DEPARTMENT OF ARCHITECTURE AND CIVIL ENGINEERING DIVISION OF CONSTRUCTION MANAGEMENT CHALMERS UNIVERSITY OF TECHNOLOGY Gothenburg, Sweden 2021 www.chalmers.se Digital Transformation in Commercial Real Estate A Case Study on Creating Digital Twins of Existing Buildings Master’s thesis in Design and Construction Project Management REBECCA ENGVALL ANNA LARSSON Master’s thesis 2021 Digital Transformation in Commercial Real Estate A Case Study on Creating Digital Twins of Existing Buildings ANNA LARSSON REBECCA ENGVALL Department of Architecture and Civil Engineering Division of Construction Management Chalmers University of Technology Gothenburg, Sweden 2021 Digital Transformation in Commercial Real Estate A Case Study on Creating Digital Twins of Existing Buildings ANNA LARSSON REBECCA ENGVALL © ANNA LARSSON, 2021. © REBECCA ENGVALL, 2021. Examiner: Mikael Johansson, Division of Construction Management Master’s Thesis 2021 Department of Architecture and Civil Engineering Division of Construction Management Chalmers University of Technology SE-412 96 Gothenburg Telephone +46 31 772 1000 Gothenburg, Sweden 2021 iv Digital Transformation in Commercial Real Estate A Case Study on Creating Digital Twins of Existing Buildings LARSSON, ENGVALL Department of Architecture and Civil Engineering Chalmers University of Technology Abstract Smart cities use information and communication technologies in response to urban challenges to ensure sustainable development. Such cities are networks of digital in- frastructures which comprise multiple integrated and interactive components, where smart buildings are real estate components that adapt their operations in response to gathered data. Real estate has ample opportunities to contribute to smart cities by digitalizing their portfolios, however, the effort for existing buildings with lim- ited digital infrastructure is unclear. This thesis aim was to research how existing buildings can be transformed into smart buildings using digital twins by studying a case where a commercial real estate portfolio is digitalized. To realize the aim, ten qualitative interviews were held and provided material was analyzed. The results suggest that digital twins in real estate can be described as a digi- tal ecosystem where data flows from existing facility systems to a data cloud that enables applications to access the data. This is an industrial approach where in- teroperability is solved using ontologies and real estate owners utilize its existing platforms rather than purchasing a packaged product that hampers interoperability further. Previous research identified the benefits of digital twins in operations and maintenance. However, real estate is undergoing a transformation from asset to service provision where emerging business models entail service packages of flexible nature. In those business models, data can play a significant role in analyzing how spaces are used, best utilized, and easily accessed. The modeling to support this does not require high fidelity and existing basis can be sufficient. The models are in- stead refined to higher fidelity levels as changes occur in facilities’ life cycle and thus enable object-oriented benefits. Life cycle management, i.e., keeping information updated, is argued as the primary challenge. However, it appeared that it would be sufficient incentives for life cycle management if the business model is dependent on it. To succeed, strict guidelines are required in project commissioning and day-to- day information management. User provisioning is identified as an essential tool for operative real estate personnel to easily add or alter information. Keywords: commercial real estate, digital transformation, digital twins, existing buildings, smart buildings. v Acknowledgements This thesis was carried out at Chalmers University of Technology in the spring of 2021. We would like to begin by thanking our supervisor Mikael Johansson at the Division of Construction Management for his support, critical feedback, and all our inspiring discussions. The idea of the topic has been developed by the authors with support from Fredrik Ahl at Sweco. We would like to thank him for his guidance, encouragement, and time spent discussing this interesting topic with us. Furthermore, we would like to thank all the interviewees who contributed with knowledge, reflections, and insights. Without them, this thesis would not have been possible to complete. Finally, we would like to thank our families for their eternal support throughout our education at Chalmers. Anna Larsson & Rebecca Engvall, Gothenburg, May 2021 vii Contents List of Figures xi List of Tables xii List of Abbreviations xii 1 Introduction 1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.2 Aim and Research Questions . . . . . . . . . . . . . . . . . . . . . . . 3 1.3 Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 1.4 Outline of the Thesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2 Theoretical Framework 5 2.1 Real Estate and Facilities Management . . . . . . . . . . . . . . . . . 5 2.1.1 Facilities Planning . . . . . . . . . . . . . . . . . . . . . . . . 6 2.1.2 Project and Transaction Management . . . . . . . . . . . . . . 7 2.1.3 Service, Operations, and Maintenance . . . . . . . . . . . . . . 7 2.2 Information Management . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2.1 Facility Information . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2.2 Building Information Model . . . . . . . . . . . . . . . . . . . 9 2.3 Digital Twin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 2.3.1 Data Acquisition Layer . . . . . . . . . . . . . . . . . . . . . . 12 2.3.2 Digital Modeling Layer . . . . . . . . . . . . . . . . . . . . . . 13 2.3.3 Data and Model Integration Layer . . . . . . . . . . . . . . . 13 2.3.4 Service Layer . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.4 Implementing Technology in Organizations . . . . . . . . . . . . . . . 15 2.4.1 Benefits Management . . . . . . . . . . . . . . . . . . . . . . . 15 2.4.2 User Acceptance . . . . . . . . . . . . . . . . . . . . . . . . . 17 3 Method 19 3.1 Research Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.1.1 Literature Research . . . . . . . . . . . . . . . . . . . . . . . . 20 3.1.2 Case Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.1.3 Data Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 3.2 Critical Evaluation of Method . . . . . . . . . . . . . . . . . . . . . . 23 3.3 Ethical Statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 ix Contents 4 Results: Case Study 25 4.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.2 SimpleBuilding: Creating Spatial Models . . . . . . . . . . . . . . . . 25 4.3 Twinfinity: Contextualizing Digital Twins . . . . . . . . . . . . . . . 27 4.4 Collaborative Digital Twins: An Approach . . . . . . . . . . . . . . . 29 4.4.1 IT Strategy . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.4.2 Digital Aim . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30 5 Results: Case Interviews 32 5.1 Digital Transformation in Real Estate . . . . . . . . . . . . . . . . . . 32 5.1.1 Investment objectives . . . . . . . . . . . . . . . . . . . . . . . 32 5.1.2 Current Issues in Facilities . . . . . . . . . . . . . . . . . . . . 33 5.1.3 Role of Digital Twins . . . . . . . . . . . . . . . . . . . . . . . 35 5.2 Usage Areas of Digital Twins in Real Estate . . . . . . . . . . . . . . 36 5.2.1 Facilities Planning . . . . . . . . . . . . . . . . . . . . . . . . 37 5.2.2 Project and Transaction Management . . . . . . . . . . . . . . 37 5.2.3 Service Management . . . . . . . . . . . . . . . . . . . . . . . 39 5.2.4 Operations and Maintenance Management . . . . . . . . . . . 41 5.3 Challenges of Digital Twins in Real Estate . . . . . . . . . . . . . . . 43 5.3.1 Enabling Activities . . . . . . . . . . . . . . . . . . . . . . . . 43 5.3.2 Sustaining Changes . . . . . . . . . . . . . . . . . . . . . . . . 45 6 Discussion 47 6.1 Digital Ecosystem of Data, System and Service . . . . . . . . . . . . 47 6.2 Simple Modeling Enabling New Business Models . . . . . . . . . . . . 50 6.3 Enabling and Sustaining the Digital Ecosystem . . . . . . . . . . . . 52 7 Conclusion 56 7.1 Answering the Research Questions . . . . . . . . . . . . . . . . . . . . 56 7.2 Industry and Academia Contribution . . . . . . . . . . . . . . . . . . 57 7.3 Future Research Suggestions . . . . . . . . . . . . . . . . . . . . . . . 58 References 64 x List of Figures 2.1 The “built environment” life cycle (based on Ebinger and Madritsch, 2011). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 The Built Environment Management model BEM2 (based on Ebinger & Madtrisch, 2012). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2.3 The structure of digital twin (based on Boje et al. 2020). . . . . . . . 10 2.4 System architecture of a digital twin at building level (based on Lu et al., 2019 and Lu et al., 2020-a). . . . . . . . . . . . . . . . . . . . . 11 2.5 The four main analytic processes of information, structured from low- est to highest maturity level (based on Klein et al., 2019. . . . . . . . 14 2.6 Benefits Dependency Network (based on Love and Matthews, 2019 and Peppard 2016). . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 2.7 Technology Acceptance Model (based on Davis, 1993). . . . . . . . . 18 3.1 Research process (authors own figure). . . . . . . . . . . . . . . . . . 20 4.1 SimpleBuilding (by Sweco). . . . . . . . . . . . . . . . . . . . . . . . 26 4.2 SimpleBuildingPlus (by Sweco). . . . . . . . . . . . . . . . . . . . . . 26 4.3 Sweco’s process flow of creating SimpleBuildings (based on provided material from Sweco). . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.4 Twinfinity’s extraction engine (based on provided material from Sweco). 27 4.5 Twinfinity’s viewer (by Sweco). . . . . . . . . . . . . . . . . . . . . . 28 4.6 Case company’s strategy of information technologies (based on pro- vided material by the case company). . . . . . . . . . . . . . . . . . . 29 6.1 Current state where facility systems are placed in silos and managed in separate applications (authors own figure). . . . . . . . . . . . . . 48 6.2 Digital twin as an ecosystem of data, system, and service (authors own figure). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50 6.3 The Benefits Dependency Network of the studied case (authors map- ping based on model by Love and Matthews, 2019 and Peppard, 2016). 55 xi List of Tables 3.1 Overview of interviewees. . . . . . . . . . . . . . . . . . . . . . . . . . 22 List of Abbreviations AI BAS BEM BIM ICT IoT LOD Artificial Intelligence Building Automation Systems Built Environment Management Model Building Information Model Information and Communications Technology Internet of Things Level of Detail 1 | Introduction The first chapter gives an introduction of why digital twins are an important enabler of the pursuit for sustainable solutions in the real estate industry. It will also introduce the barrier for investing in digital twins which will form the foundation for the research questions which the thesis will examine. Lastly the chapter will present the limitations and outline of the thesis. 1.1 Background The world faces issues of climate change, demographic change, and urbanization (Baum et al., 2019). These issues are pressuring cities to innovatively manage overcrowding, energy consumption, resource management, environmental protec- tion, and assure a high quality of urban inhabitants’ lives (Brutti et al., 2019). Smart cities is a concept where Information and Communication Technologies (ICT) are used in response to the urban challenges to ensure sustainable development (Brutti et al., 2019; Baum, 2017). Smart cities are networks of digital infrastructure net- works which comprise multiple smart components which form the cities (Brutti et al., 2019). Smart buildings are real estate components using ICTs to adapt their operations and physical form in response to gathered data (Lecomte 2019). Few real estate companies have approached smart buildings (Baum et al., 2019). Historically, the real estate sector has been able to stay profitable using traditional business models, but commercial real estate has experienced a shift in demands from tenants (Lequeux and Guiot, 2019; Starr et al., 2020). In combination with the pursuit of corporate sustainability, companies have started to reflect on their current business models (Lequeux and Guiot, 2019). Therefore, commercial real estate companies are starting to examine how an investment in digital innovations might affect their sustainability and competitive positioning (Baum et al., 2019). Digital innovations have changed the structure of industries for many years. In the mid-twentieth century, it concerned incorporating digital computers and the advancement of information technologies. The fourth industrial revolution (industry 4.0), the digital transformation advanced to gaining knowledge from data analytics, cloud computing, and Internet of Things (Starr et al. 2020). Industry 4.0 emerged from the manufacturing industry but is now influencing real estate. Proptech is the real estate industry’s response to industry 4.0, whereas one technology that has risen in interest is the digital twin (Adamenko et al., 2020; Starr et al., 2020). 