Building Information Modelling Facility Management (BIM-FM)

Abstract

Defined digital Facilities’ Management (FM) systems will contribute to the realisation of the United Nations’ Sustainable Development Goal (SDG) 11. Of the available digital FM systems, Building Information Modelling (BIM) for FM, herein referred to as BIM-FM, is the least developed. Where BIM-FM varies from existing digital FM tools is its advanced 3D visualisation capabilities. A semi-structured literature review is undertaken to assess the current implementation of BIM-FM and identify opportunities to engender its increased adoption. This paper is part of an ongoing piece of research aimed at defining a standard methodology for the application of BIM to historically significant structures, otherwise known as Historic BIM (HBIM). Two existing approaches to BIM-FM, current and developing, are outlined. The potential value BIM-FM can provide according to the literature is discussed but there exists minimal practical evidence to justify these claims. Barriers to its adoption are discussed, with a key underlying barrier found to be a lack of defined user requirements. Consequently, functional, modelling and information requirements established within the literature are identified, and existing attempts at realising the requirements are discussed. Six information categories and two functional requirements are identified. It is theorised that the tendency to utilise simplified geometric models for FM is primarily due to software and practical limitations as opposed to actual end user needs, and it is suggested that this should be investigated further in future work. Attempts at realising BIM-FM user requirements using other advanced technologies, primarily Digital Twins, are investigated and found to be an area of increasing commonality. A new conception of BIM-FM is proposed.

1. Introduction

1.1. Digital Facilities’ Management

Facilities’ Management (FM) is defined as an “organisational function which integrates people, place, and process within the built environment with the purpose of improving the quality of life of people and the productivity of the core business” [1]. It typically involves the management of the day-to-day use and ongoing maintenance of an asset. Asset Management (AM) is a term frequently used interchangeably with FM, which involves “coordinated action of an organisation to realise value from assets” [2]. The key difference between FM and AM is that AM is focused on higher level organisational objectives and the realisation of value, whereas FM focuses more on asset functionality for immediate use. AM and FM have many common features and principles; so, whilst FM will be the primary topic of discussion, AM resources will not be excluded from this review.

Effective FM will help contribute to the United Nations’ (UN) Sustainable Development Goals (SDGs), most notably, SDG 11: sustainable cities and communities [3]. As such, the need for defined FM systems, in line with organisational requirements, is internationally recognised [4]. One approach has involved the creation of digital FM systems, which are now used extensively. Popular systems for existing buildings can typically be categorised as Asset Information Models (AIMs) [5,6], Computer-Aided Facilities’ Management (CAFM) [7,8,9] and Computerised Maintenance Management Systems (CMMSs) [10,11,12,13]. Often, these systems will be spreadsheet-based. However, sometimes, systems are integrated with Geographic Information Systems (GISs) [8,14,15], which depict a wide overview of a site with more focus on the location than the details of the asset itself. This integration is championed by Moretti et al. [16], who justified it by stating “management [takes] place at different scales depending on the scope of the specific task”.

A less developed approach to digital FM is Building Information Modelling (BIM). BIM is a 3D modelling and information management process, which typically involves the creation of parametric model objects with additional information attributed to the objects at higher dimensions, e.g., 4D involves the inclusion of time data [17]. A key benefit of BIM is its 3D visualisation capabilities, a capability not normally native to other digital FM techniques [18]. The majority of BIM development has been focused on the use of BIM for the construction stage of an asset lifecycle with minimal consideration of the use of BIM for FM, herein referred to as BIM-FM. BIM is becoming increasingly prevalent in construction. However, existing BIM and digital FM tools have developed separately to each other, and as such, there is often very little integration and understanding between the two disciplines [11]. Both the ISO standard for BIM (ISO 19650 [19]) and BIM terminology reflect this, with ISO 19650 not fully considering FM and there not being an agreed dimension within BIM referring to FM [17]. Thus, if no integration is achieved, the FM sector may increasingly suffer from a lack of relevant building information. as suggested by Quinn et al. [20], who said “the move toward digitized building information has removed some access to building information from FM”.

