematic Analysis

To conduct a thematic analysis using Braun and Clarke’s [67] six steps for the themes and patterns provided within the data were systematically identified and analyzed as seen in Table 3.

Table 3. Six-stage process in thematic analysis [67].

Step	Description
Step 1: Familiarization with the Data	Reading and re-reading data related to AI-driven digital twins (DTs) for enhancing indoor environmental quality (IEQ) and energy efficiency (EE) in smart building systems (SBS).
Step 2: Generating Initial Codes	Identifying key features within the data and creating specific codes related to IEQ and EE via DT.
Codes	“Energy optimization”, “Improved indoor comfort”, “Sensor data utilization”.
	“Real-time monitoring”, “Operational efficiency”, “Decision-making enhancement”.
	“Asset data requirements”, “Operational efficiency”, “Data integration”.
	“Fault detection”, “Lifecycle management”, “Asset optimization”.
	“Data governance issues”, “Legacy systems”, “DT awareness”.
	“Sustainability enhancement”, “AI-driven maintenance”, “Resource optimization”.
	“Predictive analytics”, “Energy consumption forecasting”, “Data-driven decisions”.
	“Energy use reduction”, “Building systems optimization”, “AI-driven controls”.
	“Data quality issues”, “AI literacy”, “Validation challenges”.
Step 3: Searching for Themes	Grouping initial codes into broader themes that reflect significant patterns in the data.
	Theme 1: Advancement of IEQ and EE via DT (how DT contributes to improving indoor environments and optimizing energy use).
	Theme 2: Real-time Operational Benefits of DT (operational advantages of DT, including improved decision-making and efficiency).
	Theme 3: Requisite Asset Information for Operational Efficiency (identifying necessary asset information for efficient building management).
	Theme 4: DT in Asset Management Efficiency (how DT enhances asset management, fault detection, and lifecycle management).
	Theme 5: DT Adoption: Organizational Challenges (barriers to DT adoption, including data governance and legacy issues).
	Theme 6: AI and Sustainable Facility Management (the role of AI in sustainable facility management).
	Theme 7: AI in Decision-Making and Predictive Optimization (AI’s role in predictive maintenance and energy optimization).
	Theme 8: AI’s Influence on Building Energy Consumption Optimization (AI’s impact on reducing energy consumption).
	Theme 9: AI Implementation: Organizational Challenges (challenges in AI adoption, such as data quality and system validation).
Step 4: Reviewing Themes	Refining themes to ensure they accurately reflect the data. Merging, splitting, or redefining themes as needed.
Step 5: Defining and Naming Themes	Clearly defining and naming each theme to capture its essence.
Step 6: Producing the Report	Compiling the analysis into a report explaining each theme, how they were identified, and their significance in BIM-based sustainability assessment and the broader application of DTs and AI in the construction industry.

These themes are delineated in Table 4, which provides a concise summary of the key concepts and challenges articulated during the interviews.

Table 3. Six-stage process in thematic analysis [67].

Step	Description
Step 1: Familiarization with the Data	Reading and re-reading data related to AI-driven digital twins (DTs) for enhancing indoor environmental quality (IEQ) and energy efficiency (EE) in smart building systems (SBS).
Step 2: Generating Initial Codes	Identifying key features within the data and creating specific codes related to IEQ and EE via DT.
Codes	“Energy optimization”, “Improved indoor comfort”, “Sensor data utilization”.
	“Real-time monitoring”, “Operational efficiency”, “Decision-making enhancement”.
	“Asset data requirements”, “Operational efficiency”, “Data integration”.
	“Fault detection”, “Lifecycle management”, “Asset optimization”.
	“Data governance issues”, “Legacy systems”, “DT awareness”.
	“Sustainability enhancement”, “AI-driven maintenance”, “Resource optimization”.
	“Predictive analytics”, “Energy consumption forecasting”, “Data-driven decisions”.
	“Energy use reduction”, “Building systems optimization”, “AI-driven controls”.
	“Data quality issues”, “AI literacy”, “Validation challenges”.
Step 3: Searching for Themes	Grouping initial codes into broader themes that reflect significant patterns in the data.
	Theme 1: Advancement of IEQ and EE via DT (how DT contributes to improving indoor environments and optimizing energy use).
	Theme 2: Real-time Operational Benefits of DT (operational advantages of DT, including improved decision-making and efficiency).
	Theme 3: Requisite Asset Information for Operational Efficiency (identifying necessary asset information for efficient building management).
	Theme 4: DT in Asset Management Efficiency (how DT enhances asset management, fault detection, and lifecycle management).
	Theme 5: DT Adoption: Organizational Challenges (barriers to DT adoption, including data governance and legacy issues).
	Theme 6: AI and Sustainable Facility Management (the role of AI in sustainable facility management).
	Theme 7: AI in Decision-Making and Predictive Optimization (AI’s role in predictive maintenance and energy optimization).
	Theme 8: AI’s Influence on Building Energy Consumption Optimization (AI’s impact on reducing energy consumption).
	Theme 9: AI Implementation: Organizational Challenges (challenges in AI adoption, such as data quality and system validation).
Step 4: Reviewing Themes	Refining themes to ensure they accurately reflect the data. Merging, splitting, or redefining themes as needed.
Step 5: Defining and Naming Themes	Clearly defining and naming each theme to capture its essence.
Step 6: Producing the Report	Compiling the analysis into a report explaining each theme, how they were identified, and their significance in BIM-based sustainability assessment and the broader application of DTs and AI in the construction industry.