1 1. Introduction A digital twin is a digital replica of physical assets, processes, and systems, that learn from multiple sources to predict current and future conditions (Lu et al., 2020-a). The technology is described as valuable for real estate owners to optimize develop- ment, operation, and service processes by analyzing the building lifecycle (Adamenk et al., 2020). It is also described as a tool for smart building management to min- imize consumed energy and carbon emissions to enhance corporate sustainability (Baum et al., 2019; Lu et al., 2020-b). However, today, several challenges and bar- riers obstruct a company’s willingness to invest in digital twins (Baum et al., 2019; Jones et al., 2020). Foremost, there is no consolidated view in the real estate and construction indus- try regarding digital twins, resulting in confusion (Jones et al., 2020). Additionally, there are gaps in research regarding if the benefits will outweigh the effort of develop- ing and integrating the digital twin for existing buildings in the real estate business and operation. The creation of digital twins is dependent on building information which, in many cases, is lacking for existing buildings (Lu et al., 2020-c). Few stud- ies have been conducted on the creation and benefit of digital twins in the scope of existing buildings. However, some literature suggests that digital twins might be more profitable for newly built buildings where models developed during the design and construction stages already exist (Koch, 2018). Around 80 % of existing build- ings are constructed before 1990 and consequently lack efficient digital information (Volk et al., 2014). At the same time, these buildings are the majority of real estate portfolios. Therefore, existing buildings are a great resource for commercial real estate to enable its business and sustainability objectives and realizing smart cities. Smart cities are far from realization, and one obstacle is the inadequate digital infrastructure (Baum et al., 2019). The smart building components are, on the contrary, achievable to pursue by a digital twin aided the digital transformation of real estate companies. However, the barriers of digital twins in the real estate sector need to be addressed to attract early movers (Baum et al., 2019). To fill the knowledge gaps, this thesis will investigate a case composed of developing and implementing digital twins for existing commercial buildings. The case includes the consultancy firm Sweco’s BIM-platform Twinfinity and SimpleBuilding concept, followed by a real estate company’s approach to digital twins and intended use of Twinfinity. 2 1. Introduction 1.2 Aim and Research Questions This master thesis aims to contribute to the research of smart cities by investigat- ing how existing buildings can be digitalized from a commercial real estate owner’s perspective. This aim will be realized by examining the concept of digital twins and how one can be formed in a real estate context. Three research questions were stated to support the aim. • RQ1: How can digital twins be described in a real estate context? • RQ2: How can digital twins create benefits in commercial real estate consid- ering existing buildings? • RQ3: What are the prerequisites for real estate companies to enable and sustain the digital twins? 1.3 Limitations This master thesis will conduct a single case study composing an ongoing collab- oration between Sweco and the case company to answer the research questions. The case company is a commercial real estate owner, including offices and retail. Therefore, the thesis is limited to their reality and values. The empirical data is collected from this case study and is not compared to other empirical cases. The methodological choice is further motivated and evaluated in chapter 3 Method. The thesis will not study the development of smart cities but rather the smart build- ing as an isolated component which, in the future, is an essential enabler of smart cities. Moreover, several ethical questions, e.g., privacy and security, arise from smart technologies. Such are excluded in the scope of this study. Last, the litera- ture that constructs the theoretical framework was mainly collected from realities that resemble the Swedish’s. 3 1. Introduction 1.4 Outline of the Thesis • Chapter 1 - Introduction: Introduces the thesis’s background and problem formulation, followed by aim, research questions, and limitations. • Chapter 2 - Theoretical framework: The theoretical key concepts used from previous research to analyze the empirical material are framed. • Chapter 3 - Method: Describes the methodological strategy that was used to conduct the thesis. It includes research design, data collection from case study, critical evaluation, and an ethical statement. • Chapter 4 - Result: Case study Provides the case background and an overview of the technologies. • Chapter 5 - Result: Case interviews: Provides the case company’s reflec- tions on the topic, including digital transformation, usage areas, and challenges of digital twins. • Chapter 6 - Discussion: Compares the theoretical framework to case results directly related to the research questions. • Chapter 7 - Conclusion: Concludes the research questions, states the thesis contribution to academia and industry, and suggests future research. 4 2 | Theoretical Framework The theoretical framework intends to create an understanding of key concepts that are used to investigate the research questions. Real estate and facilities management and the key processes areas it comprises be defined. This is followed by introducing information management in facilities. Then, digital twin and its related concepts will be explored, focusing on the building level. Lastly, theories in change management will be addressed to grasp the implementation of new technology in an organization. 2.1 Real Estate and Facilities Management The core business of a real estate organization is to aggregate and exploit a portfolio of real estate assets (Glickman, 2013). The processes which this entails can jointly be summarized under the name real estate management. As Wirdzek (2010) states, the emerging integrated data models for the built environment need a clear classification of its processes. There are many attempts to complete such categorization, but many labels that are used fragmentize the real estate literature (Ebinger and Madritsch, 2011). Figure 2.1: The “built environment” life cycle (based on Ebinger and Madritsch, 2011). This thesis’s notion of real estate management is derived from an industry-neutral classification framework in real estate literature, namely the built environment management model (BEM2) by Ebinger and Madritsch (2011) and Ebinger and Madritsch (2012). It is a comprehensive and process-based framework that compiles 5 2. Theoretical Framework real estate and facilities management from a two-dimensional approach, comprising the built environment management functions and the value stream from strategic to portfolio and operational. The researchers recognized a cyclical pattern in the built environment management functions and arranged the functions accordingly (see Figure 2.1). The life cycle of the built environment was further elaborated in an organizational environment. A series of sequential and interdependent key process areas (KPA) were then identified: KPA 1; strategic planning, KPA 2; facilities planning, KPA 3; project and transaction management and KPA 4; services, operations, and main- tenance management. Similar to methodologies in project management, it is advo- cated that the processes occur in a hierarchical organizational structure. A tacti- cal setting should support the operational project work to reach the organization’s strategic aim (PMI, 2008). This resulted in the industry-neutral and process-based framework, as it is the foundation of all real estate organizations to plan, provide, service, and maintain their built environment (Ebinger and Madritsch, 2011) (Fig- ure 2.2 ). Strategic planning, KPA 1, is an enterprise function that revolves around defining vision and mission, business strategies, and strategic objectives. An in- depth description of KPA 2 - 4 follows. Figure 2.2: The Built Environment Management model BEM2 (based on Ebinger & Madtrisch, 2012). 2.1.1 Facilities Planning The strategic objectives of KPA 1 are translated in facilities planning, KPA 2, with the aspiration to optimize capital investment decisions in the facilities portfolio. In 6 2. Theoretical Framework a real estate portfolio, reinvestments for maintaining existing facilities compete with requirements for new facilities. The different categories of reinvestments and new requirements are balanced altogether, and approved projects are handed over to KPA 3, project management, and transaction (Ebinger and Madritsch, 2012). 2.1.2 Project and Transaction Management KPA 3, project and transaction management, materializes the decisions in KPA 2 by performing a construction project or real estate transaction. The strategic goal of this key process area is to achieve optimal project performance with minimal disruption and where the project delivery satisfies the organization’s requirements (Ebinger and Madritsch, 2012). According to the BEM2, the tactical level involves coordinating all capital projects and transactions that are planned in a real estate portfolio. The sub-processes are defined as: program management of resources and team members for effective operation, monitoring risk, meeting regulatory require- ments, maintaining client relationships, and assessing performance measurement. The operational level of project and transaction management revolves around im- plementing the capital projects or transactions established by facility planning. An operative real estate representative often manages the project, but other processes are outsourced to architects, engineers, construction managers, and contractors. A project is then divided into the sub-areas: planning, implementation and control, and commissioning. Project planning expands the project assumptions identified in KPA 2, facilities planning, with a detailed initial study of the project. Project plans and specifications are then developed for implementation. During implementation, the construction is controlled by the real estate representative. After construction competition, the project is commissioned to service, operations, and maintenance management, KPA 4, that manages the facility. The commission process includes handing over relevant building information, i.e., maintenance schedules, warranties, 2D/3D models, and as-built documents (Ebinger and Madritsch, 2012). 