There have been some attempts to transfer BIM models made for construction into an FM-ready state [21], but these are often isolated, local solutions. Regardless of their success level, this fails to acknowledge that BIM is a relatively new occurrence and many existing assets do not have BIM models, such as in the work by Marmo et al. [22]. This paper will therefore investigate BIM-FM for existing assets under the assumption that, for the majority of cases, no BIM model exists.

1.2. Aim

There have been a number of reviews in the field of BIM-FM previously, with Volk et al. [23] being, arguably, the most notable and one of the only reviews limited to the study of existing buildings. Other reviews have limited their evaluation to specific sectors, such as Pinti et al. [24], who investigated the use of BIM-FM in the public sector, or Lu et al. and Mannino et al. [25,26], who have reviewed the use of additional technologies alongside BIM, such as sensor integration. All areas were found to be in the early stages of development. Barriers to BIM-FM development have been identified in previous reviews [21,23]. However, in previous work, there appeared to be some disparity between perceived issues and the actual views of FM practitioners [21].

Therefore, the aim of this paper is to establish the current implementation of BIM-FM to existing assets via a semi-structured literature review. In so doing, it should complement the existing reviews in this area by providing an updated overview of BIM-FM for the existing assets in all sectors as well as providing an overview of the existing barriers incorporating recent developments.

The objectives are as follows:

• To identify the literature regarding BIM and FM;
• To review the literature and evaluate the current implementation of BIM-FM;
• To identify barriers effecting BIM-FM implementation;
• To identify opportunities to increase BIM-FM implementation.

This paper is part of an ongoing piece of research [27] aimed at defining a standard methodology for the application of BIM to historically significant structures, otherwise known as Historic BIM (HBIM). HBIM assets are by nature of their existence pre-existing assets, so they fit within the scope of this paper. However, this review is not limited to historic assets, as there is a greater degree of research into the field of BIM-FM for more recently constructed existing structures and many reviews in the subject area actively omit historical assets [23].

2. Methodology

2.1. Literature Identification

The Scopus database was used to search for the literature. The key word search (“BIM” OR “HBIM”) AND (“Asset management” OR “Facilities Management”) was used to search for papers. HBIM was included for the completeness of data but only six of the reviewed papers were on the topic of HBIM. The literature search excluded papers looking to integrate BIM models into alternative modelling formats (where the alternative modelling format acts as the primary source of data). This primarily involved GIS. The reason for this was that since different modelling formats have different exchange requirements, the effort and resources required to create a BIM model and then transfer it to an alternative format would be formidable. The exceptions were papers looking to integrate alternative modelling formats where BIM acts as the primary source of data. For instance, utilising BIM tools and modelling but incorporating elements of processes or tools utilised by a different format (e.g., incorporating GIS datasets into a BIM authoring tool).

The Industry Foundation Class (IFC), an ISO standardised data exchange format (ISO 196739-1:2018 [28]), is used to transfer data in a BIM-ready format. Whilst it enables data transfer, its current scope is a limiting factor for BIM-FM integration, as it dictates the interoperability of systems. Moreover, it is regularly updated, meaning the limitations of one release are often addressed by subsequent releases. Therefore, papers prior to 2018 were generally excluded as, at the time of writing, it was when the latest official IFC update (IFC version 4.0.2.1 [29]) was released. From an initial literature search, it was confirmed that earlier papers encountered issues caused by previous IFC versions, which were rectified by subsequent releases. Unless otherwise specified, all standards given herein refer to international standards.

Section 3 provides further analysis on the publications reviewed. All bibliometric data were retrieved from Scopus.

2.2. Evaluating the Current Implementation of BIM-FM

From the reviewed literature, the current state of BIM-FM integration was found to be minimal (Section 4), with few examples of BIM being used as a live FM tool. It was also found that the implementation was greatest when there was a clear understanding of the potential value that BIM-FM could provide, as discussed in Section 5.

2.3. Identifying Barriers

Since the actual application of BIM-FM was found to be minimal, the barriers to its implementation identified from the literature are discussed in Section 6.1. The differing opinions between academicians and FM practitioners are discussed in Section 6.2. The key barriers are subsequently identified as follows:

• Lack of data exchange formats (Section 6.3);
• Lack of FM alignment (Section 6.4);
• Lack of standardisation (Section 6.5);
• Lack of data retention (Section 6.6).

