Link to article

ABSTRACT

This chapter presents a theoretical framework for implementing Digital Twins (DTs) in several stages of the building lifecycle, including design, construction, operation, and maintenance. The analysis begins by exploring the function of Building Information Modeling (BIM) in the process of design and construction. It emphasizes the significance of incorporating data obtained from physical buildings through implementing the Internet of Things (IoT) across the building lifecycle. The chapter discusses how the level of detail in BIM models contributes to the creation of DTs by utilizing both 3D data and asset specifications from BIM models, along with real-time data gathered from IoT devices. A comparative matrix of different platforms used for developing DTs also evaluates visualization, user interface, and sensor connectivity. The matrix is organized based on the roles of various user groups aligned with distinct activities in the building lifecycle: the development group (architects, engineers, construction managers) responsible for uploading and updating data, the control group (facility managers) tasked with data analysis and control, and the innovators group (researchers and academia) involved in data testing and validation. Furthermore, the chapter explores future advancements in DTs, focusing on integrating artificial intelligence algorithms. In summary, valuable insights are provided into the effective utilization of DTs across various stages of the building lifecycle: design, engineering, construction, operation, and maintenance.

KEYWORDS: Digital Twins; Building Life Cycle; BIM; DT implementation

1. INTRODUCTION

1.1 An overview of Digital Twins (DT) in building lifecycle

The building lifecycle—design, construction, operation, and maintenance—requires seamless data exchange and interoperability. BIM, a digital model that stores both geometric and semantic data, has long facilitated collaboration among AECO professionals. However, BIM’s static nature limits its ability to update in real time. The integration of IoT, which collects real-time data through sensors, extends BIM's capabilities by enabling live monitoring of construction processes and indoor conditions. This combination forms the basis for emerging Digital Twin (DT) technologies, originally developed in aerospace and manufacturing, but still in their early stages within AECO. Consequently, a systematic review of the BIM-to-DT transition is needed to guide future research and technology adoption in the industry.

1.2 IoT data from building life cycle

BIM is widely used in construction to create a unified virtual model that enhances collaboration during design, engineering, and construction by integrating diverse data and facilitating communication among stakeholders. However, its focus on static data limits its effectiveness for operation and maintenance. In contrast, Digital Twins overcome these limitations by incorporating IoT and sensor-actuator systems to capture real-time data, perform simulations, and enable dynamic facility management with AI and machine learning. A Digital Twin is a virtual replica of a physical entity that continuously mirrors its actions and functionalities through bidirectional connectivity.

Integrating sensors and simulation data into Digital Twins transforms building design, engineering, construction, operation, and maintenance. By equipping DTs with IoT devices, sensors, and data loggers, real-time indoor environmental data can be collected. This data can be used with tools like EnergyPlus for real-time energy simulations and with AI/ML algorithms to predict failures, enhancing the overall efficiency and reliability of the building lifecycle.

2. BIM IN DESIGN AND CONSTRUCTION PRACTICE

2.1 BIM Levels of Detail

Digital Twins (DTs) offer virtual representations of physical systems that can transform construction by making design, building, and maintenance more efficient, cost-effective, and sustainable. However, DTs are still in early development in the architecture, engineering, and construction (AEC) industry, and standardized practices are needed for their widespread use. Currently, Building Information Modeling (BIM) is the conventional tool for digital representation, providing detailed data on a building's design, construction, and operation, with its detail varying by Level of Detail (LOD).

• LOD 100. The Model Element may be graphically represented in the Model with a symbol or other generic representation. However, it does not satisfy the requirements for LOD 200. Information related to the Model Element (e.g., cost per square foot, tonnage of AC., etc.) can be derived from other Model Elements. (Menassa, 2021)

• LOD 200. The Model Element is generically and graphically represented within the Model with approximate quantity, size, shape, location, and orientation.

• LOD 300. The Model Element, as designed, is graphically represented within the Model such that its quantity, size, shape, location, and orientation can be measured. (Pan & Zhang, 2021)

• LOD 350. The Model Element, as designed, is graphically represented within the model so that its quantity, size, shape, location, orientation, and interfaces with adjacent or dependent Model Elements can be measured.

• LOD 400. The Model Element is graphically represented within the Model with detail sufficient for fabrication, assembly, and installation.