These themes are delineated in Table 4, which provides a concise summary of the key concepts and challenges articulated during the interviews.

Table 4. Results of thematic analysis from qualitative interviews.

Theme	Description
Advancement of IEQ and EE via DT	DTs offer a sophisticated control system, enabling facility managers to gauge and enhance occupant comfort via adaptive tuning of mechanical and electrical systems.
	Utilization of real-time sensor data through DTs allows for the regulation of indoor temperature, air quality, and optimal building usage.
	DTs are instrumental in realizing maximal operational efficiency and in the reduction of anticipated costs and carbon emissions of buildings.
Real-time Operational Benefits of DT	DTs facilitate the acquisition of data for predictive analyses and asset optimization.
	Continuous, data-informed decision-making is enabled through the application of DT technology.
	DTs permit the simulation of future scenarios, aiding in the comprehension of asset lifecycle events across various demographics.
Requisite Asset Information for Operational Efficiency	The necessity for precise and structured data is emphasized to avert mismanagement and energy loss.
	The integration of BIM models with sensory input is critical for informed decision-making and automatic fault detection.
	Precise data acquisition is fundamental for effective predictive maintenance and conditioning.
DT in Asset Management Efficiency	DTs offer clarity and direct fault detection capabilities for asset managers, thus streamlining the asset management process.
DT Adoption: Organizational Challenges	Challenges include data governance dilemmas, insufficient records for legacy buildings, and a deficit in DT awareness among key industry stakeholders.
AI and Sustainable Facility Management	AI’s predictive capabilities based on historical data contribute to energy consumption forecasts and indoor climate regulation.
	AI’s data-driven control of building systems promises enhancements in sustainability via reduced energy usage.
	AI’s application in industrial facilities coordinates complex logistics, diminishing the reliance on manual processes.
AI in Decision-Making and Predictive Optimization	AI is adept at fault detection by assessing asset lifespans and environmental conditions to inform maintenance strategies.
	AI contributes to organizational process improvement by identifying systematic inefficiencies.
	AI facilitates time-saving data organization for enhanced big-data analytics.
AI’s Influence on Building Energy Consumption Optimization	AI is leveraged to autonomously refine HVAC systems, taking into account weather patterns and energy tariffs.
	AI dynamically adjusts environmental controls in response to occupancy fluctuations, maintaining comfort and EE.
	In high-occupancy buildings, AI modulates humidity and temperature to cater to transient populations effectively.
AI Implementation: Organizational Challenges	Challenges encompass the verification and validation of AI, the scarcity and disorganization of data, and a lack of AI literacy and precedent examples.
	The quest for high-quality data remains a challenge, highlighting the repercussions of unreliable data on decision-making outcomes

Table 4. Results of thematic analysis from qualitative interviews.

Theme	Description
Advancement of IEQ and EE via DT	DTs offer a sophisticated control system, enabling facility managers to gauge and enhance occupant comfort via adaptive tuning of mechanical and electrical systems.
	Utilization of real-time sensor data through DTs allows for the regulation of indoor temperature, air quality, and optimal building usage.
	DTs are instrumental in realizing maximal operational efficiency and in the reduction of anticipated costs and carbon emissions of buildings.
Real-time Operational Benefits of DT	DTs facilitate the acquisition of data for predictive analyses and asset optimization.
	Continuous, data-informed decision-making is enabled through the application of DT technology.
	DTs permit the simulation of future scenarios, aiding in the comprehension of asset lifecycle events across various demographics.
Requisite Asset Information for Operational Efficiency	The necessity for precise and structured data is emphasized to avert mismanagement and energy loss.
	The integration of BIM models with sensory input is critical for informed decision-making and automatic fault detection.
	Precise data acquisition is fundamental for effective predictive maintenance and conditioning.
DT in Asset Management Efficiency	DTs offer clarity and direct fault detection capabilities for asset managers, thus streamlining the asset management process.
DT Adoption: Organizational Challenges	Challenges include data governance dilemmas, insufficient records for legacy buildings, and a deficit in DT awareness among key industry stakeholders.
AI and Sustainable Facility Management	AI’s predictive capabilities based on historical data contribute to energy consumption forecasts and indoor climate regulation.
	AI’s data-driven control of building systems promises enhancements in sustainability via reduced energy usage.
	AI’s application in industrial facilities coordinates complex logistics, diminishing the reliance on manual processes.
AI in Decision-Making and Predictive Optimization	AI is adept at fault detection by assessing asset lifespans and environmental conditions to inform maintenance strategies.
	AI contributes to organizational process improvement by identifying systematic inefficiencies.
	AI facilitates time-saving data organization for enhanced big-data analytics.
AI’s Influence on Building Energy Consumption Optimization	AI is leveraged to autonomously refine HVAC systems, taking into account weather patterns and energy tariffs.
	AI dynamically adjusts environmental controls in response to occupancy fluctuations, maintaining comfort and EE.
	In high-occupancy buildings, AI modulates humidity and temperature to cater to transient populations effectively.
AI Implementation: Organizational Challenges	Challenges encompass the verification and validation of AI, the scarcity and disorganization of data, and a lack of AI literacy and precedent examples.
	The quest for high-quality data remains a challenge, highlighting the repercussions of unreliable data on decision-making outcomes

