Advancing BIM and game engine integration in the AEC industry: Innovations, challenges, and future directions

Abstract

The integration of building information modelling (BIM) with real-time game engines (GEs) promises to transform the Architecture, Engineering, and Construction (AEC) industry by enhancing design processes, streamlining construction workflows, and fostering more effective stakeholder collaboration. To understand the status of BIM and GE integration, this study employs a mixed-method approach by combining meta-analysis and meta-synthesis to assess the current state of BIM and GE integration. Drawing on data from academic databases such as Scopus and Web of Science, it classifies applications into key domains (e.g. design visualization, modular construction workflows, and real-time simulation) and examines the technologies, workflows, and evolving trends. The findings demonstrate that integration notably improves productivity, decision-making, and collaboration, yet widespread adoption remains hindered by persistent challenges, including interoperability barriers, high implementation costs, scalability constraints, and limited exploration of emerging technologies, such as generative AI, robotics, and reinforcement learning. Overcoming these issues is crucial to realizing the full potential of BIM–GE integration. Building on these insights, this paper proposes a future research agenda, encouraging the development of standardized integration protocols, more intuitive user interfaces, and advanced interoperability solutions. It also advocates for incorporating extended reality, automation, and other advanced technologies to support real-time, scalable, and collaborative environments. While this work provides a robust foundational resource for both researchers and industry practitioners, it acknowledges certain limitations, such as a reliance on academic literature and a greater emphasis on methodological aspects over practical implementations.

Graphical Abstract

Open in new tabDownload slide

List of Abbreviations

 
• AEC

  Architecture, Engineering, and Construction

 
• AI

  Artificial intelligence

 
• AR

  Augmented reality

 
• BIM

  Building information modelling

 
• CSV

  Comma-separated values

 
• DES

  Discrete event simulation

 
• DfMA

  Design for manufacturing and assembly

 
• DTS

  Data transmission system

 
• FBX

  Filmbox (File format for 3D models)

 
• GEs

  Game engines

 
• GPU

  Graphics Processing Unit

 
• HMD

  Head-mounted display

 
• HBIM

  Historic Building Information Modelling

 
• IFC

  Industry Foundation Classes

 
• IoT

  Internet of things

 
• LOD

  Level of detail

 
• ML

  Machine learning

 
• NLP

  Natural Language Processing

 
• PoE

  Post-occupancy evaluation

 
• PSC

  Public safety communication

 
• Revit API

  Revit Application Programming Interface

 
• SaaS

  Software as a Service

 
• SMEs

  Small and medium enterprises

 
• SQL

  Structured Query Language

 
• UE

  Unreal engine

 
• UGV

  Unmanned ground vehicle

 
• UI

  User interface

 
• VR

  Virtual reality

 
• VRML

  Virtual Reality Modelling Language

 
• WOS

  Web of Science

 
• XR

  Extended reality

1. Introduction and Background

Innovation in the Architecture, Engineering, and Construction (AEC) sector is driven by significant advancements in design methodologies, materials, constructability, and organizational practices, with a pronounced focus on sustainable and technological advancements (Lo et al., 2022). This shift towards innovation is crucial in enhancing performance and efficiency, underscoring the role of innovative technologies in construction processes and project management as pivotal to the industry's evolution towards sustainability, efficiency, and optimized performance (Alaloul et al., 2020). Concurrently, collaboration and stakeholder engagement emerge as central pillars of this innovative thrust. Qualitative research highlights the critical impact of stakeholder engagement on the sustainability and success of projects, notably in prefabricated construction (Luan et al., 2022). The drive for stakeholder participation in large-scale projects is fuelled by the potential for collaborative innovation, closely aligning with sustainable construction objectives (Zhang et al., 2020). Together, these dimensions emphasis the integral role of stakeholder collaboration in advancing innovation and sustainability within the AEC industry.

BIM, a transformative technology in the AEC industry, enables the creation of digital representations of physical and functional characteristics of places (Wu et al., 2023). BIM facilitates a collaborative environment, allowing for the integration of distinct aspects of the design, construction, and management of buildings within a single digital model (Fernández Rodríguez et al., 2023; Salleh et al., 2023). It enhances visualization, improves productivity, reduces costs, and streamlines project timelines. Significant benefits include better-coordinated designs, reduced errors and conflicts, and improved decision-making processes (Azhar et al., 2011). Despite its numerous advantages, BIM adoption faces several challenges. These include the excessive costs associated with BIM software and training (Alam et al., 2023), the complexity of BIM technology (Kineber et al., 2023), and resistance to change within organizations (Ismail et al., 2022). There are also issues related to interoperability between different BIM software and the management of BIM data (Waqar et al., 2023). Legal and contractual issues arise concerning the ownership and liability of BIM models (Murguia et al., 2023). Furthermore, the AEC industry faces a notable challenge in the domain of complex architectural interactive design and immersive visualization, areas where BIM alone may fall short (Xu et al., 2022; Liu et al., 2023a). Addressing this gap necessitates the integration of BIM with interactive and immersive visualization technologies, enabling enhanced interactivity for various stakeholders. This integration promises to elevate the design and visualization capabilities within the sector, fostering a more collaborative and engaging project development environment.

GE are software frameworks designed for the creation and development of video games and interactive digital environments (Politowski et al., 2021). GE simplifies the game development process by providing functionalities such as graphics rendering, physics engine, sound, scripting, animation, artificial intelligence (AI), networking, streaming, memory management, threading, and a scene graph (Varsami et al., 2022; Khameneh et al., 2023). Beyond gaming, these engines are increasingly used in other fields such as virtual reality (VR), augmented reality (AR), architectural visualization, and even in automotive design for virtual prototypes (David et al., 2022). The interactive and immersive capabilities of GE make them ideal tools for a wide range of applications beyond traditional game development. The most widely utilized GE in the domain of game development include Unity, Unreal Engine (UE), GameMaker, Godot, and CryEngine (Vohera et al., 2021). The integration of GE with BIM technology effectively narrows the gap between intricate architectural designs and interactive visualization, significantly enhancing stakeholder engagement and decision-making processes. This convergence leverages the robust visualization and interactive strengths of GE, augmenting collaboration, and comprehension of spatial dynamics (Ehab et al., 2023). Such convergence facilitates the real-time creation and evaluation of design alternatives, significantly elevating stakeholder participation in the design phase (Buhammood et al., 2020). Moreover, the application of BIM in constructing simulation models, streamlined by GE, diminishes the time and effort necessitated for model creation, thus promoting the reuse of models, and ensuring efficient communication of outcomes to stakeholders (Osorio-Sandoval et al., 2022). Additionally, the adoption of this integrated approach within engineering education has demonstrated advantages in augmenting learning outcomes and enriching the educational journey (Sanchez et al., 2022).

To date, multiple research endeavours have focused on the integration of BIM with GE, aiming to explore various aspects of the AEC sector, including but not limited to realistic rendering, visualization, AR/VR applications, and simulation. The compilation and analysis of these studies through a comprehensive review article are crucial to distilling their benefits for both the academic and industrial domains within the AEC domain. While numerous reviews have been published concerning the role of BIM in the AEC sector and its synergies with industry 4.0 technologies such as machine learning (ML), robotics, and the internet of things (IoT), a prominent gap remains in the literature regarding a dedicated review on the integration of BIM with GE for fostering innovation in the AEC sector. To address this gap, this article presents an in-depth review of the current state of the art concerning the integration of GE technologies and BIM within the AEC domain. The scope of this paper extends to a critical review of scholarly articles that explore various domains of AEC innovation facilitated through the integration of BIM and GE technologies. This includes, but is not limited to, areas such as architectural visualizations, real-time realistic renderings, construction simulations, construction safety, stakeholder engagement, facility management, project management, and project planning.

This paper aims to explore current trends in BIM and GE integration and propose future directions. The objectives are: (1) to conduct a meta-analysis for a quantitative examination of state-of-the-art publications, (2) to perform a meta-synthesis for a qualitative assessment identifying trends and gaps, and (3) to recommend future research directions. The main research questions are: (i) what are the current trends, challenges, and future directions in BIM and GE integration? (ii) Which GEs are commonly used with BIM? (iii) What is the typical workflow for BIM–GE interoperability? To address these questions, relevant studies from Scopus and Web of Science (WOS) were analysed both quantitatively and qualitatively to clarify trends, challenges, and future directions.