• LOD 500. The Model Element is a graphic representation of an existing or as-is condition developed through a combination of observation, field verification, or interpolation. The level of accuracy shall be noted or attached to the Model Element.

Key data formats are crucial for developing Digital Twins in construction. IFC provides a structured model for building components, while BCF links real-world observations to DT elements for effective communication. Visual aspects are enhanced by 3D model formats like Revit and SketchUp, sensor data formats deliver real-time updates, and the emerging USD format boosts visual fidelity and interoperability across different software

3. DATA FORMATS FOR CONSTRUCTION’S DIGITAL TWIN

Although no universal standard exists yet, several key data formats are essential for developing Digital Twins in construction. The Industry Foundation Classes (IFC) provide a structured model for building components, while the BIM Collaborating Format (BCF) enables communication by linking real-world observations to DT elements. Visual representations are enhanced by 3D model formats like Revit and SketchUp, and sensor data formats (CSV, JSON, or proprietary) supply real-time information. Additionally, the emerging Universal Scene Description (USD) format improves visual fidelity and interoperability for complex 3D scenes across various software applications

Here's how USD benefits Digital Twins:

• Enhanced Visual Quality: USD allows for the import of high-fidelity 3D models from various sources, resulting in a more visually stimulating and immersive digital twin experience.

• Improved Collaboration: The interoperable nature of USD makes it easier for different teams using different software to collaborate on the visual aspects of the digital twin.

• Efficient Data Management: USD's efficient data structure allows for managing complex scenes without sacrificing performance, which is crucial for real-time rendering within digital twins.

While IFC, BCF, and 3D model formats remain the cornerstones of data exchange for DTs, USD offers a compelling option for projects that demand exceptional visual quality and seamless collaboration across diverse software environments. Understanding these data formats enables the user to navigate the digital twin development process and choose the right tools to construct a robust and informative digital representation of a building project.

4 USER IDENTIFICATION

While the overall concept of DTs follows similar concepts for most applications, the practical implementation varies depending on factors such as the level of detail and use cases for DTs. Additionally, setting up a DT on each platform requires a distinct procedure encompassing different levels of complexity. This implies that a careful decision and thorough study of user identification and case studies will be essential before implementation. Consequently, a comprehensive comparison matrix can serve as a roadmap for selecting the most appropriate platform based on DTs’ requirements and potential use cases. Given the focus on the construction field encompassing both academia and industry sectors, the users of DTs within this domain can be broadly classified into three main groups as shown in Table 1:

Table 1. User description

Development Group:

The development group consists of architects, engineers and construction managers involved in the design and construction phases of projects. Their role involves defining and providing essential building-related data while establishing mechanisms for the continuous updating of data within the DT

Control Group:

Following the completion of construction, the control group, comprising facility managers, scrutinizes and maintains the data provided by the development group, and from any embedded IOT sensors. Their focus lies in ensuring optimal operational performance and efficient maintenance scheduling for the facility.

Innovators:

Researchers and innovators within the realm of DTs explore new frontiers, experiment with emerging technologies, and contribute to the ever-evolving landscape of this technology through dedicated research and development initiatives.

User identification is key to defining project constraints and scope, as a user-defined perspective highlights critical DT aspects tailored to specific groups. For example, effectively integrating BIM data with sensors and actuators into DTs requires meeting specific criteria. The growing use of AI across various domains adds a new layer of criteria that innovators and researchers must consider when deploying DTs. Each component—user-defined needs, BIM integration, and AI incorporation—has its own criteria essential for selecting the right DT platform. The subsequent section outlines these detailed criteria in a comparison matrix based on user needs.

5 EVALUATION CRITERIA TO ANALYZE DT PLATFORMS

5.1 Visualization: Development group (architects, engineers)

As the digital representation of physical assets and processes becomes increasingly integral to operations, the ability to render and interact with complex 3D models, dynamic data, and diverse levels of detail emerges as a cornerstone requirement. Therefore, it is critical to identify key criteria and the corresponding methods or tools for evaluating them. In this subsection, a selection of criteria to consider will be discussed (see Table 2).