4.2. Statistical Analysis

4.2.1. Descriptive Statistics

Descriptive statistics involve the use and analysis of quantitative measures (mean, median, and standard deviation) to summarize and interpret data from a given sample [68]. Descriptive statistics reporting the mean and median values and standard deviations of questionnaire responses are presented in

Table 5. Descriptive statistics, factor analysis, and reliability test.

Questionnaire Statement	Mean 1	Med 2	Stan Dev 3	Factor Loading	Cronbach α	Rank
Digital Twin (DT)
A digital twin can display the real-time status and conditions of an asset	4.04	4.0	1.023	0.812		10
Digital twins can help predict when maintenance is required	4.02	4.0	1.012	0.794		13
Digital twins can identify issues regarding the condition of assets quickly and accurately	3.93	4.0	0.967	0.833	0.870	8
Using digital twins can enhance the quality of indoors. Environments	4.12	4.0	0.929	0.781		14
Using digital twins can efficiently improve energy usage and promote better environmental conditions	4.30	5.0	0.920	0.840		7
Asset Information Requirement (AIR)
Asset information requirements support the creation of asset information modeling in digital twins	4.02	4.0	0.922	0.862		4
Asset information modeling (AIM) facilitates asset management throughout operation and maintenance faces	3.89	4.0	0.975	0.882		2
It is essential to have precise and current asset information to ensure efficient indoor environments and quality	3.95	4.0	0.980	0.800	0.898	12
Monitoring the condition of energy-related assets can help optimize energy consumption, leading to reduced energy costs and a smaller carbon footprint	4.09	4.0	0.996	0.846		6
Accurate information about asset conditions permits improved operational efficiency in terms of comfort, quality, and energy consumption	4.09	4.0	1.009	0.824		9
Artificial Intelligence (AI)
AI can improve efficiency and sustainability in monitoring and controlling assets in building facilities	4.23	4.0	0.915	0.860		5
AI contributes to developing smart buildings with different control systems for efficient optimization, and fault detection, specifically in HVAC systems	4.14	4.0	0.939	0.807		11
AI helps decision-making for the operation of assets in building facilities by enabling the integration and analysis of raw data	4.33	5.0	0.861	0.902	0.885	1
AI can detect potential faults and predict for optimization of future operations by analyzing historical data	4.23	5.0	1.016	0.878		3
AI can automate HVAC, indoor air quality, and other building systems to optimize energy consumption	4.27	5.0	0.930	0.695		15

. The summarized statistics specify interesting comprehensions as an overview of the building industry’s perception of the impact of AI-driven DTs on enhancing IEQ and EE in SBS. According to the results, the mean scores for 12 out of the 15 questions were higher than 4.00 out of 5.00. The proposed model’s overall mean rating was 4.11, which means that industry professionals support the AI-driven DT enhancing occupants’ comfort and energy-efficiency performance and enabling decision-making on automatic fault detection and maintenance conditioning to improve buildings’ serviceability and IEQ in real time responding to the key industrial needs in BEMS, and interrogative and predictive analytics for maintenance.

The argument that sought respondents’ opinions on AI helping decision-making for the operation of assets in building facilities by enabling the integration and analysis of raw data had the highest mean value of 4.33 in the relative importance of the variables. “Using DT can efficiently improve energy usage and promote better environmental conditions” had the second-highest mean value of 4.30. “AI can automate HVAC, IAQ, and other building systems to optimize energy consumption” had the third-highest mean value of 4.27.