2.4. Identifying Opportunities

From the findings of Section 6, it was concluded that defining comprehensive requirements for BIM-FM could help overcome some of the identified barriers. As such, attempts within the reviewed literature to define and facilitate information requirements (Section 7.1), functional requirements (Section 7.2) and modelling requirements (Section 7.3) are discussed. The application of other technologies, such as Digital Twins, to meet some of the requirements is discussed in Section 8. A new conception of BIM-FM is outlined in Section 9.

3. Bibliometric Analysis

A total of eighty-four publications were identified as part of the literature search. Figure 1 depicts the country of origin for the reviewed papers, and Figure 2 depicts the number of citations per country. The United States, the United Kingdom and Italy had the greatest apparent research output, with the United States having the greatest number of publications and the United Kingdom having the greatest number of citations. Neither result was unexpected. For instance, the COBie data exchange format [30], the exchange format created specifically for the transfer of operational information from a BIM model, was created in the United States. Furthermore, both the United Kingdom and Italy mandate the use of BIM on government-funded projects [31].

Figure 3 depicts the co-occurrence of keywords for the literature reviewed. As suggested in Section 1, there is much overlap between FM and AM, which can be seen from the large co-occurrence of papers using both “facilities management” and “asset management” as key words. Figure 3 also suggests that there is increasing interest in the application of other technologies within the field of BIM-FM such as GIS (discussed in Section 6.3) and the Internet of Things (IoT) (discussed in Section 8).

4. Current State of BIM-FM

Currently, the most utilised approach to BIM-FM involves the BIM model being used to generate information about an asset and then the information being transferred to another format or system [11,33]. Generally, this occurs at the handover stage of a construction project using a pre-existing BIM model. Currently, the only officially recognised method for this is the COBie data exchange format [30], which results in a spreadsheet output, meaning the visualisation benefits of BIM are lost [18]. The use of spreadsheets may be due to the apparent skills gap between the people in charge of managing and inspecting assets and BIM practitioners [34,35]. Even when COBie is used, information and models are typically manually updated and input into FM systems from BIM, a time-consuming process prone to human error.

To overcome this, an increasingly common approach to BIM-FM is the use of a visual programming software, such as Autodesk’s Dynamo [36], which creates a link between the software [37]. The program produced is usually bespoke for a project and is often only uni-directional, so works similarly to COBie. Sadeghi et al. [38] were able to create an automatic data validation and exchange process using Dynamo but stated that “automatic BIM data retrieval, verification, and transfer to FM requires systematic and robust identification, clarification, and conveyance of the model requirements including syntax, semantics, and geometrics in early stages of the project lifecycle” [38]. This sentiment is also echoed by Matarneh et al. and the Global Forum on Maintenance and Asset Management [33,39]. Visual programming was also employed by Jofré-Briceño et al. [40]; however, there does not seem to be any validation of their proposed method by the end users, a common occurrence in the existing BIM-FM integration work. This is perhaps due to a current lack of defined ‘best practice’, meaning consistent quality control is not possible. Furthermore, the validity of just passing on a BIM from the construction stage to FM is questioned by Rogage and Greenwood [7], who found that “simply ‘passing on a model’ does not necessarily mean that the model was used or even usable”.

Whilst it is true that some works have attempted to use BIM as a live tool for FM [41] or have created new, bespoke BIM-FM platforms [42], these are minimal [43]. Volk et al. [23] likewise identified a critical focus towards activities common to construction with limited investigation of activities unique to FM. Failing to use the BIM itself for FM arguably makes the creation of a new BIM model obsolete for an existing asset. However, the apparent research focus on BIM-FM integration from the construction phase may be attributed to the fact that if a BIM is already being created at the construction stage, the investigation of its transferal to the FM stage requires much less investment both in terms of time and cost than a building with no pre-existing BIM.

Figure 4 summarises the two most common instances of BIM-FM outlined in this section. As discussed, the most utilised (current) approach to BIM-FM consists of the information contained in the BIM being transferred to other FM systems in a uni-directional transfer. For the purposes of this paper, a BIM created according to the specifications of ISO 19650 and incorporating information up to the handover stage of an asset lifecycle will be referred to as an AEC-BIM. Whilst this transfer of information from the AEC-BIM helps populate the FM systems with relevant asset information, the visual benefits of BIM are usually lost, and there is rarely a single FM system capable of completing all FM functions, resulting in siloed information. The developing approach (developing) is BIM as a live tool, where there is a bi-directional information transfer between the AEC-BIM and the FM systems. This is an improvement from the current approach as, whilst the FM information is still siloed, the BIM provides a single source from which to access the other FM tools. However, as previously stated, instances of this approach are minimal and still developing.

