Development of a BIM-Based Metaverse Virtual World for Collaborative Architectural Design

Abstract

The rapid evolution of the metaverse is driving the development of new digital design tools that integrate Computer-Aided Design (CAD) and Building Information Modeling (BIM) technologies. Core technologies such as Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR) are increasingly combined with BIM to enhance collaboration and innovation in design and construction workflows. However, current BIM–VR integration often remains limited to isolated tasks, lacking persistent, multi-user environments that support continuous project collaboration. This study proposes a BIM-based Virtual World (VW) framework that addresses these limitations by creating an immersive, real-time collaborative platform for the Architecture, Engineering, and Construction (AEC) industry. The system enables multi-user access to BIM data through avatars, supports direct interaction with 3D models and associated metadata, and maintains a persistent virtual environment that evolves alongside project development. Key functionalities include interactive design controls, real-time decision-making support, and integrated training capabilities. A prototype was developed using Unreal Engine and supporting technologies to validate the approach. The results demonstrate improved interdisciplinary collaboration, reduced information loss during design iteration, and enhanced stakeholder engagement. This research highlights the potential of BIM-based Virtual Worlds to transform AEC collaboration by fostering an open, scalable ecosystem that bridges immersive environments with data-driven design and construction processes.

1. Introduction

As construction projects become increasingly complex and large-scale, architectural design teams face growing challenges in implementing Building Information Modeling (BIM)-based collaboration. The amount of required design and project information has drastically increased, placing new demands on collaborative workflows [1]. BIM, as a digitized representation of the building design process, combining 3D geometric modeling with construction documentation, has significantly improved communication and collaboration among project stakeholders. However, collaboration challenges persist despite the adoption of BIM, prompting the development of various technological solutions to enhance design coordination.

Extended Reality (XR) technologies encompassing Virtual Reality (VR), Augmented Reality (AR), and Mixed Reality (MR) have been introduced to support BIM-based collaboration throughout all phases of building projects [2,3]. While XR integration has shown potential in enhancing immersive project interaction, it has also introduced new technological barriers that limit effective collaboration. These challenges can be broadly categorized into technological, organizational, and industry-wide factors, each restricting BIM’s full potential in improving design efficiency, communication, and project delivery.

This study specifically addresses the technological limitations of BIM-based collaboration, with a focus on the integration of XR, particularly VR, with BIM for collaborative design activities. Although recent advancements have improved VR incorporation into BIM workflows, several critical issues remain unresolved:

Limited User Support: Current BIM–VR systems primarily support small-scale or individual interactions, typically limited to visualization and clash detection tasks. They lack the capacity for large-scale, real-time design collaboration across multidisciplinary project teams [4].

Data Loss and Version Management: The process of exporting BIM models into VR environments often leads to data loss, version conflicts, and redundant modeling efforts when updates occur, undermining collaborative efficiency [5].

Lack of Persistence and Continuity: Existing VR sessions are generally temporary and non-persistent, requiring repeated restarts for each collaborative task, which disrupts workflow continuity and leads to delays and increased cost [6].

System Complexity: Most VR–BIM systems are technically complex, requiring specialized expertise. This complexity limits participation from non-technical stakeholders, such as clients or contractors, who are often excluded from immersive design collaboration [7].

Fragmented Communication Tools: Current platforms lack integrated communication channels, relying on external chat or conferencing tools, which results in fragmented collaboration and slower decision-making processes [8].

To address these persistent technological challenges, this research proposes a BIM-based Virtual World (BIM–VW) as a core component of the AEC metaverse. The proposed framework introduces a persistent, immersive 3D environment specifically designed to support synchronous real-time, multi-user design collaboration. This system allows participants to directly engage with BIM objects and other users, collaboratively manage design iterations, and communicate seamlessly using embedded collaboration tools.

The research also addresses a critical barrier to the adoption of metaverse technologies in the AEC industry: the lack of standardized definitions and frameworks, particularly concerning the role of Virtual Worlds (VWs) in design and construction workflows. Establishing shared terminology and industry standards is a necessary prerequisite for advancing research, development, and implementation of metaverse-based collaboration in AEC practices.

The methodology of this study involves developing a BIM–VW prototype using a combination of BIM, immersive and desktop VR, cloud computing, and network technologies within a 3D virtual environment created in a game engine. The platform prioritizes design collaboration as a core function, enabling scalable, mass multi-user interaction in a persistent, data-rich virtual space that maintains BIM interoperability, data integrity, and workflow continuity.

By proposing this BIM–VW framework, the research contributes to the development of an accessible, scalable, and continuous collaborative platform, laying the foundation for the practical integration of metaverse technologies in architectural design and construction processes.

2. Background and Related Research

2.1. Definition of Metaverse in the AEC Industry

The term “metaverse” was first introduced by Neal Stephenson in his 1992 science fiction novel Snow Crash, where it is depicted as an expansive VR space shared by millions of users [9]. Today, the metaverse is recognized as a broad digital ecosystem composed of multiple immersive VWs that coexist alongside the physical world. These virtual environments enable users to engage in a variety of activities, including social interaction, entertainment, education, and commerce, all facilitated by advanced digital technologies [10,11].

Since 2005, academic and industry research on the metaverse has evolved steadily, reflecting growing interest driven by technological advancements, platform innovation, and the emergence of a digital-native generation [12]. During the COVID-19 pandemic, the metaverse gained significant attention as people sought alternative modes of interaction and collaboration while maintaining social distancing measures [13].