2.1.3 Service, Operations, and Maintenance This operative key process area has the strategic goal to uphold the desired perfor- mance of existing facilities and provide an optimal environment for the organization leasing the premise (Ebinger and Madritsch, 2012). The tactical level is responsible for the operational function’s delivery. Like project and transaction management, the tactical level includes managing resources, risks, client relationships, and per- formance. It also constitutes a constant reviewing of facilities condition and iden- tification for renewals. The sub-areas of service, operations, and maintenance are described below. Service management includes all the service processes supporting the facility’s core business requirements and working environment. It includes a broad spectrum of operational services such as: office support, technical, janitorial, food, and lease and space management (Ebinger and Madritsch, 2012). Lease management concerns 7 2. Theoretical Framework finding potential tenants that match the facilities’ available spaces and continuous management of the tenant relationship (Ebinger and Madritsch, 2012; Glickman, 2013). Space management is a critical function in facilities management, including efficient and cost-effective use of space (Atkin and Brooks, 2015). This is done by incorporating flexible and adaptable features that support different activities of dif- ferent tenants and systematically collecting information about space utilization to increase cost-efficiency (Atkin and Brooks, 2015). Operations management entails performing and managing the processes that ensure the work environment facilitates an effective operation of core businesses in organiza- tions (Ebinger and Madritsch, 2012). These processes include operating the facility systems (HVAC, Electrical, Plumbing) and the utility and energy management of these systems (Ebinger and Madritsch, 2012). Maintenance management is responsible for the preventive and reactive maintenance of the existing asset portfolio. Preventive maintenance activities are either time- based or condition-based (Mangano and de Marco, 2014). Time-based maintenance is planned maintenance that is set to be completed in a specific interval based on manufacturing information of building components. Condition-based maintenance is based on consistent monitoring of building component’s conditions and identification of abnormalities. The opposite to preventive maintenance is reactive, or corrective, maintenance. It is the maintenance of unplanned breakdowns where the action is only taken when a failure occurs (Mangano and de Marco, 2014). Errandonea et al. (2020) include additional categories for maintenance that rely on computer analytics, namely predictive and prescriptive maintenance. The former is based on statistical analytics of asset information to predict the remaining life. The latter is based on analytics of information of the condition monitoring to predict the current asset status. 2.2 Information Management Information management processes generally include collecting, structuring, storing, analyzing, and updating information and data (Atkin and Brooks, 2015). Data is an objective fact, while information is processed data and is, therefore, the resource that is valuable in decision making-processes (Parsanezhad, 2015). This chapter will describe information management in facilities, followed by introducing Building Information Model. 2.2.1 Facility Information Atkin and Brooks (2015) state that information is the “lifeblood” for efficient real estate and facilities management. The processes generate a large amount of data, and if it is captured, it can be used for efficient decision-making. In facilities, there are many tools and systems used to improve the processes, and those are filled with information. Various systems usually steer the processes in service, operations, and maintenance management. Building Automation Systems (BAS) is used to manage 8 2. Theoretical Framework HVAC, electrical, mechanical, and plumbing systems. Daily maintenance and other activities within service management are supported by Computer-Aided Facilities Management (CAFM). The commissioning process from project finish to facilities management is too filled with facility information. Technical information concerning the facility’s as-built in- formation, including its design and construction, is suggested to be considered in the context of Building Information Models (BIM) (Atkin and Brooks, 2015). However, the exchange and re-usability of the data in these systems and the fragmented na- ture of the industry make the information transfer among stakeholders challenging (Yalcinkaya and Singh, 2014). 2.2.2 Building Information Model Building information model (BIM) is growing in interest in the real estate sector due to its effective information management across the facilities’ life cycle (Atkins and Brook, 2015). BIM can hold information about the physical attributes of a facility, i.e., geometry, components, systems, the spatial relationship, and attached non-geometric information (Atkins and Brook, 2015). Some common examples of physical attributes are floor plans, windows, structural information (e.g., columns and beams), and arrangement of the HVAC system (Klein et al., 2012; Lu et al., 2019). Besides geometric information, objects may also have non-geometric at- tributes, commonly in text format (Fitz and Saleeb, 2019). By adding such at- tributes, objects can hold information about, e.g., where its service zone is, how much energy is used and when it was last maintained (Becerik-Gerber, 2012). In- dustry Foundation Classes (IFC) is a format for BIM with standard specifications that makes it possible to hold and exchange information between various software applications (Atkins and Brook, 2015). A common way of characterizing a BIM model is the level of detail (LOD) (Alshora and Ergen 2021, Bedrick, 2008). LOD is used to define the attributes of geometric and non-geometric data in the design of construction projects (Volk et al., 2014). Therefore it is not fully applicable in facilities management. There are five levels in the scale of LOD ranging from 100 to 500 explained below. • LOD 100 - Conceptual: Non-geometric data or lines, areas, volumes zones. • LOD 200 - Approximate geometry: Generic elements shown in 3D. • LOD 300 - Precise geometry: Specific elements confirmed in 3D geometry. • LOD 400 - Fabrication: Shop drawing or fabrication. • LOD 500 - As-build: Representation of the component as it is built. In a mature process-oriented industry, BIM is used during the whole life cycle (Thy- 9 2. Theoretical Framework dell, 2017). The model is then categorized depending on when it was last configured. As-designed models emerge during the design phase and detailed planning. The as- designed model is updated after construction with observed differences to an as-built model. As-is models are modeled from existing in-use buildings where no previous reliable building document or representation exists (Becker et al., 2019). The bene- fits of BIM are well recognized in the design and construction phases (Becker et al., 2019). It is expected to be beneficial in facilities management as well. However, the adoption of existing facilities is less confident compared to the cost of creating the model initially (Atkins and Brook, 2015; Becker et al., 2019). 2.3 Digital Twin The concept of digital twins has greatly increased in interest over the past couple of years, both by industries and as a research topic by academia (Jones et al., 2020; Lu et al., 2019). It was introduced in 2003 when Michael Grieves held a lecture in Product Lifecycle Management. It was then defined as "a reengineering of structural life prediction and management" (Boje et al., 2020). The concept was initially used within the aerospace field, later appeared increasingly in the manufacturing indus- try and more recently within the built environment and smart cities. Most research applications of digital twins are within maintenance because its significant impact on companies (Errandonea et al., 2020). However, the definition of a digital twin varies between literature and industries; hence there is no consolidated view of what it is (Errandonea et al., 2020; Jones et al., 2020). The idea behind the digital twin is to deliver and receive product data that can be used to understand a physical product during its lifecycle (Errandonea et al., 2020). Several studies describe it as a virtual-physical integration. In other words, a combination of the physical component, its virtual counterpart, and the data which connects them (see Figure 2.3) (Boje et al., 2020; Errandonea et al., 2020; Jones et al., 2020). The physical component represents the real-world spaces and real- time data that the virtual component mirror (Al-Ali et al., 2020; Boje et al., 2020). The virtual model involves data aggregation of the collected data and the virtual modeling of the physical component (Al-Ali et al., 2020). Figure 2.3: The structure of digital twin (based on Boje et al. 2020). 10 2. Theoretical Framework The term fidelity described the transferred data between the physical and virtual components. Fidelity is defined as "The term fidelity describes the number of param- eters, their accuracy, and level of abstraction that are transferred between the virtual and physical twin/environment" (Jones et al., 2020, p.40). Jones et al. (2020) ex- plain three levels: fully comprehensive, ultra-realistic, and high fidelity. In industry, other characterization methods have been identified. Arup (2019) uses five levels. The first level has low fidelity and is described as a conceptual model, and the fifth level has high fidelity and is described as a model with a high degree of accuracy. However, this method has not been acknowledged in the literature as an accepted categorization of a digital twin (Jones et al., 2020). Both Boje et al. (2020) and Jones et al. (2020) argue that high fidelity is an unachievable goal and that the level of precision instead should mirror what is required by the use case. Hossain and Yeoh (2018) conclude that the higher the fidelity, the higher the cost. The fidelity level should therefore be appropriate to maximize benefits while minimizing expenses and technical difficulty (Hossain and Yeoh, 2018; Jones et al., 2020) The technical difficulty is determined by the different systems and components that construct the digital twin (Jones et al., 2020). To explain how a digital twin at a building level, a system architecture framework was constructed by Lu et al. (2019), Figure 2.4. The architecture is built on four layers where data flows from a data acquisition layer through a transmission layer into a data/model integration layer. The layers are complemented with a modeling and data complementary layer. The service layer is where applications are formed. Figure 2.4: System architecture of a digital twin at building level (based on Lu et al., 2019 and Lu et al., 2020-a). 11 2. Theoretical Framework 2.3.1 Data Acquisition Layer Data is the foundation of the digital twin (Boje et al., 2020; Uhlenkamp et al., 2019; Lu et al., 2019). Data acquisition entails collecting data from the physical product and its environment. Errandonea et al. (2020) explained that a digital twin used for maintenance should contain building data along with operational, organizational, and technical data. Furthermore, some common data sources for digital twin at the building level addressed in literature are: building management data (e.g., BAS); asset management and space management data (e.g., CAFM); and real-time sensor data (Cheng et al., 2020; Fitz and Saleeb, 2019; Fukuda et al. 2014; Lu et al. 2019). Sensing data is real-time data describing the operational data in the building and environment with physical sensors (Cheng et al., 2020). Internet of Things (IoT) network allows the transfer of data from the physical sensor device to the data/- model integration layer (Cheng et al., 2020; Jones et al., 2020). Lu et al. (2019) mentioned several types of data collected through IoT-enabled wireless sensing net- works, among other sensors measuring indoor climates such as temperature, humid- ity, carbon monoxide, motion, vibration, and light detectors. As data is acquired, the data is transferred to the data/model integration layer by the transmission layer, which composes the next step in the digital twin system architecture (Lu et al., 2019). Lu et al. (2019) explain that the main challenge of data integration is how to integrate various sources from various autonomous, disparate, and heterogeneous sources. Lack of data is not the main issue of existing digitalization approaches but instead the lack of structured data (Cajias, 2020; Lu et al., 2020-b). Unstructured data is data that lack the required organization for being able to be readable and analyzed by machines (Gandomi and Haider, 2015). Structured data is consequently digitally organized in the format required by the machines and, in this case, the analyzing functions of the digital twin. Lu et al. (2019) add that another important aspect is data quality. This implies the data’s ability to support the various applications by ensuring the quality meets its requirements. Lu et al. (2019) explain four main reasons why low data quality may occur. The first reason is simply that the quality of the original data source is low. Another reason could be that the extraction process is of low quality resulting in the degrading of data, or the integration process of data could result in quality loss. The last reason is that the process for acquiring and integrating the data is developed for supporting an application of lower data quality requirements than the intended application (Lu et al., 2019). Cajias (2020) summarizes that defining the pipelines of collecting, cleaning, and organizing data is a challenging process that demands a substantial amount of human resources. Many actors struggle with this process. However, Lu et al. (2019) adds that the process has to be well designed to generate value (Caijas, 2020). 