4.2.2. Factor Analysis and Reliability

Factor analysis is a statistical technique used to explain variability among observed, correlated variables by identifying a smaller number of unobserved variables, known as factors. It is a widely applied inter-dependency method, particularly when a set of variables exhibits systematic interrelationships, aiming to uncover the latent factors responsible for common variance. Factor loadings, including communality, represent the sum of squared loadings for all factors associated with a given variable, indicating the proportion of variance explained by these factors. Communality quantifies the percentage of variance in a variable accounted for by all identified factors and serves as a measure of the reliability of the indicator within the factor structure [69]. Factor analysis was utilized to identify the primary dimensions within the variables of digital twins, asset information requirements, and artificial intelligence. Principal component analysis was employed to empirically examine and validate the variables, with a summary of the results presented in Table 5. To evaluate the suitability of the data for factor analysis, both overall and individual measures of sampling adequacy were computed, with values exceeding 0.5 considered acceptable. The reliability of each extracted factor was assessed using Cronbach’s alphas, which evaluates the internal consistency of the factors based on the average correlation between variables within each factor. A minimum acceptable Cronbach’s alpha value is 0.7. The analysis of Cronbach’s alpha values revealed that all reliability coefficients α for the constructs listed in Table 5 demonstrated satisfactory levels of reliability, albeit with some constructs exhibiting higher reliability than others. Specifically, the constructs “Asset Information Requirements” and “Artificial Intelligence” exhibited the highest reliability coefficients α, with values of 0.898 and 0.885, respectively. Among the variables, “AI helps decision-making for the operation of assets in building facilities by enabling the integration and analysis of raw data” and “Asset Information Modeling (AIM) facilitates AM throughout operation and maintenance faces” demonstrated the highest factor loadings, with values of 0.902 and 0.882, respectively.

4.2.3. Correlation Analysis

Spearman’s rank correlation coefficient is a nonparametric statistical measure of rank correlation that evaluates the degree of association between two variables based on their rankings. It quantifies how well the relationship between two variables can be represented by a monotonic function [70]. A high Spearman correlation indicates that the relative ranks of observations in one variable correspond closely to those in the other, with a perfect correlation (ρ = 1) signifying identical ranks and a perfect negative correlation (ρ = −1) indicating completely opposite ranks.

The correlation analysis conducted using Spearman’s rank-order correlation reveals key interdependencies among critical components that influence the sector’s digital transformation. One of the most significant correlations identified is the strong positive association between Support of Asset Information Modeling (AIM) Creation through Asset Information Requirements (AIR) and Facilitation of Asset Management (AM) Across Operation and Maintenance Phases by AIM (ρ < 0.01, r = 0.743). This finding underscores the essential role of structured information management in ensuring effective operation and maintenance (O&M). The use of AIM, structured through AIR, enables seamless data integration, reducing inefficiencies and facilitating proactive asset lifecycle management. Furthermore, the correlation between Facilitation of Asset Management (AM) Across Operation and Maintenance Phases by AIM and Improvement of Asset Efficiency and Sustainability by AI (ρ < 0.01, r = 0.581) suggests that well-managed asset data significantly enhance the predictive and prescriptive capabilities of AI. AI-driven insights, powered by structured AIM, contribute to enhanced asset performance, energy optimization, and sustainability metrics. The results indicate that Enhancement of IEQ through DT has a moderate yet significant correlation with both Support of AIM Creation through AIR (r = 0.439) and Improvement of Asset Efficiency and Sustainability by AI (ρ < 0.01, r = 0.515). This highlights the role of DTs in dynamically monitoring and optimizing environmental parameters within buildings, ensuring adaptive and occupant-centric indoor conditions. The moderate association suggests that while DTs provide a foundation for IEQ improvement, their full potential is unlocked when combined with AI-based predictive analytics and automated asset management workflows. Detailed correlation computations concerning respondents’ perceptions of the AI-driven DT on enhancing IEQ and EE in SBS are presented in

Table 6. Correlation matrix of AI-driven DT for enhancing IEQ and EE in SBS.

Construct	Enhancement of Indoor Environment Quality (IEQ) Through DT	Support of Asset Information Modeling (AIM) Creation through Asset Information Requirements (AIR)	Facilitation of Asset Management (AM) Across Operation and Maintenance Phases by AIM	Improvement of Asset Efficiency and Sustainability by AI
Enhancement of Indoor Environment Quality (IEQ) through DT	1.000
Support of Asset Information Modeling (AIM) Creation through Asset Information Requirements (AIR)	0.439 **	1.000
Facilitation of Asset Management (AM) Across Operation and Maintenance Phases by AIM	0.387 **	0.743 **	1.000
Improvement of Asset Efficiency and Sustainability by AI	0.515 **	0.531 **	0.581 **	1.000

Note: N = 85. ** The correlation is significant at the 0.01 level (2-tailed).

.

The results have several implications for the smart building sector:

i.

Strengthening AIM-AIR Integration: Given the strong correlation between AIM creation and AM facilitation, policymakers and industry stakeholders should emphasize the standardization and implementation of comprehensive AIR frameworks to enhance AIM usability.

ii.