Previous reviews on BIM–GE integration have largely focused on visualization and VR/AR applications. In contrast, this study employs a data-driven approach using meta-analysis and meta-synthesis, to systematically classify applications, assess interoperability challenges, and explore emerging technologies. It evaluates both technical and economic feasibility by discussing interoperability solutions, middleware applications, and standardized protocols to overcome adoption barriers. Moreover, the scope extends beyond design visualization to include real-time simulation, modular construction, automation, and project management. Addressing a gap in the literature, the review also assesses the economic feasibility for small and medium enterprises (SMEs) and proposes cost-effective strategies such as open-source solutions and subscription-based Software as a Service (SaaS) models. Additionally, a generic framework for BIM–GE interoperability is introduced, outlining data conversion pipelines, automation mechanisms, and multi-user collaborative environments for scalable, real-time integration. By tackling both technological and economic challenges, this study bridges critical research gaps and serves as a practical guide for advancing BIM–GE integration.

This paper is organized as follows: Section 2 details the methodology, including data collection and screening processes. Section 3 presents a meta-analysis of the papers collected, offering quantitative results. Section 4 provides a meta-synthesis, highlighting identified trends in the literature. Section 5 discusses findings, including integrations, workflows, research challenges, gaps, and future development opportunities. Section 6 concludes with the contributions of this research to both industry and academia.

2. Methodology

A mixed review approach was used to fully understand the topic within the defined scope, as shown in Figure 1 which outlines the process of data collection and screening, succeeded by a mixed review analysis comprising both quantitative and qualitative aspects. Following these analyses, insights derived from the mixed review, including identification of trends, challenges, research gap, and future research directions. The subsequent sections will briefly describe methodology and tool used for each step of the research process.

2.1. Finding relevant citations: database selection and keyword search

The selection of databases and strategic keyword searches are critical to ensuring relevant and high-quality sources. Data collection and screening followed PRISMA guidelines (Page et al., 2021) using Scopus and WOS. Searches targeted ‘Title/Abstract/Keywords’ fields with four keyword combinations: '(i) (game AND engine) AND (BIM), (ii) (game AND engine) AND (construction), (iii) (4D AND simulation) AND (construction), and (iv) (4D AND simulation) AND (game AND engine). This strategy, inclusive of journals, conferences, and book chapters, ensured alignment with research objectives without restrictions on publication year or field. The use of ‘AND’ refined results to focus on AEC-related publications, emphasizing temporal and procedural aspects of construction relevant to BIM–GE integration. The specific keyword combinations minimized the need for additional filtering by inherently narrowing results to the thematic scope.

2.2. Collection and screening of relevant literature

The initial search yielded 252 articles (90 from WOS and 162 from Scopus), as shown in Figure 2. The fourth keyword set in WOS produced no relevant results. Screening for duplicates excluded 65 articles, leaving 187 for retrieval. Of these, 12 were inaccessible and excluded. The remaining 175 articles were screened using the following criteria: (i) English language, (ii) incorporation of GE technology, and (iii) academic research sources, excluding reviews and non-academic materials. This reduced the corpus to 87 articles. An additional 31 articles identified through references and citation tracking were screened, with 25 meeting the criteria. The final data set comprised 112 articles for the mixed review methodology.

2.3. Analytical methods

The study employs a mixed review approach combining meta-analysis and meta-synthesis techniques. VOS viewer software was used for the meta-analysis due to its effectiveness in visualizing keyword co-occurrence and bibliometric networks, identifying key topics, trends, and relationships (Lin et al., 2011). To address potential gaps in the automated meta-analysis, a meta-synthesis was conducted, providing a qualitative review of research themes and findings. This dual approach integrated empirical evidence with thematic exploration to analyse the integration of BIM and GE technologies and their impact on the AEC sector. Key relationships and integration strategies were identified and visualized using a Sankey diagram (SankeyMATIC: A Sankey diagram builder for everyone, n.d.), which effectively represents complex data flows through varying link widths (Glover et al., 2022). This visualization method clarified multifaceted connections, highlighting trends and interaction volumes.

3. Meta-analysis

The meta-analysis incorporated 112 publications, whose annual scholarly output is visualized in a bar chart in Figure 3A. There is a visible trend of fluctuating research output over the years, with a notable increase in publications starting from 2016 and peaking in 2021. Figure 3B quantifies the usage of GE tools in the literature, revealing a predominant preference for Unity 3D, used in 84 instances, followed by UE with 20. Unity 3D and UE are particularly favoured for their VR/AR support and comprehensive features that enhance architectural visualization and simulation. In contrast, engines such as Godot, Microsoft XNA Game Studio, Torque 3D, Panda 3D, Radiant, CryEngine, and Havok show minimal usage, likely due to their limited features and lower applicability to BIM tasks. The dominance of Unity 3D as the primary GE for BIM applications underscores a preference for user-friendly, widely supported, and feature-rich platforms. However, this trend also suggests an overreliance on a single proprietary solution, which could limit interoperability, hinder innovation in alternative GE, and create barriers for SMEs due to licensing costs. UE, despite its advanced graphical capabilities, remains underutilized, indicating that ease of use and accessibility may outweigh technical superiority in adoption decisions. This highlights an opportunity for future research to explore alternative open-source GEs such as Gadot that could offer cost-effective and customizable solutions.

The literature is categorized into seven main domains, each with its corresponding subcategories, based on specific classification criteria, including the study's objective, application, and contribution. The author analysed the findings of individual studies and grouped them according to their predominant functionality or application context. For instance, studies that primarily utilized GEs for ‘design visualization’ were grouped together, while those focusing on ‘safety’ or ‘quality control’ were categorized under construction management. By following this thematic approach, the classification ensures that each domain aligns with a coherent set of research objectives, effectively capturing the purpose and technical contributions of GE integration in BIM-related applications.

Figure 4 illustrates the distribution of literature on integrating BIM and GE technologies in contemporary research and practice. Project Management dominates with 24.11%, reflecting the technologies potential to enhance project tracking, scheduling, and resource allocation through real-time capabilities and detailed information models. Simulation and Analysis, at 4.46%, highlights the limited use of GE for dynamic BIM data simulation, showcasing its utility in performance analyses, construction sequence simulations, and design validation. The seven domains were subdivided into 16 subdomains for deeper analysis. Enhancing stakeholder engagement during the design phase emerged as a key theme, covering 18.75% of the literature. In contrast, the least explored themes include applications for ‘Emergency Evacuation Scenarios’ (2.68%), ‘Discrete Event Simulation (DES)’ (1.79%), and ‘Daylighting, Energy Usage, and Sustainability’ (2.68%). This distribution underscores the diverse applications of BIM and GE technologies while identifying areas requiring further research.

3.1. Analysis of keyword co-occurrence

To analyse the knowledge framework and progress in BIM–GE integration, keyword co-occurrence analysis was conducted using Scopus and WOS data. Synonymous terms were manually merged for accuracy. The analysis, performed with VOS viewer, employed cluster- and time-based methodologies, setting a minimum keyword occurrence threshold of five to identify significant terms. The resulting networks and density maps highlight thematic clusters and interrelated concepts, providing a clear depiction of the research landscape.

3.1.1. Time-based visualization

Figure 5 illustrates a time-based keyword co-occurrence analysis, highlighting the evolution of research themes through colour-coded years. Early studies, marked in blue, focus on BIM and GE applications for visualization and 3D modelling. Persistent themes like architectural design, BIM and VR remain central over time. Recent advancements include AR, laser scanning, and mixed reality for construction management, spatial conflict resolution, and workflow optimization. Modular construction emerges as a significant new focus, reflecting the field's ongoing evolution and technological integration.

3.1.2. Keywords co-occurrence cluster networks

Keyword co-occurrence network (KCN) analysis maps keyword relationships in the literature, using nodes (keywords) and edges (co-occurrences) to indicate linkages, with thicker edges signifying stronger connections (Radhakrishnan et al., 2017). Figure 6 shows terms like GE, BIM, architectural design, and VR form strong linkages, while thinner edges denote weaker associations. The network, adjusted for modularity and time sequence, places frequently occurring keywords at the core, diminishing outward.