12 2. Theoretical Framework 2.3.2 Digital Modeling Layer The digital model is the geometric representation of the building components and systems (Boje et al., 2020). The most common modeling approach for the digital model in buildings is BIM (Lu et al., 2020-c). However, Boje et al. (2020) state that BIM is a static model, and it must be adapted to a semantic web paradigm to be used as a digital model. Otherwise, cyber-physical integration will be lost (Boje et al., 2020). Some natural sources of collecting building geometry information are drawings, mod- els, and other documents created during design and construction phases (Klein et al., 2012; Lu et al., 2020-c). Jones et al. (2020) note that this information needs to be updated to create value. This means that the as-designed should be up- dated to as-built representation and continuously updated throughout the life cycle when changes are made. However, Klein et al. (2012) address that commonly, the as-built drawings used within facility management are not updated into as-is docu- mentation when changes are made. These documents and models representing the building geometry are usually not available for existing buildings (Klein et al., 2012). Lu et al. (2020-c) stated that literature describing how digital modeling should be efficiently executed is limited, but there are established processes. If some materials, such as 2D drawings and hard-copy text documents, are available, one method is to model from the existing material and complement manual verification (Klein et al., 2012; Lu et al., 2020-c). This method has shown to be error-prone and time- consuming to achieve accurate measurements (Klein et al., 2012). Laser scanning and photogrammetry are two building data acquisition techniques that are com- monly used to gather spatial information. Those are more accurate than manual field surveys (Klein et al., 2012). Laser scanning is a technology that calculates the distance to objects by emitting a laser, where the distances are later used to form a 3D point cloud. Digital software is then registering the images to create texture and 3D data. Digital photogrammetry relies on camera-captured images processed through digital software to create 3D geometric information (Hossain and Yeoh, 2018; Klein et al., 2012). These digital data acquisition technologies can either sup- port the modeling and as-is verification of acquired drawings or be used to collect information when no previous building geometry information is available (Klein et al., 2012; Lu et al., 2020-c). However, laser scanning is time-consuming, relatively expensive, and inappropriate for regularly updating as-is models (Lu et al., 2020-c). 2.3.3 Data and Model Integration Layer The data and model integration is the kernel layer where the data are combined (Lu et al., 2019). Processes included are storing, integrating, processing, and analyzing data and models (Lu et al., 2019). Lu (et al. 2019) explains that Application Program Interface (API) enables the transfer and integration of data between the layers. API allows applications to easily communicate with one another (Mathijssen et al., 2020). However, to allow 13 2. Theoretical Framework communication, the data needs to be structured as previously described. Gunes et al. (2014) declare that interoperability is a leading technical challenge with a phys- ical to a digital communication network. The digital twin should be interoperable, which means that it has effective communication and can exchange information be- tween systems. This often requires a common ontology (Gunes et al., 2014; Klein et al., 2019). An example of data and model integration is the combination of sensing data such as temperature with a modeled component representing the geometric positioning of that sensor (Stenberg, 2018). Analytics is a part of the data/model integration layer, as stated in the system ar- chitecture. Jones et al. (2020) explain that analytics are virtual processes where software, algorithms, and other computational techniques are used. Klein et al. (2019) describe four levels of data analytics (see Figure 2.5 ). Descriptive analytics aims at describing what is and what has happened by visualizing the current and past performance, for example, in the form of visualizing sensing data. Diagnostic, predictive, and prescriptive analytics are described as advanced data analytics where prescriptive analytics represents the highest maturity level. These can be achieved through artificial intelligence and machine learning (Boje et al., 2020; Klein et al., 2019). However, data analytics also require a lot of data and high data quality for efficient analytics (Boje et al., 2020). Diagnostics analytics answer why something happens by finding causes, reasons, and patterns in the descriptive data. The diagnostic analysis could be used to support condition maintenance (Errandonea et al., 2020). Furthermore, predictive analytics aims at describing what will happen based on statistics or machine learning. Pre- scriptive analytics aims to answer how the future performance can be improved by evaluating alternatives, using simulations and optimization processes to recommend action and support decision making (Klein et al., 2019). The analytics processes depend on the previous step’s results from the previous steps. A higher analytics maturity level decreases human involvement decreases and increases the dependency on intelligence and learning ability. Nie et al. (2019) add that most buildings pas- sively react to change, not yet using real-time strategies to describe status. Figure 2.5: The four main analytic processes of information, structured from lowest to highest maturity level (based on Klein et al., 2019. 14 2. Theoretical Framework 2.3.4 Service Layer All the processes composing the digital twin aim to facilitate smart construction ser- vices and applications (Boje et al., 2020). Lu et al. (2019) explain the application layer as the layer which “interacts with the facility managers and provide services for users” (Lu et al., 2019, p.70). Each digital twin application requires its specific data and data quality (Lu et al., 2019). In Lu et al. (2019), the following applications were mentioned: security and health management, energy management, space management, as-is asset monitoring, and facilitation of preventive maintenance. Other expressed applications facilitate the design and construction of facilities and information management, including minimizing information loss and facilitating operational activities (Lu et al., 2020- c). 2.4 Implementing Technology in Organizations To optimize processes and achieve organizational improvement, the PPT theory is a widely used approach which address that the focus for change should be on people, process and technology (Prodan et al. 2015). Change management is a management discipline that includes capabilities for managing individuals, organizations, and in- stitutions through a change process and making the intended transition successful. There are several change management models and theories presented in the litera- ture. This thesis will use the benefits management model by Love and Matthews (2019) to grasp implementing technology in an organizational context by address- ing the drivers. Secondly, this chapter will cover an intention model, namely the Technology Acceptance Model by Davis (1989), for evaluating the user acceptance of information technology. 2.4.1 Benefits Management The “whys” of implementing technologies are usually well understood and agreed upon. Though, the “hows” of realizing the “whys” are rarely documented. The ex- pected benefits are often exaggerated and the management effort is often overlooked (Love and Matthews, 2019). A strategy for successful implementation and realiza- tion of digital technologies benefits is the benefits management strategy composed by Love and Matthews (2019). The strategy is described as a frame of reference for planning change management processes to ensure that the intended technology effectively generates value (Love and Matthews, 2019). Before initiating the benefits management processes, Love and Matthews (2019) address that the organization needs to acknowledge the basic principles of imple- mentation of digital technologies. These are: benefits only materialize from the technological use and when the technology enables people to do things differently. The organization should also remember that technology adoption is a cost, and not all use of technology produces benefits. An unsuccessful implementation might even 15 2. Theoretical Framework affect the organization’s competitive positioning negatively. The change process needs to be carefully planned and actively managed to obtain benefits from tech- nology. Furthermore, benefits management manages and aligns project outputs, results, ben- efits, and organizational strategy. Love and Matthews (2019) present a five-stage process that addresses this management and alignment. The first stage, Identifying and structuring benefits, entails understanding the business drivers for introducing the new technology and identifying the aspired benefit realization from the change. The second stage,Planning benefits realization, entails planning how the benefits will be achieved and the required changes necessary for these processes. The third stage,Executing the benefits realization plan,is where the designed change manage- ment program is being implemented. This execution is later evaluated in the fourth stage,Evaluating and reviewing result, which is continuously done during the whole system lifetime. The fifth and last stage,Discovering the potential for further benefits, is where the organization learns from the evaluation to identify additional opportu- nities. The stages continuously progress from each stage to the next in a circular manner and affect previous stages as results are evaluated (Love and Matthews, 2019). Identifying and structuring benefits entail understanding the business context and identifying the fundamental drivers for the change. Since the benefit of implement- ing a technology only materialize from its use, it is essential that the motivation for implementation meets business and user demand rather than satisfy a trend within the industry. To develop a business case for investing in digital technology, Love and Matthews (2019) suggests that that the organization should answer the following questions: • Why do we need to improve performance? • What improvements do we want/could achieve? • Where will improvements (benefits) occur? – How can we measure the benefits, quantitative (e.g., time, money) as well as qualitative (e.g., customer and employee satisfaction)? – Can a financial value for the realized benefits be determined? • What changes are needed to ensure improvements materialize? – How can the changes be enabled and sustained? • How can the changes be enabled and sustained? • Who is responsible for making changes? • Who will be affected by the change? • How and when can changes be made? These questions will support the mapping of the desired change. After a driver has been identified, the organization should work backward to recognize which business 16 2. Theoretical Framework benefits support the driver. Then, how the technology can enable those benefits and which changes are needed to implement the technology. Related to the digital twin, Lu et al. (2019) suggest that the objectives and how the digital twin will create value should be clear before initiating the process of developing a digital twin. This includes what data the digital twin needs to fulfill this value creation and a well-designed process for collecting, updating, transferring, and integrating the data and the digital model throughout the whole building life cycle (Lu et al., 2019). To map required changes and activities to ensure expected benefits Love and Matthews (2019) used a mapping framework called the Benefits Dependency Network (BDN) developed by Peppard (2016), see Figure 2.6. The role of technology is the technology included in the change. Enabling changes are typically one-of and include both the prerequisites for creating the change and the prerequisites for bringing the new system to operate. Sustaining changes include the technology- enabled permanent changes of working practice, which assures the change’s long- lasting operation and benefit realization. Business benefits are the aspired benefit that will satisfy the driver (Love and Matthews, 2019; Peppard, 2016). Figure 2.6: Benefits Dependency Network (based on Love and Matthews, 2019 and Peppard 2016). 