Leveraging AI for Proactive Management: The positive association between AIM-driven AM facilitation and AI-enhanced efficiency suggests that AI applications in predictive maintenance, fault detection, and performance optimization should be further explored.

iii.

Expanding DT Capabilities Beyond Monitoring: While DTs contribute to IEQ improvement, their integration with AI-driven automated control systems could further enhance building adaptability and occupant satisfaction.

iv.

Interdisciplinary Collaboration: Given the interdependent nature of AIM, AI, and DT technologies, fostering collaboration among data scientists, facility managers, and sustainability experts is crucial for achieving optimized, data-driven building operations.

5. Discussion

Smart buildings are at the forefront of energy reduction strategies and environmental impact mitigation because of the growing demand for sustainable solutions within the built environment. This need is accompanied by an improvement in the IEQ. The purpose of this study is to shed light on the function that DT technology plays in improving the IEQ and enhancing the EE of smart buildings. In addition, the purpose of the research is to outline the asset information needs that are essential for the implementation of this technology. In addition, it considers the possibility of combining DT with AI to improve the accuracy of fault detection systems, enhance predictive maintenance procedures, and maximize energy efficiency. In conclusion, the purpose of this research is to analyze and appreciate the complex difficulties that companies confront when it comes to the adoption and deployment of AI inside their operational system.

5.1. Theoretical Contributions

5.1.1. Enhancement of IEQ Through DT

The strategic deployment of DTs provides facility management teams with comprehensive insights into occupant experiences within built environments. By leveraging sensor-driven data analytics, these virtual models facilitate precise regulation of mechanical and electrical systems, enabling real-time adjustments to parameters such as indoor temperature and air quality. This capability not only fosters an occupant-centric indoor environment but also enhances operational efficiency, ensuring that building performance aligns with predefined cost and energy benchmarks. The real-time monitoring capabilities of DTs, particularly in tracking CO₂ concentrations and dynamically adjusting ventilation systems, underscore their critical role in facilitating continuous decision-making. Moreover, their adaptive functionality in response to external climatic conditions enhances energy efficiency while maintaining optimal IAQ [23,44,45].

5.1.2. Support of Asset Information Modeling (AIM) Creation Through Asset Information Requirements (AIR)

AIM’s significance in automating fault detection, enables predictive maintenance strategies, and offers a comprehensive understanding of building asset configurations. Merino et al. [71] assert that AIM serves as a fundamental mechanism for prompt problem resolution and performance optimization, thereby contributing to effective asset management (AM). Furthermore, Desogus et al. [46] emphasize that integrating sensor-enhanced AIM facilitates real-time monitoring of various parameters, including internal comfort levels, occupant well-being, and overall energy consumption patterns in existing buildings. This integration fosters data-driven decision-making, ensuring that maintenance strategies are both proactive and responsive to evolving operational demands.

5.1.3. Facilitation of Asset Management (AM) Across Operation and Maintenance (OM) Phase by AIM

The convergence of DT and AI presents a promising approach to efficient asset administration by enabling customized building systems that adapt dynamically to real-time data. This integration effectively harmonizes occupant comfort with operational efficiency, promoting environmentally sustainable and health-conscious building environments [46]. Despite these advantages, key challenges such as stakeholder comprehension, structured data organization, and specialized expertise requirements must be addressed to fully realize the potential of AI and DT in AM [72]. Merino et al. [71] highlight that integrating building automation systems (BAS), BIM, and real-time IoT-generated data facilitates the implementation of dynamic AM applications, ultimately optimizing operational efficiency and supporting net-zero building objectives (e.g., enhancing HVAC system performance). Moreover, integrating facility management practices with cutting-edge technologies such as DTs and AI augments decision-making processes, imbuing building operations with advanced predictive capabilities that enhance long-term sustainability and performance efficiency [73].

5.1.4. Improvement of Asset Efficiency and Sustainability by AI

Despite existing challenges, the systematic integration of AI with DT technology offers substantial benefits, including enhanced decision-making capabilities, increased operational efficiency, and improved sustainability metrics within SBS. Addressing these challenges necessitates a structured approach centered on high-quality data acquisition and well-designed technological integration strategies, which can yield significant environmental and economic benefits. In the OM phase, organizations encounter numerous obstacles when incorporating AI and DT solutions, emphasizing the critical need to resolve these issues to achieve effective building management [44,45,51,74]. Karatzas et al. [74] underscore that DTs enable precise real-time data acquisition, thereby improving operational control of building assets and refining the ability to predict individual thermal comfort preferences through novel spatial analysis methodologies. Similarly, Tahmasebinia et al. [56] highlight that DT technology offers substantial potential for optimizing physical infrastructure performance, supporting plant operations, and facilitating AM while effectively mitigating energy-related emissions.