In the context of the AEC industry, the metaverse is often discussed in relation to the convergence of physical, augmented, and virtual realities within a shared online environment. According to Hadavi and Alizadehsalehi (2024) have explored its application during the design and construction phases of building projects, utilizing models such as the ICAM DEFinition for Function Modeling (IDEF0) to map construction workflows and examine the technological motivations behind the migration of BIM into the metaverse. These studies include empirical reviews and case studies that highlight the potential for improving collaborative processes across project phases [14].

Further research by Chen et al. (2024) proposed frameworks for metaverse adoption in AEC, focusing on both the opportunities and barriers. For example, a cyber-physical-social systems (CPSS) approach has been employed to analyze the adoption challenges related to technological, political, and organizational factors, including accessibility limitations for non-programmers, scalability issues, networking and communication constraints, interoperability challenges, and unstructured data management [15].

Moreover, the integration of Construction 4.0 technologies such as BIM, VR, Internet of Things (IoT), and digital twins is not isolated from the metaverse but represents a synergistic technological ecosystem. However, due to varying maturity levels and separate developmental trajectories, the combined application of these technologies remains underexplored. The lack of standardized integration strategies leads to dispersed use cases, and selecting the appropriate combination of smart technologies often requires substantial time, effort, and expertise in practical applications

The concept of the metaverse represents more than a simple convergence of virtual and physical realities; it signifies a paradigm shift in the conception, execution, and management of construction projects [16]. By utilizing shared standards and protocols, alongside advancements in reality capture, digital twinning, and game engine technologies, the construction metaverse has the potential to drive unprecedented levels of collaboration, efficiency, and innovation within the AEC industry [17]. As the technology advancement cycle accelerates, these capabilities are becoming increasingly accessible and transformative.

The development of a metaverse tailored for AEC workflows requires a deliberate and systematic approach, with VWs serving as the foundational infrastructure. VWs create the essential ecosystem for metaverse applications in construction, enabling core activities such as model and data visualization, simulation, communication, and coordination among project stakeholders. Building expansive, persistent VWs provides the basis for developing collaborative, immersive, and data-driven environments that can be applied across the design, construction, and project management phases. These environments must remain adaptable, with evolving functionalities designed to meet the dynamic needs of users in the AEC industry.

2.2. BIM-Based Integrated Systems for Collaborative Design

Collaboration in building projects, with BIM as the driving technology, has fundamentally transformed the way architects, engineers, and contractors interact. BIM facilitates seamless information sharing and enhances visualization capabilities, streamlining the design and decision-making processes [18].

BIM-based collaborative design can be broadly approached from two perspectives: The technological aspect, which involves the integration of tools, systems, and software [1]. The organizational aspect, which focuses on the implementation of policies, workflows, strategies, and standards to support BIM collaboration [19].

In practice, participants in BIM-based design typically use different software tools tailored to their specific tasks. This diversity in tools often leads to data loss and redundancy during information exchange, requiring additional efforts to recover or replicate lost data [18]. According to Oh et al. [1], effective BIM-based collaborative design requires software that supports essential functions across all phases of a project. Based on the four design phases defined by the American Institute of Architects (AIA)—Pre-design, Schematic Design, Design Development, and Construction Documents—the critical functional requirements include:

(1)

BIM data generation and documentation,

(2)

Quality evaluation of BIM data, and

(3)

BIM data storage and management.

Further contributions by Choi et al. [20] include the development of a BIM collaboration framework designed to support compliance with BIM mandates. This framework draws from analyses of prior studies, including the ISO 19650-1:2018 [19] standard and the Common Data Environment (CDE) requirements identified through surveys of BIM practitioners. The CDE requirements for joint projects during the design phase are categorized into four key sectors:

(A)

BIM model management,

(B)

BIM data management,

(C)

Project participant management, and

(D)

Tool support.

This research builds upon these foundations by investigating the technological architecture of BIM-based collaborative design, with a focus on the development and integration of advanced software tools that improve communication among project stakeholders. The study highlights the need for a collaborative design platform that not only supports BIM models and data management but also enables real-time storage, interoperability, and dynamic visualization. Such a platform would allow multidisciplinary project teams to collaborate effectively throughout the design and construction process, enhancing both workflow continuity and project outcomes.

2.3. Evolution of BIM–VR to BIM–VWs for Collaborative Design

The concept of the metaverse is not confined to a single technology but emerges from the synergistic integration of multiple advanced systems, including VR, AR, Artificial Intelligence (AI), Blockchain, and the Internet of Things (IoT) [21]. These technologies collectively enable immersive, interactive, and interconnected digital environments. In the AEC industry, this technological convergence is actively reshaping project delivery processes across design, construction, and operation phases. Among these technologies, VR has demonstrated particular promise in enhancing design collaboration, stakeholder engagement, and decision-making.

The integration of BIM with VR has already transformed how AEC professionals engage with complex architectural and structural models. By providing immersive, real-time visualization, BIM–VR systems facilitate better spatial understanding, streamline design reviews, support client presentations, and optimize construction planning. Applications of BIM–VR include clash detection, construction sequencing, remote walkthroughs, and design evaluation, contributing to improved project outcomes. Recent research has expanded these capabilities from isolated individual experiences to multiuser collaborative environments. For example, Zaker and Coloma [22] explored VR-based workflows for on-site task simulations, Astaneh Asl and Dossick [23] examined multidisciplinary collaboration in VR settings, and Tea et al. [24] developed a real-time, multiuser VR platform for remote design reviews.

Despite these advances, current BIM–VR systems are constrained by fragmented workflows, task-specific implementations, and a lack of persistent collaboration. VR sessions often involve manual exports of BIM models, leading to data loss, versioning conflicts, and redundant modeling when updates occur. Furthermore, the complexity of standalone VR platforms (e.g., Revizto, Varjo, Fuzor, Enscape) presents accessibility challenges for non-expert users. While game engines like Unreal Engine (UE) offer enhanced customization, networking, and multiuser capabilities, their adoption in AEC remains limited due to technical barriers and a lack of standardized workflows.