2.4.2 User Acceptance Another aspect of implementing technology in organizations is the behavioral re- sponse from users. This is relevant because no system can benefit organizations if they are not being used (Love and Matthews, 2019). Theories in this area of change management are built on behavioral science and understand how human behavior is associated with technology usage (Kukafka et al., 2003). These theories and models are intention models and explain the mechanisms of how technology is accepted by individuals and subsequently used. 17 2. Theoretical Framework Hilal et al. (2019) explain that the Technology Acceptance Model (TAM) is a widely used intention model. It was developed by Fred D. Davis in 1989 and later tested in a field study in 1993. TAM illustrates the relationship between system design fea- tures, perceived usefulness, perceived ease of use, attitude toward using, and actual usage behavior (Figure 2.7). The model is based on attitude psychology principles that divide behavior into four components: external stimulus, cognitive response, affective response, and behavioral response. The aim of TAM is not only to address why users may accept technologies or not but also aims to improve user acceptance through the design of the system features (Davis, 1993). Figure 2.7: Technology Acceptance Model (based on Davis, 1993). The model suggests that the attitude of a prospective user is the primary deter- minant of behavioral response and actual system use. The attitude toward using the technology is then dependent on two cognitive response variables: perceived usefulness and perceived ease of use. The two variables are affected by the system design features. Perceived usefulness is defined as “the degree to which an individual believes that using a particular system would enhance his or her job performance”, where “job performance” concerns both process and outcome (Davis, 1993, p. 477). Perceived ease of use is defined as “the degree to which an individual believes that us- ing a particular system would be free of physical and mental effort” (Davis, 1993, p. 477). Perceived ease of use has a direct effect on perceived usefulness, but not vice versa. A system can be perceived as more useful either by adding new functionalities or facilitating the functionalities that already exist (Davis, 1993). Mongogole and Jokonya (2018) argue that the attitude towards using is additionally affected by the organizational culture. Organizational culture is described as “the morals, values, views, beliefs and unseen assumptions that staff publicly share in the organization.” (Mongogole and Jokonya, 2018, p. 839). 18 3 | Method This chapter aims to describe the research strategy. The main method to research the topic was a single case study encompassing a collaboration between technical consultancy company Sweco and a real estate company. First, an introduction to the chosen strategy will be presented. This section includes three subsections that describe how literature was reviewed, what the case study encompassed, and how data analysis was conducted. The last sections include a critical evaluation of the chosen method and an ethical statement. 3.1 Research Strategy This thesis research strategy is based on a qualitative approach. It is an open process where the material collection and analysis are parallel procedures (Bell et al., 2019). It gives the research conditions to reach an in-depth understanding of the phenomena since the process becomes iterative and investigators can return to empirical material with new ideas. Moreover, qualitative research methods are preferable when conducting a case study since it allows unstructured interviewing which are helpful when examining a case (Bell et al., 2019). The research began with performing an extensive literature review parallel to initial discussions with consultants at Sweco. The research questions were defined along with strategies for retrieving empirical data. The literature review research pro- vided an understanding of the context by constructing a foundation from previous research. It created a theoretical framework that was used for analyzing the case study. The case study included assessing material provided by Sweco and conducting ten qualitative interviews with relevant actors from both companies. The empirical data was later categorized and analyzed in relevance for its research question. The arrangement for the individual elements of the research will be further explained in forthcoming subchapters. 19 3. Method Figure 3.1: Research process (authors own figure). 3.1.1 Literature Research A literature review is an essential element when conducting a research (Bell et al., 2019). It helps to identify what is already known about the research topic and previously applied theories. This literature review was conducted accordingly to the narrative approach. This approach focuses on quality rather than quantity by limiting the research to the most interesting contributions. The main focus is on the connections between found sources and research questions, allowing theories to emerge from the search and subsequently revise the research questions (Bell et al., 2019). The collection of research was mainly done through the literature databases Chalmers Library, Google Scholar, and Scopus. Keywords such as digital twin, BIM, digital transformation, real estate, and facilities management were used in different combi- nations to collect relevant literature. Since digital twins within real estate still is a new topic with limited conducted research, the literature was combined with reports from companies developing digital twins. Furthermore, an important realization was that the BIM concept had developed fur- ther than the static model. Many researchers use the concept synonymously to the ideas of the digital twin, e.g., “BIM and IoT integration with real-time data” (Cheng et al., 2020), which makes such, but not all, literature relevant for this thesis. This was a necessary realization to broaden the scope of relevant articles. The literature research resulted in the theoretical framework that explains the key theoretical concepts in real estate and facilities management, information manage- ment, digital twins, and change management of technologies. The theoretical frame- work was used as a reference for evaluating the case and supported the analysis of the empirical data. 3.1.2 Case Study The selection of case was made according to an intrinsic case study. In an intrinsic case study, the case is chosen based on interest by the researchers and conducted to 20 3. Method achieve a better understanding of this particular case (Stake, 2000). The case study consisted of two companies, one consultancy company Sweco, which served the role of product and service provider, and a real estate company. The real estate company is transforming its data infrastructure, and that is where digital twins have emerged. As a part of their IT strategy, Sweco has developed technolo- gies called Twinfinity and SimpleBuilding to enable the digitalization of buildings. In 2020, Twinfinity was implemented as pilot projects in a few different facilities owned by the case company and situated in different geographical locations in Swe- den. In 2021, the technology is planned to be implemented in 30-40% of the real estate portfolio. The empirical data consisted of both materials describing the technologies used, and ten semi-structured interviews held online. The interviewees were employees involved in the case from both companies and with different professions. Semi- structured interviews were chosen as the method for conducting the interviews since it allows flexibility. It was beneficial as the interviewees had various insights into the case. Semi-structured interviews with open-ended questions are one of the main strengths of qualitative research (Mohajan, 2018). It creates the perquisites for ob- taining new or unanticipated information (Mohajan, 2018). The interview candidates were chosen in snowball sampling, where the researcher makes contact with a small group of people relevant to the research topic. This sampling helps to establish contact with other appropriate interviewees (Bell et al., 2019). In this case, the initial contact and interviewee was Consultant 1 (C1), who introduced relevant participants from both Sweco and the case company. Those participants aided in the further sampling of interviewees. The ten chosen interviewees had various backgrounds and roles in the studied case. The interviews aimed to gain knowledge and opinions from employees within all the identified key process areas: facilities planning, project and transaction man- agement, and service, operation, and maintenance management. Since the studied case was in the initiation phase, the number of people with knowledge was limited. However, the intended aspiration of interviewing employees within each of the key process areas was obtained by interviewing three employees who were not directly linked in the case but involved within the company. Two of these interviewees had prior knowledge of digital twins, but not precisely the solutions of this specific case. The interviewees and their role descriptions can be seen in Table 3.1. Henceforth their names will be shortened to C1 for Consultant 1 and R1 for Real estate repre- sentative 1. The three interviewees who were not directly involved in the case will be represented by a letter instead of a number, for example, CA. 21 3. Method Table 3.1: Overview of interviewees. Interviewee Short Role Description Consultant 1 C1 Service developer, involved in the case on several levels Consultant 2 C2 System developer, involved in the case on several levels Consultant 3 C3 Business developer, involved in the case on several levels Consultant A CA BIM strategist, not involved in the case Consultant B CB BIM strategist, not involved in the case but familiar with other cases of digital twins in real estate and FM Real estate representative 1 R1 IT manager, strategically involved in the case Real estate representative 2 R2 Facility development, projects, strategi- cally involved in the case Real estate representative 3 R3 Facility development, technical, tactically involved in the case Real estate representative 4 R4 Research and development, strategically in- volved in the case Real estate representative A RA Facility manager of several facilities, not in- volved in the case 3.1.3 Data Analysis Qualitative content analysis was used to interpret meaning from the content of interview transcriptions. Such analysis is defined as “Qualitative content analysis is defined as a research method for the subjective interpretation of the content of text data through the systematic classification process of coding and identifying themes or patterns.” (Hsieh and Shannon, 2005, pp. 1278). The process entails going through the text data and highlight words based on categories that organize the information in meaningful clusters (Hsieh and Shannon, 2005). This thesis used a conventional content analysis approach to determine the categories and subcategories for the coding process. With the conventional approach, the authors identify patterns and key concepts after collecting and getting a sense of all the data. Then forming categories based on the emerging themes (Hsieh and Shannon, 2005). By analyzing both the interviews and the provided materials, six categories emerged. The first three related to the technologies and methodologies of the case study. These formed the basis of chapter 4 case study. The latter three related to: digital transformation in real estate, usage areas of digital twins, and challenges of digital twins. These formed the chapter 5 interview study. Lastly, the conventional content analysis was used once more to divide the data into subcategories. 