The findings of this study provide critical insights into the integration of AI with DTs within SBS. Empirical evidence derived from survey responses (Table 5) and thematic analyses of qualitative interviews (Table 4) establishes a strong foundation for the practical implementation of AI and DT technologies in SBS. This integration is poised to revolutionize decision-making processes, enhance predictive maintenance capabilities, and optimize energy performance across various operational contexts. The synthesized guidance from this research will serve as an essential framework for embedding AI within DT-driven environments, underscoring its pivotal role in addressing contemporary challenges in SBS.

Figure 3 visually represents the framework for procedural dynamics of AI-DT integration, serving as a conceptual guide for researchers and practitioners. This visualization helps in understanding the complex interplay between systems-based, model-based, data-based, and analytic-based DT framework, highlighting their synergistic potential. Overall, this study not only clarifies the practical aspects of AI and DT integration in SBS for BEMS and IEQ but also paves the way for future research. It invites academic and professional communities to explore how AI and DT can collaboratively advance SBS for BEMS and IEQ, ensuring sustainability, efficiency, and improved operational efficacy.

5.2. Practical Implications

The integration of AI with DT presents a transformative approach to improving IEQ and EE in SBS. The practical implications of this advancement span across design, construction, AI, and policy domains, offering significant opportunities and challenges that need to be carefully considered.

5.2.1. Design Implications

i.

Dynamic and Adaptive Building Designs: Incorporation of DTs: AI-driven DTs necessitate the integration of dynamic and real-time monitoring systems in the design phase, enabling continuous adaptation of building designs to changing environmental conditions and occupant behaviors. This results in designs that are not only energy-efficient but also capable of maintaining optimal IEQ.

ii.

Energy-Efficient System Layouts: Optimized HVAC Systems: Designers must consider the placement and configuration of (HVAC systems to fully leverage AI’s predictive capabilities. This ensures the systems can respond to AI-driven insights, reducing energy consumption while maintaining comfort levels.

iii.

Smart Material Selection: Intelligent Material Use: The selection of materials with properties that complement AI-driven DTs, such as smart windows or energy-efficient insulation, becomes critical. These materials should work in harmony with the predictive models to enhance overall building performance.

5.2.2. Construction Implications

i.

Integration of Sensors and IoT Devices: Construction Readiness: The construction phase must account for the installation of IoT devices and sensors necessary for DTs to function. This includes ensuring that buildings are equipped with the infrastructure needed to support real-time data collection and AI-driven decision-making processes.

ii.

Enhanced Construction Processes: Precision and Efficiency: AI-driven DTs can monitor construction progress, ensuring that energy efficiency and IEQ goals are met during the construction phase. This leads to fewer deviations from design specifications and improved construction quality.

iii.

Retrofit and Upgrade Considerations: Adapting Existing Structures: For existing buildings, construction processes will need to include retrofitting capabilities to integrate AI-driven DTs. This involves upgrading electrical systems, HVAC setups, and installing new sensors, which could be costly but necessary for achieving desired outcomes.

5.2.3. AI Implications

i.

Advanced Predictive Analytics: Real-Time Decision-Making: AI models must be sophisticated enough to analyze vast amounts of data in real time, predicting future conditions and optimizing system performance accordingly. This requires continuous learning and adaptation to new data inputs.

ii.

AI-Driven Automation: Autonomous Control: The implementation of AI-driven automation within smart building systems can lead to automatic adjustments in lighting, HVAC, and other systems to maintain optimal IEQ and energy use without human intervention. This enhances operational efficiency and reduces the need for manual control.

iii.

Data Management and Security: Data Handling: The vast amount of data generated by AI-driven DTs necessitates robust data management strategies. Ensuring data privacy, security, and integrity becomes crucial, particularly in protecting sensitive information related to building operations and occupant behavior.

5.2.4. Policy Implications

i.

Regulatory Standards for AI and DT Integration: Policy Development: Governments and regulatory bodies must establish standards and guidelines for the integration of AI and DTs in building systems. These policies should address data privacy, energy efficiency benchmarks, and safety standards, ensuring that buildings equipped with these technologies comply with legal and ethical norms.

ii.

Incentives for Sustainable Construction: Government Support: Policymakers should consider providing incentives for the adoption of AI-driven DTs in construction, such as tax breaks or subsidies for developers who incorporate these technologies to improve IEQ and energy efficiency. This could accelerate the adoption of sustainable building practices.

iii.

Building Codes and Certifications: Updating Standards: Existing building codes and certification processes, such as LEED or BREEAM, may need to be updated to reflect the capabilities of AI-driven DTs. This includes incorporating new criteria for real-time monitoring, adaptive systems, and energy optimization.

iv.

Public Awareness and Education: Stakeholder Engagement: To ensure widespread adoption, policies should also focus on raising awareness among stakeholders, including building owners, developers, and the public, about the benefits and potential of AI-driven DTs. Educational initiatives can help in overcoming resistance to adopting new technologies and ensure that the workforce is prepared for the shift.