The network divides into four clusters: Cluster 1 emphasizes BIM, GE, and architectural design, focusing on efficiency and integration of digital and physical processes, with keywords centralized along the timeline. Cluster 2 explores technological impacts, including robotics, modular methods, and e-learning, with keywords near the timeline's end, reflecting recent innovations. Cluster 3 highlights early applications like 3D graphics and laser scanning for historic preservation, with keywords appearing at the timeline's start. Cluster 4 examines VR and AR in GE, enhancing design and construction processes, with keywords at the timeline's centre, showing growing adoption alongside BIM (Wang et al., 2009). This analysis underscores thematic trends, technological integration, and the evolving focus of BIM–GE research.

4. Meta-synthesis

Meta-analysis, as a quantitative approach, fails to fully capture trends, challenges, and gaps in the literature. To address this, a meta-synthesis using thematic analysis was conducted through an in-depth review of 112 articles, categorized into subcategories from Figure 4. Each subcategory was examined using triangulation, comparing data interoperability, technology usage, findings, and evaluation metrics to ensure validity and robustness. The subsequent section outlines research trends within each sub-category.

4.1. Design and visualization

4.1.1. 3D architectural visualization

The integration of BIM and GE has followed a clear trajectory from basic 3D visualization towards real-time, immersive, and interactive applications, particularly in areas such as smart buildings, geospatial modelling, and construction management (Pouke et al., 2018). Latest trends summarize in Table 1 indicate a shift from static BIM-based renderings to dynamic environments where users can engage with real-time simulations, digital twins, and VR-enabled design reviews (Janovský, 2022). However, interoperability remains a persistent challenge, as BIM's parametric models do not seamlessly translate into the mesh-based environments of GEs, often requiring manual adjustments, texture reassignments, and model simplifications (Virtanen et al., 2020). Additionally, cost and scalability concerns arise, as high polygon counts in BIM models strain real-time rendering capabilities, necessitating LOD (level of detail) techniques, AI-driven optimization, and procedural modelling to enhance performance (Virtanen et al., 2020; Janovský, 2022). For instance, Bogen & East (2011) attempted to transform IFC-based BIM models into GE for multi-user visualization but encountered significant difficulties in maintaining traceability and metadata consistency due to conversion limitations. Their findings illustrate how current workflows lack automated and standardized tools for seamless BIM–GE integration, highlighting the need for open BIM formats (e.g. IFC 4.0, glTF) and AI-assisted automation pipelines to enable cost-effective, scalable, and real-time interoperability between BIM and GE.

4.1.2. Architectural heritage visualization and preservation

The integration of BIM and GE in 3D architectural heritage visualization has evolved to enhance model accuracy, real-time performance, and user interaction. Emerging trends emphasize the transition from static 3D renderings to immersive, interactive environments, utilizing VR, AR, and serious gaming technologies (Pybus et al., 2019; Ma, 2021). Tools such as AutoCAD, Revit, Unity3D, and photogrammetry have been used for precise and interactive reconstructions, particularly in projects like Historic Jeddah and the Canadian Parliament (Boeykens, 2011; Albourae et al., 2017). Recent studies demonstrate how HBIM and laser scanning are increasingly employed for VR-ready models in Unity3D and UE, combining heritage conservation with digital workflows (Pavelka & Michalík, 2019; Argiolas et al., 2022). However, interoperability remains a major challenge, as BIM's parametric models require conversion into mesh-based formats suitable for GEs, often leading to metadata loss and manual optimization requirements (Pybus et al., 2019; Ma, 2022). Scalability is another pressing concern, with large heritage data sets requiring LOD adjustments to ensure real-time performance in VR environments (Pavelka & Michalík, 2019; Argiolas et al., 2022). For instance, Pybus et al., (2019) study, which translated a heritage BIM of Canada's Parliament into Unity3D for VR visualization but faced difficulties in optimizing complex geometry and maintaining high visual fidelity while ensuring performance.

4.2. BIM integration

4.2.1. BIM-based simulation

Simulation-based applications have significantly improved real-time visualization, construction planning, and immersive collaboration. A major trend is the use of Unity3D and UE for 4D BIM simulations as shown in Table 2, allowing for enhanced constructability analysis, construction process modelling, and gamified deconstruction planning (Wu & Garcaí de Soto, 2022). Additionally, simulation-based approaches in heavy industrial lifting, temporary elevator logistics, and deconstruction planning have been increasingly integrated with BIM to enhance process efficiency (Sydora et al., 2021; Earle et al., 2023). However, several critical challenges hinder seamless BIM–GE integration. Interoperability remains a persistent issue, as BIM's structured parametric data does not transfer efficiently into the mesh-based environments of GE, leading to data loss, manual rework, and inconsistent model fidelity (Osorio-Sandoval et al., 2022). Another significant challenge is scalability, as large BIM data sets require high computational power for real-time rendering, necessitating LOD adjustments, pre-processing techniques, and AI-driven optimization to maintain performance (Klein et al., 2016). The complexity of 4D BIM-based simulations for constructability analysis also introduces barriers, as current workflows often lack automation for model transfer, requiring extensive manual adjustments and expert intervention. Additionally, real-time collaboration in VR-based simulations faces usability issues, as current implementations struggle with multi-user interactions, intuitive controls, and seamless data synchronization (Osorio-Sandoval et al., 2022). For instance, Boton et al., (2019) proposed an automated framework for integrating 4D BIM models into GE for virtual reality-based construction simulation. The study highlighted difficulties in maintaining data consistency during the BIM-to-GE transfer, requiring extensive manual interventions to correct errors in geometry, metadata retention, and simulation accuracy. These findings underscore the need for standardized BIM–GE conversion workflows.

4.2.2. BIM-based decision making

Decision-making has gained prominence in various domains, including energy-efficient retrofitting, cost estimation, pathfinding, and signage optimization. BIM–VR applications have been particularly effective in facilitating informed decision-making by incorporating human behaviour simulations, environmental impact assessments, and interactive signage placement tools (Tung et al., 2021). For instance, Yang et al. (2013) demonstrated how immersive virtual mock-ups improve multistakeholder decision-making in energy retrofit projects, enabling better evaluation of alternative designs and energy savings strategies. Similarly, Ju et al. (2023) assessed data conversion processes for BIM-based VR applications, offering valuable insights into compatibility issues affecting decision-support workflows. However, several challenges persist that hinder the seamless integration of BIM–GE in decision-making environments. A major challenge is data interoperability, where semantic information loss occurs during the transfer of BIM models into GE environments, affecting the accuracy of decision-making models (Ju et al., 2023). Additionally, Yang et al. (2013) highlighted the difficulty in developing reusable modules for simulation-based decision support, making it challenging to scale virtual mock-ups across different projects. Motamedi et al. (2017) further identified limitations in signage placement optimization, where current systems lack comprehensive methodologies for adaptive placement based on real-time user interactions. In the domain of cost estimation and pre-sold home customization, Tung et al. (2021) found that while BIM–VR integration allows for real-time cost estimation, manual adjustments are still required to update price changes dynamically.

An example of these challenges is Ju et al.’s (2023) study, which assessed BIM-to-VR data compatibility and found that some conversion workflows resulted in significant information loss, particularly for object metadata and spatial relationships. Similarly, Yang et al. (2013) found that traditional decision-making processes rely heavily on tabular data and lack effective integration with 3D environments, leading to inefficiencies in collaborative decision-making for energy retrofits. These challenges underscore the need for improved interoperability frameworks, automated data retention solutions, and AI-driven adaptive decision-support tools. Table 3 summarize all the study related to decision making by BIM and GE integration.