22 3. Method 3.2 Critical Evaluation of Method When conducting qualitative research, trustworthiness is an essential criterion for assessing the quality of the research (Bell et al., 2019). Trustworthiness can be di- vided into credibility, transferability, dependability, and confirmability. Credibility refers to how believable the findings are. Transferability address how well the find- ings can be applied to other contexts. Dependability tackle how well the findings can be applied later. Lastly, confirmability refers to whether the researcher’s own beliefs and personal values have influenced the study (Bell et al., 2019). To create credibility, this thesis used respondent validation by providing the re- search participants with the constructed result to gain confirmation and feedback of the conducted data interpretation (Bell et al., 2019). Moreover, to strengthen the conclusions, the results were compared to findings from the literature according to the triangulation methodology (Bell et al., 2019). Transferability is sometimes viewed as troublesome for a single case study. The concept of digital twins in real estate is an uncharted topic, and only a few companies use digital twins – as well, the approaches vary. This thesis, therefore, chose one case to answer the research questions and, therefore, provides perspectives from one approach. Another method would have been to examine a “cross-section” of the industry to display several approaches and perspectives. The consequences of this thesis method are that the results can be too one-sided to fit the whole industry. Flyvbjerg (2006) however treat several misunderstanding concerning case studies in her research, among others the misunderstanding “that one cannot generalize on the basis of a single case and that the case study cannot contribute to scientific de- velopment” (Flyvbjerg, 2006, p. 12). Flyvbjerg (2006) emphasizes that one can often generalize the results gained from a single case study. However, one should not underestimate ‘the force of example.’ Even if the knowledge gained from the case might not be generalized, it can still contribute to the process of knowledge gathering in the specific field (Flyvbjerg, 2006). The following reflections regarding transferability can be made related to this thesis: The digital twin studied in this case might not symbolize all digital twins; however, the knowledge gained can con- tribute to the understanding of digital twins within real estate. Likewise, might the case not generalize a unified strategy regarding the implementation of digital twins in all real estate companies. However, the knowledge gained from the example can contribute to understanding of the implementation. Moreover, Bell et al. (2019) argue that dependability can be gained by the auditing approach consisting of complete records of all phases of the research and peer re- view. This study conducted interview transcripts and provided documents and data analysis decisions but was not easily accessible and analyzed by the peer reviewer. Bell et al. (2019) identify that qualitative studies can generate a large amount of data that could affect the suitability of an auditing approach. This was relevant as the material was perceived as to extensive for a peer group to manage. However, to strengthen the dependability, the material and data analysis were discussed with 23 3. Method peer reviewers during auditing of the draft result close to the end of the research. This auditing also helped to strengthen the confirmability. Discussion with both the peer review and examiner was held to evaluate the data analytics to ensure that open-minded analytics was conducted with minimal influence on the researchers’ personal values. 3.3 Ethical Statement When conducting a study within social science, the authors must be aware of the ethical principles involved to avoid harm to participants and unethical activities (Bell et al., 2019). According to Bell et al. (2019), ethical consideration should be a vital part of the research process and continuously revised through the study. The ethical consideration discussed in this section is based on the main ethical codes in business research stated by Bell et al. (2019). This study included interviews with several participants from two different compa- nies. The participation was voluntary to avoid harm to the participants and achieve consent. The participants received information about the study beforehand in or- der to make a well-informed decision relating to whether they wish to participate or not. Considering the interviews recorded, the authors asked for permission and consent beforehand. One representative from each company were presented with the result from the interviews to review before the study was published. This was done to avoid misrepresentation and harm or stress regarding accidentally omitting confidential information. Moreover, the participants were anonymized, including only the professional title and name of the organization to protect privacy. Lastly, to prevent deception, one representative for the organizations was involved in the planning and execution of the research. There was a continuous dialogue between the authors and this repre- sentative when changes occurred in the research process. A final statement of this thesis is that there is no intent to advertise any products but to reflect on the knowledge that practice possesses. 24 4 | Results: Case Study The empirical result is derived from a case study where digital twins are developed in a commercial real estate company. The results are divided into chapters 4 and 5. This chapter presents the case background, followed by SimpleBuilding and Twinfinity that are the technologies used to digitalize buildings and last, the case company’s approach to digital twins. 4.1 Background The case was provided by Sweco, which is a leading technical consultancy com- pany in Sweden within engineering and architecture (Sweco, 2021). The related department is Sweco Position which specializes in IT solutions within the built en- vironment. The case study comprises Sweco and their partnership with the case company, which is one of the largest commercial real estate owners in Sweden. Sweco and the case company have a far-reaching collaboration in developing the case company’s digital information management. During their collaboration, they have long acted as partners and jointly developed solutions in line with their digital trans- formation. The collaboration initially concerned digital information management in projects but has in recent years approached facilities management. A few years ago, the product Twinfinity emerged. Simultaneously the collaboration changed to Sweco becoming a product owner, and the case company became a customer. However, there are still many consultants that work internally in the case company. The case company is an active leader in digital innovation. In recent years, they have focused on contributing to an industry development that enables real estate owners to become more successful and benefit more from their platforms. In this way, they have been involved in developing RealEstateCore and ProptechOS. It is in that journey where the interest in digital twins has been established. They have chosen to collaborate with Sweco to create “collaborative digital twins” by using an “industrial approach”. This is further explained in 4.4. 4.2 SimpleBuilding: Creating Spatial Models To create digital representations of existing buildings, Sweco has developed a con- cept called SimpleBuilding. It is a methodology and technology for buildings that 25 4. Results: Case Study either lack or have poor basis. SimpleBuilding encompasses modeling in two Levels of Detail, named SimpleBuilding and SimpleBuildingPlus. • SimpleBuilding: a space model with the simplest level of detail including only rooms. • SimpleBuildingPlus: an extended space model with higher level of detail in- cluding objects, e.g. walls, floor, doors, windows, roof and staircases. SimpleBuilding is the simple volume model for a building with a low level of detail, where the base only consists of the building’s rooms (Figure 4.1). No other building envelope or installations are specified. Figure 4.1: SimpleBuilding (by Sweco). SimpleBuildingPlus is a simple architectural model that includes the building enve- lope, mezzanine floor, indoor walls, and doors (Figure 4.2). It extends the Simple- Building by including a larger part of the building and has a higher level of detail. The modeling method enables visualization opportunities and orientation inside the building. The SimpleBuildingPlus is a greater step towards realizing the digital twin as it allows more combinations of data. Figure 4.2: SimpleBuildingPlus (by Sweco). 26 4. Results: Case Study The methodology of creating SimpleBuildings is seen in Figure 4.3. The process of constructing a model begins with performing an inventory analysis of existing mate- rial and data. Besides drawings, pdfs, and dwgs, other information of interest is, for example, room numbers and rental objects. Acquired 2D drawings are converted to a 3D model by an automized process. If no drawings exist or are of too poor quality, an inventory of the building is performed. This is either done with laser scanning, drone scanning, or manual measurement. Scanning generally creates a PointCloud in 3D. That is a geometrical and geographical correct representation of the as-is model. The spaces, i.e., those of interest of a real estate owner, are allocated. The result is a BIM model. Figure 4.3: Sweco’s process flow of creating SimpleBuildings (based on provided mate- rial from Sweco). 4.3 Twinfinity: Contextualizing Digital Twins Twinfinity is a platform that includes processes for adding one aspect of the digital twin, the digital model. Hence, it is not a digital twin itself. It should rather be seen as an essential enabler of creating and contextualizing a digital twin. Twin- finity prepares building data to be integrated with other systems or applications. The aim of Twinfinity is to create value within the real estate and facilities man- agement processes by combining several different data sources in building, business and operations. An overview of Twinfinity’s processes is shown in Figure 4.4. Figure 4.4: Twinfinity’s extraction engine (based on provided material from Sweco). The input data to Twinfinity are BIM-models, PointClouds, 360-photos, DWG, Raster, and/or PDF. Twinfinity has automized processes for transforming the for- mats into BIM models in IFC format. Areas that are of interest for real estate owners 27 4. Results: Case Study are allocated, such as LOA (premises area of facility), ATEMP (heated area), and BOA (area of household). The file, and the quality of the model, are validated in an automized process according to pre-set design guidelines to assure the model has the expected object properties. Twinfinity is then explained as an extraction engine that extracts the BIM model into objects organized by element type (see Figure 4.4). The element type can, for example, be walls, rooms, doors, windows, text information, toilets, and sinks. The objects are then stored within a database, and other systems can collect the data by Twinfinity’s API. Twinfinity has a tool for sending change requests to the IFC-file without going through a computer program. It is called provisioning and allows users to directly add or alter points of interest, e.g., assets, furniture, sensors, or spaces into the dig- ital model. The users can add the points of interest while using the viewer function in a web browser or mobile devices. Furthermore, Twinfinity includes an advanced 3D motor and viewer application that enables visualization of the spatial model and the linked data in webpages or mobile app (see Figure 4.5). This tool makes it possible for the user to digitally navigate in the digital building, click on spatial objects to visualize the linked information. It is also possible to apply a visualization setting, which determines what information should be visible and not. By using Twinfinity Embedded, it is possible to build Twinfinity’s viewer into other systems and applications. Figure 4.5: Twinfinity’s viewer (by Sweco). Twinfinity itself does not include any applications. Instead, applications are devel- oped by the customer. By combining the building data with business and operational data, there are endless possibilities are to develop tailor-made services. Potential data consumers are, e.g., mobile apps, simulations, calculations, intranet, viewers, and virtual/augmented reality. 