Figure 4 summarizes the practical implications of AI-driven DT for enhancing IEQ and EE in SBS

Table 7. Opportunities and challenges in implementing AI-driven DT for enhancing IEQ and EE in SBS.

Process	Opportunities	Challenges	Potential Improvements
Asset Identification and Classification	Enables the creation of a comprehensive asset inventory, facilitating effective asset management.	Challenges include the accurate categorization of assets and the inclusion of all critical components.	Develop standardized protocols for asset categorization and periodic reviews to ensure inventory completeness.
Real-Time Monitoring and Analysis	Employs sensors to constantly monitor crucial parameters such as temperature, humidity, and occupancy.	Ensuring the accuracy and reliability of sensor data and managing large volumes of real-time data pose significant challenges.	Implement advanced data validation techniques and scalable data management systems for efficient data handling.
AI-Driven Analytics	AI algorithms interpret data from digital twins for advanced predictive analytics, enhancing decision-making.	Crafting and refining AI models requires specialized expertise and continual tuning for precision.	Invest in continuous AI training and development to enhance model accuracy and applicability.
IEQ Monitoring and Optimization	Real-time IEQ monitoring with sensors, automatically adjusting building systems to optimize conditions.	The reliability and calibration of sensors are critical challenges for precise data collection.	Establish rigorous sensor maintenance and calibration schedules to maintain data integrity.
AI Optimizing Building Energy Consumption	Utilizing AI to analyze data for energy optimization leads to potential cost savings and improved sustainability.	High-quality, stable data input is required, alongside managing the balance between optimization and operational constraints.	Focus on data quality assurance and develop adaptive AI systems that can account for operational variability.
Decision-making and Predictive Optimization by AI	AI aids in analyzing extensive datasets to improve decision-making and assess risks and uncertainties.	Sufficient volumes of high-quality historical data are necessary to train precise AI models.	Strengthen data collection and curation processes and ensure robust validation of AI model outputs.

details the operational challenges, opportunities, and improvements arising from the AI-DT convergence, offering stakeholders a comprehensive understanding of the landscape.

5.3. Opportunities, Challenges, and Potential Improvements in AI-Driven DT Implementations for Enhancing IEQ and EE in SBS

5.3.1. Opportunities

AI-driven DT offers significant opportunities in AM by enabling precise identification, classification, and tracking of building assets. Integrating sensor data provides a comprehensive view of asset performance, supporting informed decision-making and predictive maintenance, which reduces downtime and maintenance costs. Continuous monitoring of IEQ parameters ensures optimal occupant conditions, while real-time energy consumption data facilitate quick adjustments for EE. AI-driven analytics uncover patterns and insights from vast datasets, continuously refining operations for enhanced performance, energy savings, and occupant comfort. AI optimizes building energy consumption by dynamically adjusting systems like HVAC and lighting, significantly reducing energy costs, and helping achieve sustainability goals. Furthermore, AI provides data-driven insights for informed decision-making, anticipates future conditions for predictive optimization, and improves long-term planning and resource allocation based on predictive analytics, enhancing overall building management and efficiency.

5.3.2. Challenges

Implementing AI-driven DT in SBS faces several challenges. Integrating data from different systems and formats is complex, as is ensuring sensor data accuracy and reliability. Achieving seamless communication between diverse systems and devices is critical. The technical complexity of AI and DT systems requires advanced expertise, and the initial investment costs can be significant. Scaling these solutions across large and complex buildings adds further difficulty. Security and privacy are paramount, with the need to protect sensitive data from cyber threats and ensure occupant privacy during data collection and analysis. Regulatory and compliance issues also pose challenges, as navigating various frameworks can impede implementation, and ensuring all systems meet relevant standards is essential.

5.3.3. Potential Improvements

To improve AI-driven DT implementations in smart buildings, several enhancements can be made. Unified data platforms should be developed to integrate data from all sources, ensuring cohesive systems. Rigorous validation and cleaning processes are necessary to ensure high-quality data inputs. Advanced AI algorithms, including continuously improved machine learning models, should be created to enhance prediction and optimization accuracy. AI systems must also adapt to changing conditions and requirements. Cost-effective solutions, such as affordable AI and DT technologies, should be developed to be accessible to more building owners. Government incentives and subsidies can help offset initial costs. Strengthened security measures, including advanced protocols and strict privacy guidelines, are crucial to protect data and systems from cyber threats. Proactive compliance with regulatory changes is essential, and efforts should be made to standardize AI and DT technologies to simplify compliance and integration.

5.4. Limitations and Future Studies

This research has certain limitations. The sample size was relatively small and constrained, with limited parameter exploration and a focus on only a subset of regions. The limited sample size and regional emphasis may affect the study’s statistical strength, making it difficult to detect subtle yet significant effects. A limited sample size and regional focus may hinder the generalizability of findings beyond the studied group. Due to variations in cultural, social, economic, and environmental factors across locations, conclusions drawn from small samples may only partially reflect the characteristics and practices of some organizations in the industry.