These limitations underscore the need for an evolution from BIM–VR as a visualization tool to BIM-based Virtual Worlds (BIM–VWs) as comprehensive, persistent collaboration platforms. BIM–VWs are not just immersive spaces for visualization but multiuser, real-time environments that support live design collaboration, BIM data management, and integrated communication within a shared digital space. Unlike conventional VR workflows, BIM–VWs maintain continuous data streams, allowing for iterative design development, seamless model updates, and collaborative decision-making.

The shift toward Virtual Worlds (VWs) in the AEC industry reflects broader metaverse trends, particularly the distinction between open and closed systems. Open platforms like Decentraland emphasize decentralization, user-created content, and community-led governance, promoting inclusivity but posing challenges in terms of security and content moderation. Closed platforms such as Horizon Worlds offer safer, curated environments but may restrict user creativity and flexibility. For BIM-based Virtual Worlds (BIM–VWs), a hybrid approach is essential—one that supports broad stakeholder engagement while ensuring the interoperability, reliability, and data security required for professional AEC collaboration.

A leading example of this balanced approach is the EU-funded PrismArch project, which integrates immersive technologies with collaborative design. PrismArch introduces a layered VR platform that combines esthetic visualization, simulation data, and meta-information in a contextual interface, enabling real-time interaction among AEC professionals and interdisciplinary experts. This setup allows users to collaboratively evaluate project alternatives from their own disciplinary viewpoints, fostering co-creation and shared decision-making. The system supports dynamic, multi-author contributions, encouraging consensus-driven design outcomes through personalized perspectives and immediate visual feedback. PrismArch illustrates how a carefully structured VW can blend openness and control, enhancing collaborative workflows while maintaining clarity, precision, and security in complex design environments [25].

In this context, the BIM–VW framework proposed in this research represents a significant step forward. By leveraging game engines, cloud computing, and scalable networking, the framework ensures BIM interoperability, model persistence, multiuser interaction, and mass-user collaboration. This approach supports the transition toward a robust, user-centric, and technologically integrated metaverse for AEC, enabling real-time design collaboration, simulation, and project management at an unprecedented scale. As illustrated in Figure 1, the emergence of VWs continues to evolve, reflecting ongoing efforts to improve interoperability and user participation. Ultimately, the metaverse is positioned as the next transformative advancement in digital interaction, much like how social media revolutionized communication [26].

3. Methodology

3.1. Research Agenda

This research proposes a systematic framework for BIM-based collaborative design by integrating VR and VW technologies. The core objective is to resolve the current fragmentation in BIM–VR workflows, particularly concerning collaborative design scenarios in the AEC sector. While BIM and VR are increasingly adopted in design and construction processes, their integration often remains inconsistent, confined to specific project phases such as visualization or walkthroughs. This discontinuity limits the broader benefits of immersive technology, hindering real-time collaboration and reducing the potential of VR for comprehensive project engagement.

A review of existing literature and industry practices reveals five critical barriers to effective BIM–VR integration. These include the lack of technical expertise and user training, organizational resistance to adopting new collaborative technologies, the high costs of deploying VR systems, the complexity of VR application development, and the technological immaturity of VR platforms in addressing the needs of AEC workflows. These factors often restrict VR’s role to visualization and issue detection rather than facilitating fully collaborative design processes. As a result, VR tools remain underutilized in dynamic, multi-user project collaboration.

To overcome these limitations, this study proposes a BIM–VW framework designed to enhance user interaction, reduce technical barriers, and ensure project continuity through persistent data management. The framework introduces several key components. First, it prioritizes user-friendly interfaces with intuitive interactions and customizable training modules, minimizing the learning curve and accommodating users with varying levels of BIM and VR experience.

Second, the system enables multi-user collaboration, allowing stakeholders to participate in real-time or asynchronously within the same virtual environment, regardless of geographic location. Third, it ensures persistence and continuity by supporting saving, version control, and revision tracking, enabling consistent access to updated project information throughout the design and construction lifecycle.

The framework defines the collaborative metaverse by identifying nine essential characteristics of VWs applicable to design and construction collaboration. These include shared temporality, ensuring all participants operate within the same project timeline; real-time interaction, allowing simultaneous actions with immediate feedback; and shared spatiality, providing participants with a collective sense of digital presence in the same virtual space. Additionally, a single distributed database ensures that all users interact within a unified data environment [27]. The system integrates software agents to assist users, supports non-pausable environments that operate continuously, and maintains persistence so the virtual world evolves and stores information even when users are offline. Participants interact through avatars, representing their virtual presence, and can engage with both the digital environment and other users for collaborative design tasks.

This BIM–VW framework conceptualizes a multi-user collaborative system that merges immersive features from VWs with the data-centric attributes of BIM platforms. From the VW domain, the system incorporates real-time collaboration, shared temporality, persistence, avatar-based interaction, and intelligent software agents. From the BIM domain, it integrates parametric object libraries, property management tools, platform scalability, and interoperability with industry standards. The intersection of these domains creates a seamless collaborative design environment where users can interact with BIM data in immersive, real-time virtual spaces, enabling direct manipulation of architectural models and design parameters as seen in Figure 2.

The BIM domain, represented by the right circle, emphasizes system scalability, tool interfaces, libraries of BIM elements, platform user interface consistency, extensibility, interoperability, and effective support for managing properties. These capabilities are essential for the functional and operational success of BIM systems in construction and design.