Table 2: Evaluation Criteria – Development Group

Criteria and Description

• Rendering and Performance: When evaluating DT platforms for the construction sector, rendering capabilities and performance are critical factors. The ability of platforms to efficiently handle intricate 3D models, dynamic data, and various levels of detail directly impacts user experience. Metrics such as real-time adaptability and processing speed for complex computations provide insights into platform effectiveness. Furthermore, responsiveness to user interactions, and efficient utilization of system resources playvital roles in preventing lags or delays, ensuring seamless operation throughout the project lifecycle (Cabral et al., 2015).

• Realtime Collaboration: Collaboration features are indispensable for design, construction and maintenance teams across different project phases. Platforms facilitating real-time interaction and stakeholder coordination enhance communication and streamline decision-making processes. Compatibility and Scalability: Platforms should support various file formats and seamlessly integrate with other design and 3D modeling tools commonly used in the construction industry. Moreover, accessibility to Computer-Aided Design (CAD) and BIM assets and libraries of materials and textures enhance the platform's utility. Scalability ensures that the platform can accommodate evolving project requirements and scale up to effectively meet the demands of large-scale construction projects (Sobotkiewicz & Milkowski, 2021).

• Performance Optimization: Platforms that offer performance optimization tools cater to the need for efficient resource utilization and enhanced rendering efficiency. Features such as LOD adjustments and occlusion culling improve overall platform performance, ensuring smooth operation even with complex data sets. DT platforms can deliver a superior user experience by implementing these optimization techniques, enabling stakeholders to focus on project objectives without technical constraints (Bethel et al. 2012).

• Customized Views: Offering perspectives relevant to stakeholders' needs enhances communication and decision-making within the stakeholders in the building life cycle. By allowing stakeholders to focus on the aspects pertinent to their domain expertise, customized views streamline collaboration and contribute to project success.

5.2 User Interface (UI) and User Experience (UX):

The selection of a Digital Twin platform goes beyond technical specifications to emphasize user experience and visualization. A comprehensive evaluation framework has been developed, assessing critical factors such as intuitiveness, collaboration features, visual programming support, interactivity, navigation, and community support. This framework is designed to empower three key user groups—architects and engineers (Development Group), facility managers (Control Group), and researchers/academia (Innovators Group)—to make informed decisions for efficient and effective DT creation and interaction.

Table 3: Evaluation Criteria – User Groups

Criteria & Description

• Intuitiveness: Measure user task completion time and error rates for joint actions to gauge the learning curve and ease of use. Evaluate user satisfaction through subjective feedback and assess how well interface design leverages familiar patterns, offers clear labeling and help functions, and promotes discoverability with a logical information hierarchy (Crum, 2020).

• Collaboration Features: Look for real-time editing, shared workspaces, and collaborative tools that enhance teamwork. Verify alignment with industry-standard formats and assess platform-specific features like commenting annotation and access control (Bergstrom & Schall, 2014).

• Visual Programming Language Support: Evaluate the variety and complexity of visual programming functionalities. Test ease of use for non-programmers and their ability to build efficient workflows. Assess the availability of drag-and-drop functionalities, pre-built components, and the ability to extend workflows with scripting or low to no-code approaches.

• Interactivity and Navigation: Measure time spent navigating and manipulating elements within the digital twin. Consider user feedback on the fluidity, responsiveness, and intuitiveness of controls. Evaluate support for multiple views, camera angles, drill-down capabilities, real-time data visualization, and interactive simulations (Spolsky,2008; Bergstrom & Schall, 2014).

• Security: Analyze industry compliance certifications (e.g., ISO/IEC 27001:2022) and platform-specific security features. Evaluate user authentication, data encryption, and access control mechanisms. Assess support for multi-factor authentication, role-based access control, audit logging, intrusion detection, and integration with existing security infrastructure.

5.3 Sensor-Actuator connections: Control group (facility managers) and Innovators group (researchers and academia)

The sensor connection will be evaluated based on three main categories: Model Representation, Data Streaming and Integration Criteria, and Interoperability and Collaboration Criteria (Kalantari et al. 2022, Khallaf et al. 2022).

Table 4: Evaluation Criteria – Control Group

Criteria & Description

• Level of Detail (Geometry, Materials, Systems, Operational Characteristics) Thoroughly assess the depth of representation, considering geometry intricacies, material specifics, system intricacies, and operational characteristics in the sensor connection model.