4.2.3. Modular construction and DfMA

Modular construction has gained attraction due to its ability to enhance collaboration, automation, and visualization throughout the design, manufacturing, logistics, and on-site assembly processes. Recent studies have explored the use of GE-based platforms to streamline decision-making, real-time collaboration, and digital twin applications in modular construction projects (Wong Chong & Zhang, 2019; Ezzeddine et al., 2021). Unity and UE have been increasingly adopted for process simulation, interactive visualization, and robotic integration in prefabricated construction environments (Jansson et al., 2018). One of the key advancements is the integration of modular construction workflows into immersive environments, allowing project teams to coordinate prefabrication, logistics, and assembly in a common virtual space (Potseluyko et al., 2022). Despite these advancements, several challenges hinder seamless BIM–GE integration in modular construction. Coordination and collaboration issues remain a significant concern, as project teams often work in silos, leading to misalignment between design, manufacturing, and assembly phases (Ezzeddine et al., 2021). Another major challenge is automation and robotics integration, where modular construction still relies heavily on manual labour, and efforts to introduce robotic-assisted prefabrication have encountered difficulties in workflow standardization and real-time data synchronization (Wong et al., 2019). A major interoperability challenge is that modular group parameters, such as modular group ID, component relationships, dependencies, and fabrication constraints, do not seamlessly transfer from BIM to GE, resulting in metadata loss and requiring extensive manual rework (Jansson et al., 2018). This limitation affects the ability to automate simulations and real-time progress monitoring, reducing the efficiency of BIM–GE applications in modular construction planning and execution. Visualization and real-time performance present another challenge, as high-poly BIM models need to be optimized for GE, necessitating LOD techniques and AI-driven data simplification to maintain real-time interactivity without sacrificing model accuracy (Potseluyko et al., 2022). A case study of these challenges is Wong et al. (2019) study, which proposed a game-based simulation model to analyse the integration of robotics in modular construction workflows. While the study demonstrated the potential for GE to simulate modular assembly processes and automate repetitive construction tasks, it also highlighted significant interoperability barriers, as existing BIM formats were not fully compatible with the simulation environment, requiring extensive manual adjustments (Rehman et al., 2022).

4.2.4. Data generation and management

Data generation and management have seen increasing adoption in areas such as synthetic data creation, construction site monitoring, human behaviour analysis, and virtual experiment systems. Latest trends summarized in Table 4 emphasize leveraging GE for automated synthetic image generation to train AI models for scene understanding, object recognition, and safety monitoring (Wei et al., 2021; Lee et al., 2023). Additionally, Unity and UE have facilitated automated data fusion, progress tracking, and AI-enhanced facility management workflows in VR (Fritsch et al., 2017). These advances support digital twins, predictive analytics, and AI-assisted decision-making, demonstrating the potential for scalable, real-time data generation and management in construction and facility operations (Graebling et al., 2022). However, a major issue is interoperability, as current workflows struggle to retain semantic metadata during BIM-to-GE conversion, particularly in scene understanding applications (Wei et al., 2021). These data loss affect object recognition and AI training models, requiring manual adjustments and additional labelling efforts to maintain accuracy. Another challenge is scalability, as large data sets from real-world construction sites create computational bottlenecks, making real-time data visualization and synthetic data set generation resource-intensive (Lee et al., 2023). Additionally, human behaviour data collection in virtual simulations faces limitations in data accuracy and user interaction fidelity, impacting the reliability of evacuation modelling, worker safety training, and behavioural analytics (Zhang et al., 2016). Furthermore, virtual experiment systems designed for environmental research have encountered barriers in integrating heterogeneous data sets into a unified visualization framework, leading to high technical entry barriers for non-experts (Graebling et al., 2022).

A concrete example of these challenges is Wei et al. (2021) study, which attempted to generate synthetic image data sets using BIM and UE for AI-driven semantic scene understanding. Their research highlighted data inconsistencies when transferring object attributes from BIM to GE environments, requiring custom scripting and manual intervention to maintain classification accuracy. Similarly, Graebling et al. (2022) developed a virtual experiment information system, but found that the integration of simulation results with real-time data visualization required extensive preprocessing and software adaptation.

4.3. Simulation and analysis

4.3.1. Emergency evacuation

Emergency evacuation planning has gained traction for its ability to simulate real-time fire scenarios, improve occupant safety, and optimize evacuation strategies. Recent advancements have focused on digital twin frameworks, AI-driven evacuation modelling, and immersive serious gaming to enhance fire safety planning and emergency decision-making (Wang et al., 2013). Studies have leveraged BIM- and GE-based simulations to create real-time emergency evacuation models, helping first responders and facility managers visualize emergency scenarios dynamically (Liu et al., 2014). AIoT (artificial intelligence of things)-enabled digital twins to have also been integrated into fire management systems, particularly for enclosed environments like tunnels, allowing real-time sensor-based fire detection and evacuation route optimization (Zhang et al., 2024). However, a major issue is real-time data processing and sensor deployment, where IoT-based fire detection systems require substantial investment and suffer from low cost-efficiency, particularly for large-scale environments such as tunnels. Another challenge is data coverage and accuracy, as sensor data collected during fire incidents may be incomplete or delayed, impacting real-time evacuation decision-making. Additionally, serious gaming-based evacuation models face behavioural realism issues, as participants in virtual evacuation drills do not always replicate real-world responses, leading to potential inaccuracies in evacuation modelling (Liu et al., 2014). GE-based simulations also struggle with human behaviour variability, requiring extensive data collection and AI-driven modelling to enhance evacuation prediction accuracy (Wang et al., 2013).

A concrete example of these challenges is Liu et al.’s (2014) study, which developed a BIM-based serious game for human behaviour simulation during emergency evacuations. Their findings highlighted difficulties in accurately modelling human decision-making under emergency conditions, as virtual environments lack the unpredictability of real-life evacuations, requiring extensive validation through real-world behavioural data. Similarly, Zhang et al. (2024) introduced an AIoT-enabled digital twin for tunnel fire safety but faced challenges in real-time sensor data synchronization and fire prediction accuracy, necessitating improvements in AI-based modelling and sensor network coverage.

4.3.2. DES

DES has been increasingly applied in construction planning, process optimization, and resource management. Researchers have explored the combination of BIM and DES within Unity to provide interactive 3D visualizations and decision-support tools that help stakeholders analyse construction activity durations, cost implications, and process efficiency (Osorio-Sandoval et al., 2018, 2020). By incorporating DES into BIM–GE workflows, project managers can simulate different construction strategies, evaluate resource allocations, and optimize scheduling based on real-time interactions (Osorio-Sandoval et al., 2020). These advancements have improved the accessibility of DES tools by making simulation results more visually intuitive and interactive, allowing for real-time analysis and stakeholder engagement in construction planning. Despite these benefits, a major challenge is interoperability, as BIM-to-GE data conversion often fails to retain crucial simulation parameters, requiring manual adjustments to reconstruct DES models in GE environments (Sandoval et al., 2018). Another critical issue is the difficulty of automating simulation workflows, as current DES models in BIM lack standardization for automated integration into real-time game environments, making it time-consuming to define activity sequences and resource dependencies (Osorio-Sandoval et al., 2020). Additionally, the visualization of DES results within GE presents a challenge, as most existing BIM-based DES workflows rely on post-processing animation rather than real-time, interactive simulations (Osorio-Sandoval et al., 2020). This limitation reduces user flexibility in modifying simulation parameters dynamically, impacting decision-making efficiency.

4.4. Project management

4.4.1. Design review

Design review has significantly improved collaborative visualization, real-time evaluation, and stakeholder engagement in architectural and construction projects. Recent studies have demonstrated that GE provides immersive environments that enhance design validation, error detection, and iterative feedback loops (Kumar et al., 2011). The application of mixed reality interfaces has further facilitated interactive design review sessions, enabling stakeholders to navigate and manipulate 3D models in real-time (Behmel et al., 2014). Additionally, VR-based design evaluation systems have been applied to assess building exterior preferences, allowing users to modify and select architectural elements dynamically within a GE environment (Liu et al., 2023b). Despite these advancements, current workflows rely on intermediate file formats (e.g. FBX, OBJ) that do not fully retain parametric data, material properties, or metadata during the transfer process (Kado et al., 2018). This results in information loss, requiring manual rework and reconfiguration in the GE. Another challenge is real-time collaboration, as most BIM–GE frameworks lack efficient networking capabilities for multi-user design reviews, limiting real-time stakeholder interaction and decision-making (Shiratuddin et al., 2011). Additionally, GE does not inherently support BIM parametric behaviours, meaning elements such as doors, windows, and furniture lack embedded intelligence once transferred, requiring custom scripting or external plug-ins to restore interactive functionality (Kumar et al., 2011). Furthermore, design review simulations often suffer from usability limitations, where non-expert users struggle with complex navigation controls, reducing the effectiveness of immersive evaluations (Behmel et al., 2014). A concrete example of these challenges is Kado et al. (2018) study, which developed a bidirectional data exchange system. While the system enabled bidirectional updates between BIM and GE, it struggled with synchronizing parametric behaviours and material properties, requiring manual intervention to maintain design integrity.