The overlapping region signifies the multi-user environment, where attributes from both domains converge to enhance collaborative workflows. This integration ensures seamless communication and functionality, combining the real-time and spatial aspects of VWs with the structured and data-driven attributes of BIM systems.

Based on the combination of VW and BIM attributes, we propose a technical solution that leverages advanced algorithms, user-friendly interfaces, and cloud storage and databases within the context of CDEs to create a VW. This solution aims to improve the interactive visualization of BIM models and data, enabling efficient virtual communication in real-time.

3.2. BIM-Based VW Architecture

A significant limitation of traditional BIM–VR workflows is their reliance on isolated, task-specific sessions, which primarily serve for visualization, clash detection, or client walkthroughs rather than ongoing collaboration. Traditional VR systems often require manual export of BIM models into VR platforms, resulting in data loss, version control issues, and time-consuming rework when project updates occur. Additionally, these systems lack real-time model editing capabilities, meaning any changes made during VR sessions must be manually translated back into BIM software, leading to workflow fragmentation as seen in Figure 3.

In contrast, VWs offer a persistent, multi-user environment where BIM models are not static visualizations but dynamic, editable entities integrated into a continuously updated digital ecosystem. Unlike traditional VR, VWs support real-time collaboration, allowing multiple stakeholders to interact, modify, and annotate models simultaneously within the same shared space. This shift from isolated VR sessions to interactive, persistent VWs enables iterative design development and bridges the gap between immersive experience and practical project management, as seen in Figure 4.

This research proposes a methodology for creating a BIM–VW that facilitates continuous, collaborative design and management throughout the building lifecycle. The platform brings together diverse stakeholders such as architects, BIM managers, developers, project managers, contractors, facility managers, and end-users into a shared virtual environment where they collaborate using avatars for real-time communication and design iteration.

The system supports interoperable BIM data exchange through standards like IFC from BIM authoring tools, enabling seamless integration from various BIM and CAD tools. Real-time synchronization and version control allow users to track and manage design changes collaboratively, reducing conflicts and ensuring data consistency. A multi-layered content structure organizes geometry, metadata, simulations, documentation, and decision logs, supporting a comprehensive project workflow from conceptual design to facility operation.

The platform is hardware-agnostic, allowing access through VR, AR, MR, and traditional desktop devices, making participation inclusive and scalable. Clients and end-users can directly interact with BIM models, make design changes, add annotations, and update data in real time, promoting a user-driven design process.

Furthermore, the system supports APIs and plugin development, encouraging developers to contribute additional tools and functionalities. This open, extensible ecosystem ensures that the platform evolves with project needs, fostering innovation and continuous collaboration within a unified virtual workspace.

The proposed BIM–VW prototype developed in this research leverages cloud computing, advanced algorithms, and is based on CDE data structures to support collaborative design tasks. The BIM–VW Multi-User Manager controls network operations, manages avatar coordination, and facilitates cloud-based communication to ensure smooth synchronization among participants. Simultaneously, the BIM–VW Model and Data Manager organizes 3D models, CAD data, and related project datasets, ensuring that all visualizations and technical details remain readily available and continuously updated. This management layer eliminates inefficiencies commonly associated with the manual data handling typical of conventional VR workflows.

To further enhance user interaction, the system includes an Interface Functions Module that provides tools for model manipulation, material editing, annotation, and collaborative adjustments. These features enable users to contribute to the design process without requiring advanced programming knowledge, promoting a more inclusive and interactive design environment. The collaborative workflow is supported by the centralized BIM CDE, which stores both Project Information Models (PIM) and Asset Information Models (AIM). By managing graphical and non-graphical data in a cloud-based system, the CDE ensures that architects, engineers, contractors, and clients can access, modify, and coordinate project information in real time. This structure enhances transparency, reduces communication gaps, and facilitates continuous project development.

The collaborative process begins with uploading BIM models into the VW platform, where multiple users can simultaneously interact with the design and dataset. Participants can collaboratively edit models, annotate design elements, and store modifications directly in the cloud. These actions synchronize automatically with the CDE, maintaining data integrity and ensuring that all stakeholders work from the latest version of the project. Visual communication is facilitated through avatars, providing an immersive and engaging medium for real-time interaction in the virtual space as seen in Figure 5. This workflow bridges the gap between digital collaboration and physical project management, aligning immersive environments with construction industry requirements.

4. Development of BIM-Based VW for Interactive Design Collaboration

4.1. Multi-User Manager

This research proposes a BIM–VW framework for collaborative architectural design, leveraging infrastructure derived from Massively Multiplayer Online Role-Playing Games (MMORPGs) developed using advanced game engines. MMORPG architectures are proven in managing large-scale, multi-user environments, offering functionalities that exceed the limitations of current AEC–VR tools, particularly in terms of scalability, persistence, and synchronous collaboration.

Two leading game engines, UE and Unity, dominate VR development in the AEC sector. UE offers advanced rendering technologies, including Lumen global illumination, Nanite virtualized geometry, and real-time ray tracing, which produce highly realistic and immersive environments. While Unity also achieves strong visual quality through its High-Definition Render Pipeline (HDRP), UE is widely regarded as superior in delivering photorealistic performance for complex design workflows. Twinmotion, built on UE’s core, provides intuitive real-time rendering and a comprehensive asset library but is limited by its fixed feature set and lack of extensibility or support for custom scripting. It is designed for single-user visualization and does not meet the multi-user, customizable demands of collaborative AEC environments.