• Data Resolution and Sensor Coverage: Evaluate the Model's capability to achieve high data resolution and ensure comprehensive sensor coverage, allowing for a detailed and encompassing representation.

• Validation Methods (Comparison with Physical Measurements, User Feedback): Establish rigorous validation procedures, incorporating comparisons with physical measurements and gathering user feedback to ensure the accuracy and reliability of the sensor-connected data.

• Realtime and Historical Data Streaming: Evaluate the system's ability to seamlessly stream both real-time and historical data, ensuring a dynamic representation of the environment over time.

• Data Acquisition (Sensors, IoT Networks): Assess the efficiency of data acquisition mechanisms, considering the integration of various sensors and IoT networks for comprehensive data collection.

• Data Processing: Examine the data processing capabilities, focusing on the system's efficiency in managing and processing the diverse data acquired through sensors and IoT networks.

• Internal Simulation Capabilities: Evaluate the Model's internal simulation capabilities, assessing its ability to simulate different scenarios and environmental changes.

• Connection to External Simulation Applications: Determine the Model's compatibility and seamless integration with external simulation applications, allowing for enhanced simulation capabilities and broader insights.

• Integration with Other Systems and Platforms: Assess the Model's capacity to integrate seamlessly with other systems and platforms, fostering collaboration and data exchange across diverse environments.

6. COMPARSON MATRIX

The adoption of Digital Twins in construction has surged thanks to advancements in GPU hardware and the launch of specialized software platforms by major companies—Autodesk’s Tandem, NVIDIA’s Omniverse, and Epic Games’ Unreal Engine. These platforms, which support the building lifecycle from design to maintenance, were compared using a structured matrix based on key criteria such as visualization, user interface, and sensor connectivity. The criteria were derived from industry standards and expert insights, and the evaluation process included feedback from architects, engineers, and researchers. This comprehensive analysis helps stakeholders select the most suitable DT platform for effective building lifecycle management.

Table 5. A matrix comparative analysis of the platforms used to create DT: Tandem, Unreal Engine, and Omniverse

7. DISCUSSION

The comparison matrix evaluates Digital Twin (DT) platforms across three dimensions tailored to different user groups. For architects and engineers (Development Group), visualization criteria include rendering quality, real-time collaboration, compatibility, scalability, performance optimization, and custom views. The user interface is judged on intuitiveness, collaboration, visual programming support, interactivity, navigation, and security. For facility managers and researchers (Control and Innovators Groups), sensor-actuator connections are assessed based on model representation, data streaming, integration, and interoperability. This framework ensures that DT platforms meet diverse user requirements and perform optimally throughout the project lifecycle.

When comparing Autodesk Tandem, Epic Games Unreal Engine, and NVIDIA Omniverse, each platform displays unique strengths. Autodesk Tandem offers robust support for multiple 3D formats but lacks convertibility and visual programming support, which can be challenging for beginners. In contrast, Epic Games Unreal Engine provides strong convertibility, visual programming, navigation, and compatibility with MR, VR, and AR, along with real-time ray tracing. NVIDIA Omniverse excels in AI compatibility and real-time ray tracing. All three platforms offer excellent support for 3D asset libraries and customizable views, though sensor connection support varies, underscoring the importance of aligning platform choice with specific project needs.

8. CONCLUSIONS

The chapter reviews the early stages of Digital Twin (DT) development in the AECO domain, emphasizing the need to systematically transition from BIM to DTs. It highlights that while digital tools have revolutionized construction and facility management, BIM’s full potential for dynamic operation remains underexploited, largely due to challenges in data interoperability and automation. A user-defined perspective is essential for comparing DT platforms, which necessitates effective collaboration between developers, facility managers, and innovators.

A comparative matrix is presented for Autodesk Tandem, Epic Games Unreal Engine, and NVIDIA Omniverse, showcasing each platform’s strengths and weaknesses—Tandem's robust 3D support, Unreal Engine’s ease of conversion and user-friendliness, and Omniverse’s AI compatibility and real-time ray tracing, despite some limitations. The chapter also stresses the importance of integrating BIM data with sensors, actuators, and AI algorithms, proposing that future research should focus on criteria for seamless AI integration to enhance decision-making and analytics within DT environments.