5. Justifying BIM-FM Integration

Woodward and Heesom [44] suggest that BIM implementation requires significant investment, so there is a need for clients to understand its worth. Furthermore, Brunet et al. [45] studied BIM for AM implementation in three companies and found that the BIM implementation was good where there were influential figures and asset owners within the company that encouraged its use. To this end, this section will provide a brief overview on the justification of BIM-FM according to the literature.

Munir et al. [46] investigated the business case for the use of BIM-FM and identified the following six areas where BIM can provide value:

• Management value—benefits to the organisation regarding data and task management;
• Commercial value—functionalities allowing cost saving and cost avoidance;
• Technology value—functionalities not typical elsewhere including increased connectivity and data visualisation capacity;
• User value—benefits to the end user regarding how to access data and undertake tasks;
• Industry value—increased collaboration and productivity as well as the potential application of new business models;
• Efficiency value—better asset optimisation and utilisation.

The values suggested by Munir et al. [46] are echoed by other works by Patacas et al., Thabet et al. and Durdyev et al. [5,10,47], also suggesting that there is management value to BIM-FM as a fully enriched BIM model would provide a structured single source of asset information. Likewise, Gurevich and Sacks [48] undertook a three-year study of BIM adoption by public construction clients. They identified “long-term benefits from knowledge and information retained in the organization and across the supply chain” [48]. Whilst their work referred to construction clients, the benefits remain, as one known issue with the management of pre-existing assets is that knowledge is rarely retained in an accessible format [39]. Jiang et al. [14] similarly cited technology value associated with visualised information analysis and reporting, and Benn and Stoy and Kameli et al. [49,50] suggested there is a commercial value due to potential cost savings. Kameli et al. [50], who created a BIM and Radio Frequency Identification (RFID) FM system, claim that whilst their approach had high upfront costs, in the long run it would reduce maintenance costs by 50%.

An assumed limitation of BIM for existing buildings is that geometric survey techniques, such as laser scanning, typically only provide surface data, meaning the identification of elements beyond the surface is not possible without the use of additional technologies [23]. However, the work described by Woodward and Heesom [44] provides justification for BIM-FM in this regard. Within both the case studies described in their work, they used a BIM environment to compare the existing condition of a building with historical drawings and consequently were able to identify areas with potential concealed elements. In both cases, hidden elements such as a horse trough and decorative plaster crests were revealed in subsequent work in those areas. It can thus be concluded that BIM can improve the identification of hidden elements in a non-destructive manner.

Whilst the assumed values of BIM-FM are numerous, there is currently a lack of case studies presenting tangible evidence for these [51]. For example, Dlesk et al. [52] suggest BIM and FM are not yet fully integrated because there is not yet much legislation mandating the use of BIM in FM, which may explain the lack of practical case studies. Interestingly, the study by Lavy et al. [13], one of the only studies attempting to provide quantitative evidence of the benefits of BIM, compared work order processing times using BIM-FM and a pre-existing CMMS system and found that BIM-FM increased rather than reduced the time taken. However, it is worth noting that the participants did not have BIM expertise. Moreover, the CMMS system was well established, so it is not unreasonable to assume the participants would use a more familiar system quicker. Thus, their findings should not be considered conclusive unless the study is continued for a longer period and the participants are actively encouraged/trained to become familiar with BIM. This argument is supported by Chung et al. [41], who utilised BIM to create an Augmented Reality (AR)-based FM system and found that it improved the efficiency of work by increasing the accessibility of data and speed of decision making. Unlike Lavy et al. [13], their validation was mostly carried out by users with some level of experience with the system, which may have influenced the apparent ease of use. Arguably, with very few practical case studies of BIM-FM implementation, it remains hard to justify any of the suggested values. Even the work by Khan et al. [53], which was validated by a team of experts and was found to achieve a good level of success overall, experienced issues with using BIM for subsequent decision-making and planning maintenance, resulting in uncertainty in BIM-FM value. Future work should be made to provide quantifiable evidence of these values to justify further investment in BIM-FM.