28 4. Results: Case Study 4.4 Collaborative Digital Twins: An Approach This section describes how the case company is approaching digital twins. The first section presents their strategy of information technologies, and the second presents the digital aim of twinning their portfolio. 4.4.1 IT Strategy The case company has not chosen to create digital twins directly but rather connect building, operational, and business data. The case company has described its ap- proach as industrial, to create collaborative digital twins. It implies that they are not purchasing a packaged digital twin product but instead focus on benefit more from their existing systems and platforms. A strategy has been developed on how to use information technologies. It consists of four steps where data flows from heterogeneous sources to the tenant, owner, or third party applications (see Figure 4.6). The steps are to (1) make data usable using RealEstateCore, (2) make data accessible using ProptechOS, (3) secure access to resources and data using Accessy, and (4) to create value for users using multiple applications, e.g., tmpl. Figure 4.6: Case company’s strategy of information technologies (based on provided material by the case company). RealEstateCore is an ontology, i.e., language, to enable communication between var- ious data sources in buildings. It is not a new standard but instead merges and bridges existing standards to find the common denominators. It is a prerequisite to enable data integration in smart buildings and prepare them to interact in smart cities. The ontology is the first of its kind developed by real estate companies for their needs (Hammar et al., 2019). It is modular, to avoid over-commitment and enable customization, covering building structures, ownership, inhabitants, techni- cal systems, and sensors. Hammar et al. (2019) explains the three domains that are 29 4. Results: Case Study bridged in RealEstateCore: • Digital representation of the building’s and their constituent elements • Control and operation of the building and its systems • Emerging IoT technologies Twinfinity uses the RealEstateCore standard to export the building elements accord- ing to the building’s knowledge graph, i.e., the structure of how different building components relate to each other. RealEstateCore divides the building structure do- main into components, for example, rooms and roofs, under the name BuildingStruc- tureComponent. These components can then be joined under different premises. Premise is a collection of spaces and objects that are leased to a tenant. Devices within the real estate are linked to the BuildingStructrureComponent they belong to. Devices are defined as a piece of electronic equipment made for a particular purpose. Every device has one or more sensor(s) or actuator(s). Sensors are things that measures or detects, while actuators are components that controls or moves a system. ProptechOS is a commercial product that is available on the market. It is a Building Operating System for importing and exporting data that is connected to buildings. It functions like a data lake, where all data is collected. It enables various data to integrate and applications to import information. Accessy is a system to manage access to virtual and physical resources. As pre- viously stated, data can be published to any stakeholder through APIs, internally with themselves or externally with the academy, colleagues in the industry, or cus- tomers. To enable or limit the sharing of data, the case company uses Accessy which manages access to resources in the digital or physical component. It can be used to manage which doors a person can access or access specific data. Applications are systems that consume data to generate services. As mentioned in 4.3, combining the business, operation, and building data open several usage ar- eas. The applications could be an interface for tenants, themselves, or third parties. Tmpl and Flowscape were two applications mentioned in the case study. Some ex- amples are communication tools between tenants and owners to report problems, or interfaces for tenants to monitor their energy consumption or receive other relevant information, or used as a booking system (Tmpl Solutions AB, 2021; Flowscape So- lutions, n.d.). Other applications could be analytical tools to optimize the actuators in systems or provide reports with results. 4.4.2 Digital Aim The case company has an aim to digitalize its entire real estate portfolio. The purpose is to lift the entire existing portfolio to a connectable level that enables pro- visioning of, for example, sensors and objects, create facilities management benefits linked to the model and contribute to a smarter business. By the end of 2021, the 30 4. Results: Case Study ambition is that 30-40% of the portfolio is connected. The SimpleBuilding methodology is used to create digital base models by relying on existing basis. However, the quality of the existing basis differs between facilities. The company has defined five levels that describe the model’s trustworthiness and related capabilities. Each facility aims to be refined and climb upwards through its life cycle, regardless of what level it starts. By Level 1 the case company can linkage technical documentation to model objects and connect the model to ProptechOS. As ProptechOS require 3D models, the first level starts at SimpleBuildingPlus. • Level 1: Spatial model without any significant evaluation based on its accu- racy. Tenant divisions and floor plans are correct to the extent of number of rooms but not area. • Level 2: Ensures the quality of the existing dimensions and provides more credible areas. However, the dimensions are not checked or inventoried with laser scanning or manual measurement. • Level 3: Correct architectural model provided by laser scanning methods. Enables 360 photos and includes correct areas. This level enables energy dec- larations. • Level 4: Provides a holistic view of the building. 3D models of installations and provisioned objects are quality ensured. The level enables tracing of sys- tems and their service areas, as well a broader data supply to ProptechOS. • Level 5: A full-scale BIM-models for every technical area. The level is as- sured by BIM-models provided from commissioning of new constructions or extensive reconstructions. The level completely enables model-based facilities management. 31 5 | Results: Case Interviews The following chapter presents the results from the interview study regarding the digital transformation in real estate, followed by usage areas and challenges of digital twins. 5.1 Digital Transformation in Real Estate This first subchapter presents the result revolving drivers for the digital transforma- tion and the current issues within digitalized information management. Last, the role of the digital twin within real estate digitalization will be presented. 5.1.1 Investment objectives According to C2 and R1, a significant driving force for digitization in real estate is a transformation in business. The real estate business logic is changing from seeing leasing a facility as simply access to a physical space towards a packaged service with several offerings and flexible lease agreements. The previous long leases of three or even 15 years put much risk on the tenant. R1 describes that the customer demand for long lease agreements based on square meters is decreasing. Instead, the customers want flexibility and combine several different offers. Softer values such as productivity, health, and well-being are as well increasing in interest. However, C2 adds that expanded service business increases the importance of digital facility information. “The customer would like to have the ability to expand and decrease [the lease], combine with different types of offers. Not a traditional office of ten thousand m2, maybe two to three thousand is enough combined with a “smart and ready” short leases and some subscriptions of co-working arena. The ability to have a hybrid solution. This demand would con- sequently in combination to our goal of delivering more services as an addition to our products, result in that we need to have greater control of our buildings, in a different way than before” - Real Estate Representative 1 R2 further explains that another focus area is making their business sustainable. R1 says that their goal is to be climate-neutral by 2030. Lowering the environmental 32 5. Results: Case Interviews impact and, in a way, “save the world,” as described by R1, is an important driver, and it is everyone’s responsibility to make their contribution. R3 says that they must find new ways to optimize their operations systems to save money and reduce the carbon dioxide load. The knowledge gained from effective data collection and analysis enables wise decisions on optimizing operations and reducing the environ- mental impact further. Moreover, R1 explains that the transaction manager has mentioned that they are willing to pay more for a facility with a digital representation. Previously, the aspect of digitalization and digital information of a facility was not considered in transac- tions. R1 says that a digital representation and well-documented information will reduce uncertainties in transactions. Risk is costly, and as it is reduced by having a digital representation in place - the willingness-to-pay increases. In the same man- ner, this would increase the value of their facilities, R4 concludes. Summarizing, R1 concluded that the case company’s drivers for digitalization are quite simple - their goals are to achieve a high return to the funds that own the company. R4 adds that they would, of course, not start a project if they only saw it as cost-driving. “For our part, when we invest in new systems, it must either: reduce our operating costs so that we get lower expenses or make the houses better so that we get higher rents and better income.” - Real Estate Representative 4 5.1.2 Current Issues in Facilities When asked about the current issues faced in real estate and facilities management, the interviewees expressed a common theme that there is a significant need for a more efficient way of collecting, updating, organizing, and analyzing data. It is an essential part of decision-making processes and can optimize the processes and gain new insights, thereby reducing costs or increasing income. Moreover, C1 stated that the unstructured management of documents results in unnecessary re-work. R4 added that it is easy to see the cost of this digitalization change but that the mon- etary benefits are not always clear. However, C3 and CB stated that an investment in the organization of data would generate benefits. “If we can help our customers [the case company] to have better organi- zation and management as well as better control of their numbers, data, and business, then they will generate more money from this.” - Consultant 3 R4 elaborates by explaining how the case company is a matrix organization where they work much in downpipes, but they also collaborate