Given the promising findings of this research, future studies should explore several key areas to advance the integration of AI-driven DTs in SBS:

-

Scalability and Implementation in Real-World Scenarios: While this study establishes the theoretical and empirical basis for AI-driven DT integration, future research should focus on large-scale, real-world implementations. Studies should investigate the long-term performance, challenges, and practical feasibility of integrating DTs with AI in diverse building types and operational settings.

-

Advanced AI Algorithms for Predictive Analytics: The development of more sophisticated AI algorithms tailored for SBS applications can enhance the predictive and diagnostic capabilities of DTs. Future studies should explore deep learning techniques, reinforcement learning models, and hybrid AI frameworks to optimize IEQ and energy performance dynamically.

-

Interoperability with Emerging Technologies: The compatibility and integration of AI-driven DTs with other emerging technologies such as blockchain for secure data transactions, edge computing for real-time processing, and 5G for enhanced connectivity warrant further investigation.

-

Human-Centric AI and Occupant Behavior Modeling: AI models should be further refined to incorporate human-centric factors, including occupant behavior patterns, comfort preferences, and adaptive responses to indoor environmental changes. Future research should explore the integration of physiological and cognitive parameters into AI-driven DT systems.

-

Sustainability and Life Cycle Assessment: Research should extend beyond operational efficiency to evaluate the environmental impact and life cycle sustainability of AI-driven DTs. This includes assessing the embodied energy, carbon footprint, and long-term sustainability metrics associated with DT implementations.

-

Policy and Regulatory Considerations: The adoption of AI-driven DTs in SBS is influenced by regulatory frameworks and industry standards. Future studies should examine policy developments, ethical considerations, and legal frameworks that govern AI and DT applications in smart buildings.

The application of AI-driven DT still presents many challenges in the industry field. It is essential to conduct studies on these challenges, perhaps the most important of which are the following:

-

Data Security: Protecting confidential information from unauthorized access is an ongoing challenge. Exploring advanced technologies like blockchain may prove beneficial in ensuring data authenticity and accuracy. Blockchain-based solutions can enhance the reliability of historical data by providing a secure and immutable method for recording and verifying data transactions.

-

Historical Data: The primary challenge lies in the quality of historical data, as most of them tend to be inaccurate. Developing quality inspection and evaluation algorithms can streamline the cleaning and preparation of old data. This approach can help researchers save time and effort in getting historical data ready for analysis.

6. Conclusions

This research contributes to the knowledge base by exploring the potential of DTs to enhance IEQ and EE in SBS. Through qualitative and quantitative analysis with field experts, the study highlights the critical role of combining DT with AI for effective building management. Key findings include the significant improvements in operational efficiency through AI-powered DTs, which enable predictive maintenance, optimize energy consumption, and adapt building features for sustainability and occupant comfort.

AI-driven DTs significantly enhance AM by accurately identifying, classifying, and tracking building assets. Integrating sensor data provides a comprehensive view of asset performance, supporting informed decision-making and predictive maintenance, thus reducing downtime and costs. Continuous monitoring of IEQ ensures optimal occupant conditions, while real-time energy consumption data facilitates quick adjustments for EE. AI-driven analytics uncover patterns and insights from vast datasets, refining operations for improved performance, energy savings, and comfort. AI optimizes energy consumption by dynamically adjusting HVAC systems and lighting, reducing costs, and achieving sustainability goals. Additionally, AI provides data-driven insights for informed decisions, predictive optimization, and better long-term planning. However, implementing AI-driven DT in SBS faces challenges such as integrating data from diverse systems, ensuring data accuracy, and achieving seamless communication between devices. The technical complexity and initial costs require advanced expertise. Scaling solutions across large buildings, ensuring data security and privacy, and navigating regulatory and compliance issues are also significant hurdles.

This study highlights the significant potential of AI-driven DTs in enhancing IEQ and optimizing EE in SBS. The findings demonstrate that DTs enable real-time monitoring and adaptive control of environmental parameters, leading to improved occupant comfort and operational efficiency. Additionally, their integration with AIM supports predictive maintenance, data-driven decision-making, and proactive AM. The synergy between AI and DTs further enhances sustainability by optimizing building performance and mitigating energy-related emissions.

While the study establishes a strong foundation for AI-DT integration in SBS, several avenues for future research remain. Key areas include real-world scalability, the development of advanced AI algorithms for predictive analytics, and improved interoperability with emerging technologies such as blockchain, edge computing, and 5G. Additionally, refining AI models to incorporate human-centric factors, evaluating sustainability through life cycle assessments, and addressing policy and regulatory considerations are crucial for widespread adoption.

Overall, the study underscores the transformative role of AI-driven DTs in SBS and provides a framework for future advancements aimed at achieving intelligent, sustainable, and occupant-centric built environments.