While Twinmotion Cloud allows model sharing, it lacks interactive multi-user capabilities. In contrast, UE supports persistent real-time collaboration through native networking tools like state replication, Multi-User Editing, and Pixel Streaming for remote access. Unity offers similar features via Netcode for GameObjects and Unity Reflect, which enables BIM integration and real-time syncing with Revit. However, UE excels in extensibility, offering full access to C++, Blueprint visual scripting, and a large plugin ecosystem, enabling rapid customization. This research utilizes Unreal Engine 5.2 (UE 5.2) for its advanced graphics, scalable networking, and seamless BIM interoperability, forming the foundation for a persistent, cloud-based virtual environment with real-time user interaction.

Moreover, UE’s Blueprint system significantly reduces the technical barrier for AEC professionals by enabling visual scripting without the need for traditional programming. This empowers designers, planners, and stakeholders to engage in interactive VR workflows, facilitating the development of an immersive, persistent, and user-friendly BIM–VW environment. As such, the proposed framework addresses key limitations in existing BIM–VR integration tools and demonstrates the potential of game engine-based platforms for future AEC collaboration.

4.1.1. Network

This research employs the Epic Online Services (EOS) plugin within UE 5.2 to establish a robust online infrastructure for the BIM–VW platform. EOS enhances UE’s native Online Subsystem (OSS) by providing comprehensive networking and user management capabilities essential for real-time, multi-user collaboration in the AEC context.

Key EOS functionalities integrated into this research include cross-platform authentication, friends management, session handling, party creation, real-time presence tracking, and peer-to-peer networking. Additionally, EOS supports cloud-based features such as Player Data Storage and Title Storage, enabling persistent data management and secure collaboration across geographically distributed teams.

By allowing participants to invite collaborators through familiar platforms like Steam, EOS simplifies remote connection setup and lowers technical barriers for users. This ensures seamless access to the VW environment, facilitating an intuitive and accessible multi-user experience essential for collaborative design, decision-making, and project management in immersive BIM workflows.

4.1.2. Avatars

The avatars developed in this research serve to identify real-world users, with each avatar assigned a unique ID. Each avatar is equipped with a menu featuring various functions designed for communication and interaction within the VW as seen in Figure 6. These include video/voice chat, interactive visualization of BIM content such as BIM objects and data, a material changer function for visualizing different materials, as well as annotation and measurement tools. Figure 6 shows the virtual avatars within the VW.

4.1.3. Cloud Storage

Cloud storage serves as a robust solution for the persistent storage of user data and interactions within the VW. By leveraging cloud technology, users can securely save their information, ensuring that their experiences and interactions are not only accessible but also seamlessly integrated across various devices and platforms. This storage solution enables real-time data synchronization, allowing users to engage with the virtual environment consistently, regardless of their location or the device they are using. Moreover, the cloud infrastructure offers scalability, accommodating increasing amounts of data as user interactions grow over time. For the user aspect of the VW, we utilized components of the EOS plugin for UE 5.2 to save unique user IDs and track their progression within the VW.

4.1.4. Communication System

Video Conferencing (VC) is an effective solution to overcome challenges associated with online face-to-face interactions. VC represents a technological advancement that enables users to participate in face-to-face meetings from different locations simultaneously, enhancing communication among users in the VW.

In this research, we utilized the Agora Unreal SDK to develop a VC system that supports real-time communication. The Agora Unreal Video SDK is a real-time communication (RTC) software development kit (SDK) that allows developers to integrate real-time video chat features into UE applications. This SDK provides a wide array of APIs and tools that simplify the integration of RTC capabilities into UE games and other software applications. It is designed to function seamlessly across various platforms, such as Windows, Mac, iOS, and Android. The SDK ensures exceptional audio and video quality, alongside reliable service for users globally.

4.2. Model and Data Management

UE with Datasmith and Twinmotion via Direct Link allow seamless one-click import of models from major AEC tools such as Revit, Rhino, ArchiCAD, SketchUp, and IFC formats. Unity offers similar functionality through the PiXYZ plugin and Unity Reflect—PiXYZ supports broad CAD/BIM import with metadata, while Reflect enables real-time sync with Revit and SketchUp for AR/VR use. Both platforms support cloud-based collaboration: Unity provides Cloud Build and Render Streaming, while UE supports Pixel Streaming with integration options for AWS and Microsoft Azure for remote real-time access and visualization.

There has been a growing interest in using BIM 3D models and data through UE to improve visualization and interactivity in construction projects. These innovations have significantly advanced in recent years, enabling real-time model and data visualization. Some notable examples include:

• Speckle Connect: This tool integrates with AEC tools like CAD, BIM, and GIS through desktop connectors, acting as plugins for software such as Rhino, Revit, and AutoCAD. These connectors facilitate data flow to and from the Speckle server, converting CAD and BIM data into Speckle’s neutral format. The Speckle Manager manages accounts and connectors, requiring users to add a Speckle account for data transmission. Figure 7, shows a model uploaded from Revit to Speckle in UE 5.2 in real-time. However, additional programming within UE 5.2 is required to visualize the BIM data effectively.
• Cavrnus Metaverse Connector: This tool enhances 3D workflows by adding collaborative features to UE and Unity projects. Its Journal tracks interactions over time, providing a detailed timeline that can be accessed via APIs. The platform supports high-performance voice, video, and screen sharing through WebRTC for real-time data synchronization. Cavrnus can be deployed in the cloud or on private servers, and is compatible with multiple platforms. Figure 8 illustrates the Cavrnus interface.
• Datasmith: This research employs Datasmith and Datasmith Runtime to streamline the real-time import and visualization of complex 3D assets and BIM models from platforms such as Revit, Rhino, AutoCAD 2024, and SketchUp into UE. Datasmith preserves key model elements, including geometry, materials, metadata, and scene hierarchy, ensuring seamless translation of architectural and engineering data. Leveraging Datasmith Runtime, the system enables dynamic manipulation of udatasmith files within UE using Blueprints, allowing real-time updates without disrupting collaborative sessions. The platform was further customized to support multi-user interaction with BIM models and embedded metadata in a shared VW. It also facilitates the real-time integration of non-graphical documents (e.g., DWG, PDF, CSV, JPEG), enhancing interoperability and streamlining design coordination and review processes as seen in Figure 9.