4.4.2. Stakeholders collaboration

Stakeholder collaboration and interactive visualization are crucial for improving project outcomes in the AEC industry. Traditional static 2D methods often lead to miscommunication and delays, highlighting the need for dynamic, immersive approaches. Integrating BIM with GEs like Unity and UE has demonstrated transformative potential in construction management, healthcare, education, and urban planning. Studies on virtual site visits and construction management (Bille et al., 2014; Balali et al., 2020; Dinis et al., 2020) reported improved communication and usability, while design communication research (Wu & Kaushik, 2015b; Huang et al., 2017; Lin et al., 2018; Ehab et al., 2023) showed enhanced stakeholder engagement, despite challenges like large file sizes. Collaborative environments (Prabhakaran et al., 2022; Panya et al., 2023) facilitated immersive urban design and proposed a BIM-based metaverse for future interconnected workspaces. Educational studies (Tayeh et al., 2019, 2020; Fonseca et al., 2020) reported improved collaboration and learning, despite initial setup challenges. BIM–AR integration (Yoon et al., 2023; Tan et al., 2024) reduced errors in civil engineering education. Healthcare applications (Figueres-Munoz et al., 2015; Nandavar et al., 2018) improved training and layout planning via real-time interaction. Design review and cost estimation studies (Balali et al., 2018; Ehab et al., 2023) enhanced real-time cost updates and design change management. These findings emphasize the shift from static methods to immersive, real-time virtual environments, fostering iterative design processes, enhancing communication, reducing errors, and streamlining project management.

4.5. Education and training

4.5.1. Engineering education and training

Traditional educational methods often struggle to bridge the gap between theoretical learning and practical application, particularly in construction and healthcare fields. Unity 3D and Revit improved staff training and spatial understanding in a Norwegian hospital project (Merschbrock et al., 2016) and enhanced spatial conflict detection in precast element installation at Polytechnic University of Marche (Messi et al., 2021a). Educational initiatives integrating Unity into construction management (Wu & Kaushik, 2015a) and civil engineering curricula (Boga et al., 2017) increased engagement and learning effectiveness. Pre-university students at the University of Algarve (Dinis et al., 2017) and VR explorations like Farnsworth House further improved understanding. Fire simulation training (Diez et al., 2016) and design review projects (Sanchez et al., 2022) demonstrated practical benefits for professionals and students. These integrations provide dynamic, interactive environments that enhance spatial awareness, decision-making, and hands-on training, making them transformative tools in construction and design education.

4.6. Facility management

4.6.1. Facility management

Facility management has significantly improved real-time monitoring, asset tracking, and predictive maintenance in building operations. Recent studies summarized in Table 5 demonstrate that digital twins, powered by BIM–GE frameworks, enable data-driven decision-making and enhanced spatial awareness in facility management tasks (Yan et al., 2021). Multi-user virtual environments developed using Unity facilitate collaborative facility inspections, remote troubleshooting, and maintenance scheduling (Shi et al., 2016). Moreover, cloud-based integration of BIM and GE has allowed facility managers to access interactive 3D models linked to real-time IoT sensor data, improving the efficiency of building operations and energy management (Yan et al., 2021). Despite these advancements, existing workflows struggle with real-time data exchange between BIM models and GE-based facility management platforms, leading to incomplete asset metadata retention and synchronization delays (Khalid et al., 2017). Another significant challenge is scalability, where large facility data sets require high computational power for real-time rendering, making it difficult to implement interactive facility management applications on cloud-based platforms (Yan et al., 2021). Additionally, multi-user collaboration remains a technical hurdle, as current BIM–GE systems lack robust networking capabilities for simultaneous remote facility management interactions, limiting their effectiveness in large-scale operations (Shi et al., 2016). Furthermore, data security and privacy concerns arise when integrating IoT-based facility monitoring systems with BIM–GE applications, requiring advanced encryption and access control mechanisms (Yan et al., 2021).

A concrete example of these challenges is Khalid et al.’s (2017) study, which developed a real-time BIM-based facility management system integrating a building management system (BMS) with Unity. While the system successfully provided interactive facility visualization and real-time asset monitoring, it faced interoperability issues in transmitting sensor data to the BIM–GE model, requiring custom web socket configurations and extensive manual adjustments to maintain synchronization. Similarly, Yan et al. (2021) introduced a web-based BIM asset and facility management system but struggled with integrating multistakeholder facility data due to disparate system architectures and lack of standardized BIM–GE interoperability protocols.

4.6.2. Daylighting, energy usage, and sustainability

Daylighting, energy usage, and sustainability have significantly enhanced building performance analysis, real-time energy monitoring, and interactive visualization. Recent studies demonstrate that GE provides immersive environments for energy benchmarking, daylighting optimization, and power consumption tracking (Natephra et al., 2017). VR-based energy simulation tools enable stakeholders to explore and assess lighting conditions and energy use dynamically, bridging the gap between numerical energy analysis and interactive design validation (Natephra et al., 2017). Furthermore, BIM-enabled power consumption platforms have been developed to analyse energy usage patterns in real-time, allowing for more efficient energy management and sustainability decision-making (Chiang et al., 2015). Despite these advancements, energy simulation data often lacks seamless integration with BIM models, requiring extensive preprocessing for visualization in Ges (Woo et al., 2016). Another challenge is real-time data synchronization, where existing BIM–GE frameworks struggle to integrate real-time IoT sensor data into interactive energy simulations, leading to delays in energy performance assessment and inefficient decision-making (Chiang et al., 2015). User interaction limitations also pose challenges, as lighting simulation tools often rely on static 2D visualizations, making it difficult for non-experts to interpret energy feedback in an immersive setting (Natephra et al., 2017). Additionally, scalability concerns arise in large-scale energy benchmarking projects, where GEs experience computational strain when processing detailed BIM energy data sets, reducing the efficiency of real-time energy performance monitoring (Woo et al., 2016).

4.7. Construction management

4.7.1. Construction safety

Construction safety has significantly improved hazard recognition, real-time risk assessment, and immersive safety training. Studies shown in Table 6 demonstrate that GE interactive simulations for hazard detection, site inspections, and accident prevention (Park et al., 2013; Hussain et al., 2024a). VR and AR technologies have been used to create safety training programs that improve worker awareness and reduce the occurrence of accidents (Lin et al., 2011). Conversational AI-powered safety training modules have also been developed, enabling real-time, multilingual, and adaptive knowledge transfer, particularly for migrant workers facing language barriers (Hussain et al., 2024a). Furthermore, 4D BIM-GE models have been employed to simulate and predict unsafe working conditions, allowing safety managers to proactively address potential risks (Zaman et al., 2024). Despite these advancements, existing BIM data often lack direct compatibility with GE, leading to manual data conversion and loss of parametric information. Another challenge is real-time risk monitoring, where safety simulations are typically static and do not update dynamically based on real-time site conditions, limiting their effectiveness in active hazard management (Zaman et al., 2024). Scalability and computational limitations also pose challenges, as high-fidelity safety models require significant processing power, making it difficult to deploy them in large construction projects (Hussain et al., 2024b). Additionally, worker engagement in VR-based safety training remains an issue, as some users struggle with immersive controls and may not fully retain safety knowledge from virtual experiences (Lin et al., 2011).