6. Barriers Affecting the Implementation of BIM-FM

6.1. The Academic View

The academic interpretations of BIM-FM agree on a number of key barriers, which apparently explain the lack of BIM-FM case studies. These can be summarised as follows:

• Insufficient data exchange formats [5,10,54];
• A lack of standards including defined information requirements and required level of development [8,11,24,41,44,49,51,52,55];
• High costs for both the software and staff training [14,46,47,50,55,56,57,58,59].

Consequently, attempts have been made within the research community to reduce the cost of model creation. Gouda Mohamed and Mousa [56] attempted to reduce costs associated with the laser scanning equipment by utilising the LiDAR scanner incorporated into the iPhone 12 Pro. However, if this was carried out as a handheld procedure, it would not be unreasonable to assume that inaccuracies may have occurred due to the intrinsic operator movement during the scans. Although Gouda Mohamed and Mousa [56] successfully built a model, neither the level of accuracy obtained, nor how, or if, these potential inaccuracies were accounted for is discussed. Solutions to training barriers have been considered less within the literature, with Pinti et al. [24], who were researching the integration of BIM-FM in the public domain, further commenting on the barrier of training, asking “how can training be achieved at such a large sector and who receives training first?”. Whilst Pinti et al. [24] were specifically referring to public sector entities such as hospitals and schools, the issue of scale is arguably considered true for many organisations that manage multiple assets. Further perceived barriers include the time taken to make models [40,60], the need to consistently update them so they remain relevant [12,16,59] and the need to make data accessible to those without BIM ability or access to BIM tools [16].

Dixit et al. [21] carried out a literature review on the issues surrounding BIM-FM integration. Similar to this research, they identified an initial set of four key areas:

• Technological issues;
• Cost issues;
• Legal issues;
• BIM information management issues.

However, when they attempted to validate their findings by interviewing FM professionals, the participants, unexpectedly, disputed the majority of their suggestions. However, the participants did agree that a lack of FM involvement at early stages of BIM planning and a lack of interoperability between stages negatively affect BIM adoption.

This disparity in views between the perception of academics and FM practitioners (discussed further in Section 6.2) could be explained by the mere fact that many of the issues perceived by academics could be easily overcome by developing a justifiable business case for BIM-FM implementation—incorporating the cost of the technology and training. Whilst organisational implementation of BIM-FM would likely require some change in working practices and upskilling of staff members, this requirement is not unprecedented for any new approach. However, it is understandable that researchers carrying out pilot studies would not have the authority to implement organisational change and thus would perceive these barriers as greater than they are.

6.2. The FM Practitioner View

As evidenced by Dixit et al. [21], FM practitioners generally agree that the lack of FM involvement at early stages of BIM planning and a lack of interoperability between stages negatively affect BIM adoption. This statement is echoed by Yusoff and Brahim [57], who initially claimed that the problems of BIM-FM (in their case specifically, HBIM) can be overcome with “well-trained BIM knowledge as well as concept and processes”. This sentiment was subsequently found to be an oversimplification when Yusoff and Brahim [57] later interviewed a range of (established) conservators and facilities’ managers and identified additional barriers including, but not limited to the following:

• Difficulties collecting data for older buildings;
• Time related to creating a BIM;
• Expenses related to servers and licensing.

Whilst Yusoff and Brahim [57] only interviewed five experts, four of the five experts had over ten years’ experience in the construction industry indicating a high level of reliability to the data. Jang and Collinge [55] suggest that the standards (COBie) and processes are well-developed, however there are systemic issues associated with how the industry actually operates (referring mainly to new build construction), which prevent the application of BIM-FM. Specifically, there is a disparity between the client and industry’s understanding of their management needs [21].

The Architecture, Engineering and Construction (AEC), FM and BIM sectors have fairly well-established working practices, however with minimal commonality. This itself is a barrier, evidenced by Moretti et al. [61], who found that information storage and management techniques during construction do not translate to the operation stage and thus result in incomplete data sets. This disparity may, arguably, have resulted from the lack of FM involvement at the planning stage, as discussed by Dixit et al. [21], in addition to an overall industry resistance to change [59].