In the VW created through this research, we employed Datasmith during runtime due to the plugin’s ability to preserve 3D models and BIM objects within a single udatasmith file format. Additionally, by utilizing Amazon Simple Storage Service (Amazon S3), which offers top-tier scalability, data availability, security, and performance, users can conveniently store and access udatasmith files, promoting smooth collaboration and access among different teams and projects in our proposed VW.

4.3. Interface Functions

This research advances user interaction with BIM content in a Virtual World (VW) by enhancing UE 5.2 with collaborative features tailored for multi-user design engagement. Initial findings showed that passive visualization was insufficient for effective collaboration, prompting the development of interactive tools that transform the VW into an active design workspace. A key addition is the BIM Object Viewer, which uses Datasmith’s preserved object hierarchies to allow users to select individual elements, view associated metadata, and toggle visibility—enabling focused design review. Another essential feature is the Annotation and Measurement Tool, which lets users annotate models, add comments, and measure distances in real time, as shown in Figure 10.

These tools support real-time problem-solving, decision-making, and design iteration among stakeholders. By enabling direct interaction with both visual and data-rich BIM components, the platform shifts collaboration from static viewing to dynamic participation. This not only improves communication but also accelerates project workflows. Looking ahead, the research aims to evolve the VW platform into an open, collaborative ecosystem by integrating plugins and APIs. This strategy invites contributions from third-party developers and industry professionals, promoting ongoing innovation and ensuring the platform adapts to emerging needs within AEC and related fields.

5. Evaluation and Discussion

5.1. BIM VW Evaluation

The VW platform’s novelty lies in its persistent system, which facilitates real-time visualization of BIM 3D models with embedded data in the udatasmith format. These models are uploaded via Amazon S3 Cloud, ensuring security through passkey access for model sharing. Over a one-month evaluation period, conducted as part of a graduate school program at KNU, the platform was tested by 15 users (9 of whom were recurrent), who collectively engaged in 106 h of interaction, averaging 20 min per user session. Participants uploaded 22 MB of data, created 27 collaborative models from SketchUp, Revit, and Rhino, and exchanged approximately 320 text messages in the shared virtual space as seen in Figure 11. These insights demonstrate the platform’s potential for collaborative digital design and remote teamwork. However, further analysis is needed to refine usability and performance.

The 9 recurrent users were surveyed to provide feedback on their experience with BIM tools, their current VR tools of choice, and their experience levels. The survey also asked users to highlight the most important functions in the VW and additional features that would be beneficial in future versions of the platform.

Table 1. Summary of User Preferences, Experience Levels, and Feedback on the BIM–VW Platform During Collaborative Design Sessions.

User	Preferred BIM/CAD Tool(s)	VR Experience	Preferred VR Tool(s)	Top VW Feature	Suggested Platform Improvements
1	Sketchup	Novice	Twinmotion	Model import	Improve Model navigation
2	Sketchup, Rhino	Novice	Twinmotion, Enscape	BIM data visualization	Improve Model navigation
3	Sketchup, Revit	Intermediate	Twinmotion, Unity 3D	Multi user interaction	Include personal user info; better navigation
4	Revit	Intermediate	Unity 3D	BIM data visualization	Improve Model navigation
5	Revit	Intermediate	Unreal engine	BIM data visualization	Improve Model navigation
6	Sketchup, Revit	Intermediate	Unreal engine	Multi user interaction	Better Model data visualization
7	Sketchup, Revit, Rhino	Expert	Unity 3D	BIM data visualization	Enable Exporting data from VW
8	Sketchup, Revit	Expert	Unreal engine, Unity 3D	BIM data visualization	Real-time Revision control of data
9	Revit	Novice	Lumion	BIM data visualization	Improve Model navigation

summarizes the insights into user preferences, experience levels, and feedback regarding the VW.

The majority of users preferred SketchUp, Revit, and Rhino as their BIM tools, reflecting the diverse range of design software supported by the platform. Users’ VR experience ranged from novice to expert, with most participants falling into the intermediate category, indicating a moderate familiarity with immersive environments. Preferred VR tools varied among Twinmotion, Unity 3D, UE, and Lumion, highlighting the flexibility of the VW platform in integrating different VR applications.

In terms of functionality, users emphasized BIM data visualization and multi-user interaction as the most important features, underscoring the platform’s effectiveness in collaborative design tasks. However, several users suggested updates for the platform, such as improvements in model navigation, user information customization, real-time data revision control, and data exporting capabilities. These suggestions point to areas for further enhancement of the platform. This feedback provides valuable insights into user needs, guiding future developments aimed at improving the usability and performance of the VW platform.

Additionally, Figure 12a–c illustrates the visualization outcomes of BIM data from different users within the VW. The table details the file formats, levels of detail, and corresponding visual representations of the models after conversion to the .udatasmith format, which facilitates their integration into the VR environment.