4.7.2. Monitoring, inspection, and quality control

Monitoring, inspection, and quality control has significantly enhanced real-time tracking, automated defect detection, and progress assessment in construction and infrastructure management. Recent studies shown in Table 7 demonstrate that GEs facilitate automated data visualization, inspection planning, and interactive quality control processes (Rahimian et al., 2020; Huang et al., 2023). Drone-based inspections combined with BIM-supported coverage path planning have improved efficiency and safety in exterior building assessments (Huang et al., 2023). Similarly, AI-driven image processing and ML have been incorporated into BIM–GE workflows, enabling automated monitoring of construction progress and deviation detection (Rahimian et al., 2020). Additionally, AR and MR platforms are increasingly being used for interactive defect detection, allowing inspectors to overlay as-designed BIM models onto real-world structures for accurate quality control (Jeon et al., 2023; Riedlinger et al., 2022). Despite these advancements, BIM models require extensive pre-processing before integration into GE and external monitoring systems, leading to data loss and manual adjustments (Rahimian et al., 2020). Automated defect detection also faces accuracy limitations, as ML models require extensive training data to correctly classify construction deviations, impacting reliability in progress tracking (Rahimian et al., 2020). Additionally, real-time synchronization of BIM with IoT and sensor data is still limited, as many BIM–GE implementations rely on static snapshots instead of continuous updates, reducing their effectiveness in dynamic progress assessment (Xiong et al., 2018). Scalability concerns also persist, as high-resolution as-built models and point cloud data require significant computational power, leading to latency in rendering and performance bottlenecks in large-scale construction projects (Abreu et al., 2023).

4.7.3. Spatial conflict

Efficient space allocation is essential for construction success, yet traditional static methods fail to address dynamic, real-time site complexities like machinery movement and personnel shifts. To overcome these limitations, advanced technologies integrated with BIM offer dynamic workspace management and spatial conflict resolution. This synthesis examines four papers focused on dynamic workspace management and spatial conflict resolution using BIM and GE technologies. Messi et al. (2021a, b, 2022a, b, 2023) introduces a BIM-based serious game within a high-level system architecture to enhance work progress management. It predicts spatial conflicts in real-time, particularly useful in retrofitting existing structures, leading to more responsive and effective site operations.

5. Synthesis of Findings

The meta-analysis offered quantitative insights into the technologies employed and identified research trends, while the meta-synthesis provided a comparative analysis of different studies within related subdomains. This section consolidates the findings of both the analysis.

5.1. Data interoperability in BIM authoring tools and GEs

Data requirements for integration evolve with project phases: initial design planning relies on site and survey data, while final facility management depends on detailed 3D building models generated using BIM tools like Revit, ArchiCAD, and QGIS. Supplementary tools like SketchUp, Blender, and 3ds Max, along with surveying and 3D scanning, enhance model detail. Interoperability between BIM tools and GEs often relies on exchange formats due to limited direct integration capabilities. Figure 7 shows a pipeline for converting BIM data into formats like FBX, CSV, and IFC using tools such as Navisworks, Dynamo, and CityEngine for compatibility with Unity and visualization tools. QGIS and SketchUp use intermediary software like Datasmith and 3ds Max to ensure seamless data transfer, enhancing collaboration and visualization.

5.2. Distinct integration approaches

A Sankey diagram (Figure 8) illustrates the relationships between GEs, their subdomains, and integration formats, offering a concise overview of diverse integration approaches. Unity and UE dominate due to their versatility and robust features. Key to these integrations are data exchange formats like FBX, IFC, CSV, and OBJ, which ensure seamless interoperability between BIM tools and GEs. These formats are crucial for bridging BIM and gaming technologies, enabling effective integration and application across domains.

5.3. Generic framework for integration of BIM and GE

Importing BIM data into a GE is the first step in the integration process, requiring significant manual processing to achieve specific goals. This includes template selection, asset importing, scene configuration, and adjustments to lighting, shading, textures, and animation. For high-fidelity visualizations, lighting and texture refinement ensures realism, while animations are critical for time-based BIM simulations. A UI is then developed to enable interaction with the model, followed by interactivity coding and rigorous testing to ensure a robust visualization environment. AR/VR technologies enhance user engagement with interactive dashboards and exploration modes. The system is published on platforms like Windows, mobile devices, and web applications for stakeholder feedback. Stakeholder input guides iterative updates, as shown in Figure 9, ensuring continuous refinement and alignment with specific requirements for an effective BIM–GE integration.

5.4. Challenges and research gap

The integration of BIM and GE has advanced visualization, simulation, and interactivity in the AEC industry, but challenges persist. Real-time optimization is hindered by complex, large BIM models, with format conversions like IFC to FBX leading to data loss and inaccuracies (Argiolas et al., 2022). Interoperability issues, platform-specific limitations, and inconsistent transformations across formats (FBX, IFC, and VRML) further complicate workflows (Pooladvand et al., 2021; Edirisinghe & Woo, 2021). User interface (UI) complexity limits adoption, especially for non-technical users (Phan et al., 2020). Balancing high-resolution graphics with processing efficiency is challenging in immersive environments, affecting real-time applications such as cultural heritage and construction hazard simulations (Cicekci et al., 2014; Li et al., 2024). Techniques like level-of-detail management improve performance but can compromise visual fidelity.

Modular construction and DfMA (design for manufacturing and assembly) face fragmented information flows, limited robotics use, and a lack of methodologies to assess productivity gains from automation (Ezzeddine et al., 2021). SMEs struggle with cost and time constraints, hindering adoption of real-time technologies (Potseluyko et al., 2022). Real-time data integration faces latency and synchronization issues with dynamic data like GPS (Chen et al., 2022). Existing tools have improved integration but fall short in preserving data integrity or achieving widespread adoption (Yoon et al., 2023).

Solutions include standardized protocols, AI-driven data mapping, and middleware to enhance interoperability and reduce manual efforts (Rahimian et al., 2020). Advanced simulation algorithms, modular UI designs, and open-source SaaS models can improve accessibility, especially for SMEs (Hussain et al., 2024a). Enhancing multi-user frameworks and wearable AR/VR technologies is critical for fostering interactivity and collaboration (Hussain et al., 2024a). Table 8 represents a comparative analysis of the current capabilities, desired capabilities and advancements, and existing research gaps in integrating BIM with GEs across several key aspects.

5.5. Future direction

Building on the advancements and challenges discussed, this section outlines both short- and long-term research priorities for integrating BIM with GEs. Short-term strategies focus on practical steps such as middleware solutions, SaaS-based cost models, incremental UI improvements, and basic AI-driven data mapping to address immediate interoperability, cost, and user experience challenges, particularly for SMEs. These measures can quickly lower adoption barriers, improve collaboration, and demonstrate tangible returns on investment. Long-term strategies call for a holistic transformation, leveraging emerging technologies (generative AI, reinforcement learning, and advanced robotics) and robust frameworks (semantic web ontologies and distributed computing) to create fully integrated BIM–GE ecosystems. Figure 10 presents a chronological roadmap for BIM–GE research, with short-term goals on the left and long-term objectives on the right. Each box builds on the outcomes of the previous one, starting with data interoperability as the foundational step that enables real-time simulation. With live project updates in place, the focus then shifts to UI improvements and cost reduction, making BIM–GE tools more accessible particularly for SMEs. As projects become more complex, scalability and multi-user collaboration ensure performance and seamless teamwork. Further stages introduce automation and robotics integration, paving the way for fully integrated DfMA and digital twin ecosystems. This flow reflects a natural evolution from basic interoperability to advanced, AI-driven construction and lifecycle management solutions. Figure 10 shows the roadmap for the future research direction with short- and long-term research priorities displayed from top to bottom, respectively.

5.5.1. Short-term research priorities

Interoperability

Interoperability between BIM tools and GEs can be divided into four main modules, 3D data exchange, metadata exchange, semantic data exchange, and bidirectional exchange, all of which face significant challenges but also have promising solutions.

3D data exchange can be enhanced by using formats like OBJ and FBX, which not only captures basic geometry but can also store colour and material attributes. Software such as 3ds Max can optimize FBX files, reduce polygon counts or consolidate texture data before importing them into a GE, thus preserving essential visual details while improving real-time performance. In addition, developing a custom plugin that directly exports BIM models into exchange formats (e.g. FBX and OBJ) can streamline workflows by automating the conversion process and ensuring metadata (colour, materials, and texture coordinates) is accurately carried over. Once imported into the GEs, AI-driven generative models, such as stable diffusion or GPT-based APIs, can further enrich or recreate textures and materials that may be lost or partially degraded during the conversion. These AI tools can automatically generate high-fidelity or context-aware materials for BIM elements like walls, floors, and furniture, offering an LOD tailored to the specific aesthetic or functional requirements of a project.