cross-sectionally. Real 33 5. Results: Case Interviews estate and facilities management processes are highly dependent on information. Keeping this information up to date and relevant has for a long time been the biggest challenge in the whole real estate industry. R4 also said that they are af- fected by an increasingly extensive bureaucracy that demands authority controls and inspections. The inspections and regulatory controls a facility needs to fulfill put a high demand on organized facility information. He addresses that anyone in the company needs to find information quickly. It is then vital to store data in a way that makes it accessible for anyone. “It is really important that we have systems as well as structure and pro- cesses that says; this is how we store data in [the case company]. Then you will know where to find it, for example, inspection material from an elevator, no matter where you work at [the case company]” - Real Estate Representative 4 Two interviewees mentioned that collecting operational information has not been the problem. The operational systems in the buildings are today relatively highly automized, according to R4. The systems have for a long time been able to show temperatures, ventilation data, or provide an operations overview. Though, the technical development is still relatively static and traditional. The problem is that each system has its own fabricate, and those cannot communicate or coordinate with each other. As a result, building automation systems may regulate contra productively, e.g., ventilation and heating. Therefore, real estate owners struggle to overview and controlling their facilities’ operations effectively. Moreover, information such as drawings, tenant information, and operations data are organized and stored in different databases and systems. Historically, R3 explains that much information has been stored in physical office binders, and then the infor- mation and knowledge has been isolated in the related department. Even though the case company is not currently using physical binders anymore, the problem of iso- lated information remains. This has made it hard to share the information between different departments or concerned stakeholders. The interviewees also state that it is hard to keep the scattered information stored updated since locating and updating all the information is time-consuming and error-prone since it is easy to miss one file. Lastly, two interviewees addressed that simply accessing the information is not enough to make well-informed decisions. Instead, the information needs to be pre- sented in its context to enable a better understanding of the presented information. BIM plays a big part in contextualizing buildings’ design and construction phase, but R4 explains that the concept is not used in the operational phase. Additionally, the current tools for conducting analyses are not optimal. Some employees have created excel matrixes to conduct relevant analyses, but it is not an ideal solution R4 concludes. 34 5. Results: Case Interviews 5.1.3 Role of Digital Twins The majority of the interviewees stated that defining a digital twin is complex and that the definition varies a lot within industries. However, the interviewees’ answers when asked to define a digital twin and its role in the digital transformation were quite similar. “[...] it should be a digital representation of the facility or building you are taking care of and it should be detailed enough to fulfill your needs. I do not think it needs to be at the screw and nut level but [detailed] enough. There are different views on it, ‘that you shall have all data’. I do not think so, but rather that you should be able to link it to other types of data - or visualize the data together, one might say.” - Consultant B All but one of the interviewees stated, in different ways, that a 3D representation of a facility does not fulfill the characteristics of a digital twin. CA, who primar- ily works with models used for the design and construction process of buildings or project, describes a digital twin as a 3D BIM model which includes all information created during the construction phase in an easily accessible and structured way. The model should remove the need for separate systems and documents. Moreover, several of the interviewees stated that to fulfill the status of a digital twin, it needs to be able to visualize a 3D representation of the building and integrate additional data to this representation. What this additional data needs to contain was described in different ways by the interviewees. However, all concluded that it should represent the reality, that the business and building should be visualized in its context. Five interviewees specifically concluded that the additional informa- tion should contain building, business, and real-time data collected from the facility. When that is realized, R1 thinks that they have come quite far towards a digital twin. However, R1 addresses that it is difficult to talk about a digital twin because it is very hyped and can mean different things to different people but realizes that their approach resembles a digital twin. 35 5. Results: Case Interviews “The digital twin is a representation of all of the facility data, all the data flows within the facility. [...] Basically, everything that happens inside a facility that could be easily followed up trough data communica- tion” - Real Estate Representative 3 Four of the interviewees included statements of what role or purposes the digital twin needs to fulfill to be defined as a digital twin. One interviewee stated that it is not until the model is combined with relevant information to facilitate the intended purposes, it will fulfill the status of a digital twin. Two interviewees stated that the information provided by the digital twin needs to be able to facilitate analyses of the current state of the building and the fourth interviewee stated that the information provided by the digital twin should facilitate change processes. “For us at [case company] it is when we connect the real time data from a facility along with the spatial information and the business informa- tion. When we connect these three parts and has a process where we can handle changes in a life-cycle perspective, then we have come a long way towards a digital twin in my point of view” - Real Estate Representative 1 R2 adds that the data is the actual interesting aspect of the digital twin and that the base model is only one way to represent and visualize data. The digital twin should mirror the whole business. In some cases, a building model as a visualization tool is beneficial, but for complex analysis other methods or tools to visualize the data might be preferable. “The basic models are in a way a bit fragmented since you look at building for building and room for room [...] but if you look at complex relations then maybe there are other types of analyzes you need to use, diagrams or other methods which may better represent data.” - Real Estate Representative 2 5.2 Usage Areas of Digital Twins in Real Estate This chapter will describe intended usages and benefits the interviewees have de- scribed categorized under its key process areas. Some of the applications have been tested during pilot projects. However, since the digital twin is not fully incorporated in the organization yet, many of the discussions covered future usage of the digital twin. 36 5. Results: Case Interviews 5.2.1 Facilities Planning R2 explains that the project and transaction department are working collaboratively in facilities planning. The transaction department conducts investment analysis for the real estate portfolio yearly. A process that includes examining each facility by considering its economic condition, potential, and future plans. Several interviewees stated that historical and real-time data of the business, building, and operation would help to make wise decisions regarding facilities planning. By accessing digi- tal facility information, like historical data about reconstructions, they will have a greater basis for deciding on a facility’s future. CB stated that having control over updated building information and the different capacities of the facility will make the company be able to make decisions regarding if it is worth initiating a reconstruc- tion. For example, what is the facility’s capacity at the moment, what is needed to increase that capacity, is there room to expand the electricity systems, or how costly would the tenant adjustment. In the project department, the facilities’ capacity for new tenants is examined by considering construction, floor heights, ventilation, and other systems, maintenance status, and legal aspects in detail planning. 5.2.2 Project and Transaction Management R1 and R4 stated that a facility with a digital twin could increase the value in transactions. First, a smart building itself is attractive. R1 mentioned that envi- ronmental certifications are getting more common and can potentially increase their buildings’ value, a process the digital twin can facilitate. Moreover, R1 implied that he would not be surprised if smart building certifications arise in the near future and that these could increase the value of a building in transactions. However, many certifications put high demand on control of the indoor climate and knowledge of the facility, which continuously needs to be validated to keep the certification. Secondly, R1 explains that the transaction manager has mentioned that they are willing to pay more for a facility with a digital representation. Previously, the aspect of dig- italization and digital information of a facility was not considered in transactions. R1 says that a digital representation and well-documented information will reduce uncertainties in transactions. Risk is costly, and as it is reduced by having a digital representation in place - the willingness-to-pay increases. In the same manner, this would increase the value of their facilities, R4 concludes. Having control of what a building contains is of great value while selling a building. During a single transac- tion, the real estate employees can use the information gained by the digital twin to communicate the value of a building – then the buyer knows what the price includes. “In facility transactions, a building with bad digital representation and bad technology density will be valued less.” - Real Estate Representative 4 Regarding projects, several interviewees find many user perspectives. The model composing the digital twin can be used as a foundation to start a new project. R2 explains that the time schedules are short once a project start, and then an organized 37 5. Results: Case Interviews basis is crucial. CA explain that during the initiating design phase of reconstruc- tions, the architects need access to correct volumes to start planning. They need to retrieve dimensions in a room, e.g., where the original and suspension ceiling is situated. For an extensive reconstruction, the contractors need millimeter accuracy of the building information. However, in minor reconstructions or changes in a fa- cility, CA believes that it may be sufficient to revise the existing basis. Moreover, CA declares that creating a BIM model for reconstruction can be very difficult. Older facilities contain many surprises that are not documented. The contractors need millimeter accuracy on the model, and therefore a laser scanning is beneficial. However, laser scanning of facilities in use comes with several barriers. Suspension ceilings, for example, obstruct the ability to read hidden elements. Additional laser scanning can be necessary as the project progress to make up for initially hidden parts. During their last reconstruction project CA was involved in, the initiating process of scanning and completing a building model fit for facilitating the design phase took six months and half a million SEK to complete. Furthermore, CA adds that 360 photos should not be underestimated; they have been valuable during project design. This information can be structured, saved, and provided by a digital twin platform. A digital Point Cloud with complementary 360 photos enables consultants to digitally inspect the building and consequently save time because they do not have to visit the site.