The models were derived from various BIM software, including IFC, Revit, and SketchUp, reflecting the diversity of data sources. The level of detail varies depending on the original file format and software. The Revit model (User 5) exhibited the highest level of detail, showcasing comprehensive BIM data such as material properties, dimensions, and object classification. In contrast, the SketchUp model (User 7) displayed the lowest level of detail, with minimal attribute information. The IFC model (User 4) demonstrated a moderate level of detail, primarily representing geometric elements with limited metadata.

This comparison highlights the impact of different BIM authoring tools and file formats on the quality and richness of BIM data in the VR environment. It suggests that while the VW platform’s ability to support diverse formats is advantageous, further optimization may be needed to enhance data consistency and visualization quality across different file sources.

The 9 recurrent users were surveyed to provide feedback on their experience with BIM tools, their current VR tools of choice, and their experience levels. The survey also asked users to highlight the most important functions in the VW and additional features that would be beneficial in future versions of the platform. Table 1 summarizes the insights into user preferences, experience levels, and feedback regarding the VW.

The majority of users preferred SketchUp, Revit, and Rhino as their BIM tools, reflecting the diverse range of design software supported by the platform. Users’ VR experience ranged from novice to expert, with most participants falling into the intermediate category, indicating a moderate familiarity with immersive environments. Preferred VR tools varied among Twinmotion, Unity 3D, UE, and Lumion, highlighting the flexibility of the VW platform in integrating different VR applications.

In terms of functionality, users emphasized BIM data visualization and multi-user interaction as the most important features, underscoring the platform’s effectiveness in collaborative design tasks. However, several users suggested updates for the platform, such as improvements in model navigation, user information customization, real-time data revision control, and data exporting capabilities. These suggestions point to areas for further enhancement of the platform. This feedback provides valuable insights into user needs, guiding future developments aimed at improving the usability and performance of the VW platform.

5.2. Application Scenario: Case Study

In this scenario, a VW platform is developed for collaborative urban design, focusing on Kyungpook National University (KNU) as a real-world case study. The project integrates Cesium for UE, a powerful plugin that streams high-resolution Geographic Information System (GIS) and geospatial data into UE, allowing for the generation of an accurate and immersive 3D map of the KNU campus and its surrounding urban context, as seen in Figure 13.

The process begins by importing Cesium’s photorealistic terrain and building data into UE, creating a large-scale, real-time 3D model of the entire university area. This GIS-based visualization provides a realistic foundation for contextual urban design discussions, ensuring all participants have spatial awareness of the broader campus and its environmental constraints.

Once the GIS map is generated, the building of interest—such as a proposed new research center or student complex—is replaced with a detailed BIM model imported using Datasmith. This BIM model contains not only geometric information but also semantic data, including material properties, energy simulations, and cost estimates.

Through the Virtual World platform, multiple stakeholders—urban designers, university planners, architects, students, and faculty—enter the shared 3D environment as avatars. They collaboratively explore the site, assess spatial relationships, analyze environmental impacts, and simulate pedestrian movement. Users can annotate, adjust design parameters, and propose modifications in real time.

This workflow bridges urban-scale GIS visualization with detailed BIM design, enabling a comprehensive, immersive urban design process that supports context-aware, data-driven decision-making within a persistent, multi-user virtual ecosystem.

In this scenario, the system operated on a high-performance workstation (Intel Core i9-13900, 32 GB RAM, RTX 3080, Windows 11 Pro). In moderately complex BIM environments, the platform consistently achieves up to 90 frames per second (FPS), ensuring smooth and responsive visual interaction. During multi-user sessions on a local network, synchronization between participants—such as avatar movements and model modifications—is typically maintained within 50–100 ms, supporting a cohesive and real-time collaborative experience.

These performance metrics demonstrate that the virtual platform remains stable, fluid, and effective even during intensive design activities. However, to ensure broad usability and scalability, the system will require extensive performance testing across diverse hardware configurations and network conditions.

5.3. Discussion

The VW methodology presented in this research offers a transformative collaborative framework for the AEC industry. While recent advances in BIM–VR integration have improved design visualization and stakeholder engagement, additional development is needed to fully support hybrid environments that integrate both 2D CAD and 3D BIM workflows, particularly in early-stage design. The establishment of standardized collaborative workflows is essential to ensure consistency, reduce errors, and enable seamless coordination across design, construction, and operation phases.

Beyond traditional design tasks, the proposed VW framework has wide-ranging applications across multiple architectural and construction disciplines. By embedding BIM data into an immersive, real-time, multi-user environment, the system fosters cross-disciplinary collaboration and informed decision-making. In urban design, large-scale virtual simulations allow planners, policymakers, and communities to collectively explore development proposals. Users can navigate streetscapes, assess visual and environmental impacts, and adjust design parameters in real time, supporting participatory urban planning.

In interior architecture, the platform enables designers, engineers, and clients to collaboratively modify spatial layouts, furnishings, and lighting in an immersive setting. All changes are linked directly to the BIM model, ensuring data integrity while enhancing client engagement and design iteration.

For Facility Management Building Information Modeling (FM-BIM), the VW serves as a persistent, interactive platform for post-construction operations. Facility managers and maintenance teams can access asset data, simulate service procedures, annotate elements with operational histories, and plan future interventions collaboratively. This integration ensures continuity from design and construction through to lifecycle management, promoting efficiency and proactive maintenance strategies.

Importantly, the VW platform is designed to accommodate emerging AEC technologies such as AI. AI-driven tools can be integrated to automate clash detection, suggest optimal design solutions, or analyze energy performance within the virtual environment. Generative design algorithms can propose real-time design variations based on constraints set by project teams, while AI-based simulations can forecast construction schedules, cost implications, or environmental impacts. Additionally, AI-enhanced natural language interfaces can simplify interactions for non-technical users, allowing voice-driven commands and intuitive querying of building data during collaborative sessions.