Beyond geometric and visual information, BIM models contain metadata such as element IDs, dimensions, and material properties that inform the design and facility management processes. This metadata often becomes truncated or lost when converting from BIM-authoring software to GEs that focus primarily on 3D rendering. To preserve this information, it can be exported to structured formats like CSV or JSON, with each object assigned a unique identifier corresponding to the 3D data. Within the GE, scripts or plugins can parse this structured data and dynamically map it to model elements, thereby providing interactive metadata queries, reporting functionality, and the capacity for more informed decision-making in virtual environments.

Semantic data exchange goes one step further by ensuring that contextual relationships such as spatial hierarchy, or building codes are retained in the transition to a GEs. BIM standards like IFC inherently store these rich relationships, but many 3D-centric formats disregard them. Approaches using linked data and ontologies (e.g. RDF and OWL) can maintain these connections, translating the BIM semantics into a format the GE can interpret and potentially manipulate. NLP techniques further automate schema matching, allowing GEs to identify element categories (e.g. ‘Wall’ versus ‘Column’) and maintain property consistency. By retaining semantic linkages, immersive environments in the GE can provide not only realistic visuals but also data-driven interactions, where users can interrogate and modify objects with full awareness of their architectural context.

Finally, real-time collaboration and iterative design demand more than a one-way flow of data. Bidirectional exchange ensures that updates in the BIM tool such as altered room layouts or revised materials, sync with the GE, and likewise, modifications in the GE (e.g. adjusting the height of a wall in a VR environment) propagate back into the BIM model. This requires robust integration layers or middleware that listen for changes and push them across platforms via APIs, WebSockets, or event-driven protocols. Employing version control repositories tailored to BIM data can track these cross-platform edits, while scripting within the GE (e.g. C# in Unity) applies updates to the corresponding 3D objects. By establishing and automating these feedback loops, architects, engineers, and stakeholders can work in a synchronized ecosystem, enhancing both the speed and quality of the design process.

Going forward, researchers and practitioners should focus on refining these interoperability pipelines. Exporting BIM metadata in structured formats that complement optimized 3D files is particularly promising, as it allows GEs to dynamically map semantic details onto geometric models at runtime. Figure 11 illustrates how a combined workflow spanning 3D data exchange, geometric data exchange, semantic data exchange, and bidirectional exchange achieves seamless integration between BIM and GEs.

Real-time simulation

Real-time simulation in BIM–GE integration is often limited by the manual processes required to update models, highlighting the need for systems that adapt to project changes automatically (Boeykens et al., 2011). Emerging technologies like reinforcement learning and decision tree algorithms can predict upcoming modifications and optimize workflows, mitigating delays, and cost over-runs. While these AI-driven approaches show promise, more concrete research pathways are required to translate predictive analytics into practical, real-world solutions that large construction firms and SMEs alike can adopt. For example, generative AI could generate project scenarios for contingency planning, enabling stakeholders to weigh alternatives before issues escalate on-site. Developing unified data models that incorporate schedules, inventories, and spatial data can help maintain consistent simulations as parameters evolve (Binni et al., 2023). XR technologies further enhance these simulations by providing immersive AR and VR experiences, enabling users to visualize current project statuses and forecast performance within realistic environments (Wu & Garcaí de Soto, 2022). Recommendations for future research include implementing cloud-based simulation dashboards that integrate seamlessly with BIM data repositories. By leveraging cost-effective SaaS models and open-source visualization engines, SMEs could also benefit from these dynamic simulation capabilities without the burden of extensive in-house infrastructure.

UI

Developing a UI that supports both expert and non-expert users is critical for BIM–GE platforms to gain wider acceptance (Kim et al., 2024). While current systems offer advanced visualization capabilities, they often overwhelm users with complex menus and fragmented workflows. To address these challenges, a user-centric design approach that includes surveys, interviews, and iterative usability testing can elucidate specific pain points. Integrating large language models like GPT for natural language queries and voice commands can further streamline interactions by minimizing menu-based navigation (Hussain et al., 2024a). Likewise, standardizing UI components across different BIM–GE applications would reduce the learning curve, particularly for smaller firms that cannot afford extensive training programs (Boga et al., 2017). Concrete solutions might involve modular design systems that adapt to the needs of varying project types, from large infrastructure developments to single-floor renovations. User training could also incorporate gamification, guiding participants through scenario-based tutorials and rewarding proficiency to maintain engagement (Castronovo et al., 2019). Ultimately, an industry-wide collaboration, possibly led by professional associations and software consortia, is necessary to establish universal design frameworks that make BIM–GE UIs more consistent, inclusive, and accessible for a broader audience.

Cost and accessibility

Excessive costs and limited accessibility are significant barriers to widespread BIM–GE integration, especially for SMEs seeking to leverage these technologies without incurring prohibitive expenses (Zhang et al., 2024). While cloud computing and SaaS models can offer pay-as-you-go solutions that reduce upfront expenditures, more granular pricing tiers and financing options could further ease adoption among smaller businesses (Balali et al., 2018). Open-source software initiatives, supported through academic-industry partnerships, can also cultivate affordable alternatives to proprietary platforms (Anifowose et al., 2023). However, bridging the gap between open-source solutions and professional-grade functionalities requires a robust community of developers, regular updates, and comprehensive training resources. Future solutions should include modular, interoperable toolkits capable of integrating with existing systems, minimizing costly overhauls and downtime. Leveraging automation—such as generative AI for rapid design prototyping or robotics for on-site tasks—could offset investments by accelerating project timelines and reducing labour costs. Additionally, targeted public-private funding programs can catalyze the initial adoption of these technologies among smaller firms, ensuring cost-sharing mechanisms that lower risk (Potseluyko et al., 2022). By pursuing these strategies concurrently, the industry can foster an ecosystem where BIM–GE tools are both technologically advanced and economically accessible.

The economic feasibility of integrating BIM with GE technologies for SMEs demands a nuanced analysis that extends beyond the conventional focus on large-scale projects, as SMEs face distinct financial, operational, and technical challenges. While large firms may capitalize on economies of scale to absorb the initial capital outlay associated with advanced digital tools, SMEs often operate under tighter budget constraints and limited technical resources, making the upfront investment in software licenses, hardware upgrades, and specialized training a significant barrier. However, recent advances such as modular software architectures such as Autodesk Revit plugins, cloud-based platforms such as BIM 360, and open-source solutions such as blender and Gadot, have begun to lower entry costs and offer scalable integration options that can be tailored to the specific needs and project scopes of SMEs. A comprehensive cost-benefit analysis indicates that despite the initial expenditure, the potential for enhanced visualization, real-time simulation, improved stakeholder communication, and reduced error margins can translate into long-term savings by minimizing costly design revisions and construction delays. Furthermore, the integration of BIM with GE can facilitate more agile project management and faster decision-making, thereby bolstering competitive advantage in markets where speed and accuracy are paramount. Nonetheless, SMEs must carefully evaluate factors such as the learning curve, system interoperability, and ongoing maintenance requirements to ensure that the anticipated return on investment justifies the integration effort. Ultimately, a balanced view reveals that, under appropriate conditions and with a strategic implementation framework, the integration of BIM and GE technologies can be economically viable for SMEs, offering a transformative toolset that aligns with their scale and operational realities while mitigating risks typically associated with large-scale digital transformations.