While this research presents a comprehensive VW framework for collaborative BIM-based design and facility management, several limitations remain. Cybersecurity and data privacy are critical concerns in multi-user, cloud-connected environments, especially when handling sensitive project information and proprietary BIM data.

To ensure the security and privacy of the BIM–Virtual World platform, it is essential to protect all shared data with secure connections and restrict user access by assigning specific roles based on their responsibilities. Trusted login systems, such as Google or Microsoft accounts, should be used to verify user identities before granting access. Only authorized users should be allowed to enter project spaces and interact with design elements according to their permissions.

Login protections like two-step verification should be implemented for users with higher privileges, while user actions must be tracked to monitor access and changes. Important files should be backed up and safeguarded against accidental deletion or unauthorized modifications. Together, these measures will maintain the confidentiality of design data and support secure, collaborative work within the virtual environment.

Additionally, the current system assumes a stable high-bandwidth network infrastructure, which may limit accessibility in regions with poor connectivity. Hardware requirements for immersive VR participation also present a barrier for some users, despite efforts to support desktop-based interactions. However, addressing security protocols, encryption, and data protection strategies is beyond the scope of this study, as these topics are covered extensively in existing cybersecurity research

Another limitation is the need for user training and change management to encourage widespread adoption across various stakeholder groups. While the platform is designed to reduce technical complexity, a learning curve remains for users unfamiliar with immersive technologies that requires refinement by an expert or User Interface and User Experience (UI/UX) experts.

6. Conclusions

This study has demonstrated a novel BIM- VW platform that substantially advances the integration of BIM and VR for collaborative architectural design. The system is built on UE 5.2, chosen for its advanced graphics and built-in support for real-time, multi-user networking. We employed EOS within UE 5.2 to establish a robust cloud infrastructure, providing cross-platform authentication, session management, and peer-to-peer networking required for scalable, multi-user collaboration.

A custom Multi-User Manager module orchestrates network operations and avatar coordination, while a central cloud-based CDE stores all project data (both graphical models and documents) to maintain version control and data integrity. By synchronizing changes automatically through the CDE, the system ensures that all participants access the latest BIM models and information in real time.

Users are represented by avatars, each with a unique identifier and in the BIM–VW. These avatars support embedded video/voice chat and interactive BIM visualization—for example, users can manipulate materials or toggle object visibility on-the-fly—and include an annotation/measurement toolkit for markups and distance calculations directly on the model. A BIM Object Viewer exploits Datasmith’s maintained object hierarchies so that individual model elements can be selected and their metadata inspected within the virtual scene. These features enable stakeholders to collaboratively review and adjust design details (verifying dimensions, noting issues, proposing changes) without leaving the VR context, thereby accelerating decision-making and reducing information loss in iterative workflows.

A key technical innovation is the seamless import and handling of BIM data. The system uses Epic’s Datasmith toolset (with Datasmith Runtime) to bring complex 3D models from Revit, Rhino, SketchUp, and other CAD platforms into the VW, preserving geometry, materials, metadata, and scene hierarchy. This live import pipeline allows models (packaged as “udatasmith” files) to be loaded and manipulated in real time within the engine. Amazon Web Services (Amazon S3) is used to store these udatasmith files and other assets in the cloud, providing scalable, secure storage and enabling consistent real-time data synchronization among distributed teams. The workflow even accommodates non-graphical documents (e.g., DWG drawings, PDFs, spreadsheets, images), which can be imported and shared within the VW environment on the fly. Together, these capabilities ensure broad interoperability—the platform natively supports multiple BIM formats (Revit, SketchUp, Rhino, etc.) and file types—and maintains a single source of truth as the project evolves.

Communication within the VW is fully integrated. We incorporated the Agora real-time communication SDK to embed high-quality voice and video conferencing directly into the platform. Users can engage in live face-to-face discussions while navigating the shared 3D model. Text chat and presence tracking are also handled through the EOS framework, enabling a rich, continuous dialog. In short, the platform creates a persistent, MMORPG-style collaborative space (leveraging game-engine networking paradigms) where all project participants—architects, engineers, contractors, even clients—can co-locate as avatars and work on the BIM simultaneously.

A one-month user trial demonstrated the system’s effectiveness for collaborative design. Fifteen participants (ranging from novice to expert VR users) collectively logged over 106 h in the VW, uploading 27 distinct BIM models (from SketchUp, Revit, Rhino) and exchanging roughly 320 messages. In follow-up surveys, users overwhelmingly highlighted BIM data visualization and multi-user interaction as the platform’s most valuable features. They also provided constructive feedback—for example, suggesting enhancements in model navigation, personalized user information, real-time revision control, and data export functions—which will guide future refinements.

These results affirm that the BIM–VW framework can reduce redundant manual effort, minimize information loss between design iterations, and foster greater engagement among distributed teams. In conclusion, this research establishes a technically rigorous foundation for an AEC-focused metaverse. By unifying BIM and VR in a single persistent environment—complete with cloud-based CDE, real-time synchronization, multi-format model import, and advanced communication tools—the proposed VW platform meets current collaborative design demands and paves the way for future innovation.

Looking ahead, the platform is designed to be extensible: planned extensions include integration of 4D and 5D BIM data (scheduling and cost information), as well as emerging technologies such as artificial intelligence, augmented reality overlays, and 5G networking. These enhancements will further augment the virtual world’s role as a central hub for building lifecycle activities. Overall, the enhanced BIM–VW prototype developed in this study offers a compelling proof-of-concept that metaverse technologies can transform AEC collaboration, laying the groundwork for more efficient, immersive, and data-driven design and construction processes.