Interactivity

Interactivity is central to effective BIM–GE integration, especially in multi-user VR and AR settings where real-time collaboration is paramount (Tung et al., 2021). While advanced networking and distributed computing solutions have started reducing latency, more robust frameworks are needed to handle synchronization across geographically dispersed users. AI-driven features ranging from predictive user guidance to generative content that populates virtual environments could significantly improve engagement and efficiency. For example, reinforcement learning can adapt interactions based on user behaviour patterns, offering context-specific tools or shortcuts to streamline tasks. Integrating intuitive input methods such as gesture recognition, voice commands, and haptic feedback further lowers the usability barrier and broadens the appeal of BIM–GE platforms to a diverse workforce (Hussain et al., 2024a). There remains a need for scalable, cost-efficient solutions that can be deployed not only in high-budget projects but also in smaller settings. High-fidelity graphics and realistic physics simulations, powered by GPU-based computing, could be offered as cloud-rendered services on a subscription basis, lowering the hardware requirements for SMEs. Meanwhile, AR overlays on wearable devices provide on-site personnel with immediate access to updated models, fostering stronger coordination and reducing errors. By investing in these interactive technologies, the AEC industry stands to benefit from more dynamic, engaging, and accurate collaborative processes.

5.5.2. Long-term research priorities

Scalability

Scalability challenges arise when handling large-scale BIM models with extensive data, often causing bottlenecks that affect real-time interaction and multistakeholder collaboration (Klein et al., 2016). Although advanced data compression and streaming algorithms such as IFC spatial compression and edge breaker techniques offer partial relief (Abreu et al., 2023), further research is needed to integrate generative AI approaches that intelligently compress and reconstruct data without compromising fidelity. Distributed computing and cloud-based solutions can also be expanded to include dynamic load balancing, ensuring computational resources scale in response to project size and complexity (Nandavar et al., 2018). For smaller firms, these strategies might be delivered through SaaS models, aligning costs with actual usage rather than requiring significant capital investments. Adaptive LOD algorithms, capable of anticipating user interactions, can render only the most relevant parts of the model whether in 2D or immersive 3D environments to maintain performance (Rehman et al., 2023). Finally, predictive algorithms that preload relevant data (e.g. structural elements on the next floor of a building) can reduce latency. Future research should thus focus on developing unified frameworks that combine compression, streaming, and predictive loading, making large-scale BIM–GE projects manageable for both industry giants and SMEs.

Automation technologies

Despite its proven benefits in manufacturing, the integration of robotics and automation technologies remains nascent in the AEC sector, particularly in synergy with BIM and GEs (Chen et al., 2022). Realizing a fully automated construction pipeline requires robust simulation platforms potentially powered by reinforcement learning and generative AI to validate robotic tasks in virtual environments before physical deployment. Here, BIM provides the foundational data layer, including precise geometry, material attributes, and parametric relationships of built assets, while GEs offer real-time rendering, physics simulation, and interactive interfaces. When combined, these technologies establish a dynamic ‘digital twin’ of the construction site, enabling robotic simulation, AI-driven optimization, and adaptive on-site autonomy. This approach minimizes errors, enhances safety, and optimizes workflows by allowing teams to test construction sequences, detect collisions, and refine scheduling virtually before implementing them on-site. In addition, bridging standards for data exchange between BIM software and robotic control systems is a key priority; unified APIs can simplify interoperability, supporting not only large corporations but also smaller-scale operations equipped with cost-efficient robotics solutions. While digital twins have proven their worth in manufacturing for optimizing assembly lines, applying this concept to construction can similarly improve resource allocation and reduce project. Concrete metrics such as productivity rates, error margins, and on-site safety improvements are needed to establish benchmarks for the effectiveness of robotic systems. Moreover, improving the usability of robotic programming through gesture controls, voice interfaces, or AI-assisted planning is vital for broader adoption among construction crews (Hussain et al., 2024a). By integrating these emerging technologies within BIM–GE ecosystems along with clearer standards, performance metrics, and user-friendly interfaces, automation in construction can become both feasible and economically viable for projects of varying scales.

DfMA

While GEs have been widely used for visualization, their full potential in DfMA for modular construction remains underexplored, particularly in real-time collaborative workflows. Integrating BIM–GE platforms can enable seamless coordination across all stages; design, manufacturing, logistics, assembly, and operations, allowing stakeholders to work within a shared virtual environment. For design and planning, immersive VR/AR-based coordination can enhance stakeholder engagement, enabling real-time modifications and clash detection. Manufacturing workflows can be optimized using interactive simulations, where AI-driven generative design and ML refine manufacturability, minimize material waste, and improve efficiency (Huang et al., 2023). In logistics and transportation, real-time 4D BIM simulations can enhance scheduling, optimize routes, and prevent supply chain bottlenecks, ensuring smoother module delivery. AR-assisted assembly can guide installation crews with overlayed assembly instructions and conflict detection, reducing errors and rework. At the operations and maintenance stage, digital twins combined with IoT sensor networks provide continuous monitoring, predictive maintenance, and AI-driven decision-making for lifecycle management.

Collaborative cloud-based BIM–GE platforms can further enhance telepresence-style collaboration, offering shared virtual workspaces, real-time annotations, and voice/video conferencing. Scalable subscription-based solutions can make these technologies accessible even for smaller modular projects. By fully integrating BIM and GEs across all DfMA stages, the construction industry can move beyond visualization, leveraging real-time simulation, automation, and AI-driven insights for more efficient, cost-effective, and data-driven modular construction. Figure 12 illustrates how BIM and GE technologies can be applied across the DfMA lifecycle. It showcases key use cases, including immersive design coordination, interactive workflow simulation, AR-assisted installation, and digital twins for maintenance and management.

6. Conclusions

This study provides a holistic examination of BIM and GE integration, elucidating how these technologies can significantly enhance productivity, decision-making, and collaboration across all construction phases. By systematically categorizing BIM–GE applications into design visualization, modular construction workflows, real-time simulations, and project management, the research highlights the varied yet interrelated domains through which BIM–GE integration can drive meaningful improvements. Notably, simulations and immersive visualization tools, facilitated by platforms like Unity and UE emerged as the most mature and widely adopted applications, supporting enhanced design validation, more informed stakeholder engagement, and dynamic scenario testing throughout project lifecycles.

Despite these promising developments, critical challenges persist. Interoperability and data exchange limitations, coupled with scalability issues, inhibit broader implementation, particularly in areas like DfMA and modular construction, where the technology's transformative potential remains untapped. While XR, ML, and automation have begun to supplement BIM–GE workflows, the study found that the integration of advanced technologies such as generative AI, robotics, and reinforcement learning remains underexplored. These gaps, influenced by complexity, cost, and the absence of standardized protocols, continue to restrict seamless, real-time collaboration and optimization.

To address these hurdles, the research puts forth actionable recommendations. These include the development of AI-driven tools for streamlined data mapping, establishment of robust industry-wide standard protocol for interoperability, and the creation of cost-effective solutions tailored to SMEs. Moreover, the study urges the exploration of advanced AI subfields to enhance real-time decision-making and scalability, as well as the implementation of multi-user interaction frameworks to facilitate immersive, synchronous collaboration among geographically dispersed teams.

While this work draws extensively from academic literature and places greater emphasis on methodological approaches rather than empirical outcomes, it lays a solid foundation for future research endeavors. Subsequent studies should validate the practical impacts of BIM–GE integration in real-world scenarios, diversify data sources beyond academic case studies, and systematically assess the influence of emerging technologies on industry-level adoption and performance metrics.

In summary, this paper underscores both the current capabilities and the latent potential of BIM–GE integration, serving as a critical reference for practitioners and researchers. By spotlighting key successes, identifying persistent barriers, and delineating clear avenues for technical and operational enhancements, it provides a roadmap toward more efficient, sustainable, and interactive construction processes that can drive innovation across the AEC sector.

Conflict of Interest

The authors stated that they have no conflicts of interest related to the research, writing, or publication of this article.

Author Contributions

Saddiq Ur Rehman: Conceptualization, Methodology, Software, Data curation, Writing-original draft, writing-Review & editing, Investigation, Formal Analysis. Inhan Kim: Conceptualization, Methodology, Funding acquisition, supervision, Project administration. Kyung-Eun Hwang: Conceptualization, Formal Analysis, Supervision, Validation, Investigation, Writing-review & editing, Methodology.

Funding

This work is supported in 2025 by the Korea Agency for Infrastructure Technology Advancement (KAIA) funded by the Ministry of Land, Infrastructure and Transport (grant RS-2021-KA163269).

Data Availability

No data were used in the preparation of this manuscript.

References