To avoid another over-simplification of the issues affecting BIM-FM, the subsequent sections will elaborate on some of the key barriers the authors believe most accurately limit BIM-FM implementation:

• A lack of data exchange formats;
• A lack of alignment with FM;
• A lack of standardisation in BIM-FM;
• A lack of data retention.

6.3. Lack of Data Exchange Formats

A number of authors have acknowledged that FM requires a complex selection of interrelated data and argue that BIM authoring tools alone are insufficient for managing all the data required for FM [25,50,62].Whilst BIM conceptually requires all information to be stored in a single space, other authors such as Munir et al. [46] suggest that this can involve multiple systems as long as the systems are linked. Thus, data must be transferrable across platforms, and hence, the Open BIM exchange formats (e.g., COBie) were created to this end [54]. However, as previously mentioned, the existing exchange formats are currently insufficient for BIM-FM. For example, Patacas et al. [5] compared IFC 4 and COBie entities with the asset register requirements given in BS 8210:2012 [63] and found that, of the twenty-two asset register requirements, IFC and COBie were not able to support ten and seven of the twenty-two requirements, respectively. Whilst they were able to overcome this issue with custom ‘Property Sets’, they experienced further issues with IFC entities which were then incompatible with the IFC exporter.

Although created for FM, COBie has a number of limitations:

• It does not encompass all elements required for FM [44,64];
• It does not encompass all types of assets [65];
• Some elements have limited practical value [66];
• The structure is not easily customisable [66];
• Verifying the quality and consistency of changing data is difficult [67];
• There are interoperability issues with BIM authoring tools [12,33,68].

There appears to be limited evidence of COBie and IFC working sufficiently except in the works by Chung et al. [41], who were able to use existing COBie formats to create an AR-based FM system from a BIM model, and Marmo et al. [37], who created a BIM-FM integration for FM at healthcare facilities and claimed that IFC 4.2 was sufficient for all their identified information requirements.

There have subsequently been many calls for the expansion of COBie [8,44,64] and IFC. Attempts have been made to extend COBie, but there is limited evidence of success to date, such as in the work discussed by Shin et al. [65], who created a COBie for ports devoid of validation of their proposed schema. Furthermore, whilst Borhani and Dossick [8] advocated the use of a standard classification system for easier data transfer and extraction, they found, based on interviewing a number of asset owners, that they needed a certain level of customisation for their specific uses. Similarly, Rogage and Greenwood [7] tried to automate the data transfer from the design and build stage to the FM stage and found that even when a full IFC standard was used, human error in object classification and incompleteness of data meant that the process could not be fully automated. However, despite potential human limitations, some evidence for the effectiveness of updating the IFC schema (and consequently other standardised exchange formats) is given by Garramone and Scaioni [69], who used the ‘IFCalignment’ entity, a new feature of IFC 4.3, to successfully export GIS raster data, a functionality not previously possible. Similarly, Ciccone et al. [62] created a federated model of a railway using infrastructure categories introduced in IFC 4.2 and state that its update is key to interoperability and functionality of models.

The issue of exchange formats is not limited to IFC and COBie. Farghaly et al. [70] investigated combining multiple ontologies to account for information from different sources, e.g., BIM, sensors and databases required for asset management, and there has been some notable success with alternative technologies. At a basic level, a lot of information needed for FM originates from the manufacturers of maintainable assets and is often provided in a PDF format. Whilst it is possible to attach PDFs to BIM, the information is then not able to be queried; manually inputting data to allow querying is time-consuming and prone to error [71]. Niknam et al. [71] suggested the use of semantic web technology to improve the accuracy and efficiency of data population from different sources. Their work showed considerable improvements in the time taken and number of errors when populating a BIM with manufacturer information compared to manually inputting data from PDFs. However, Azzran et al. [9] tested four different approaches to BIM-FM integration (IFC, COBie, manual and middleware software) and found that no approach was sufficient for all requirements and no approach could handle native linked models (a functionality key for federated models). Furthermore, works attempting to integrate BIM and GIS have found recurrent issues around differing geometric representations across software and a lack of automation in the integration [16].

Fundamentally, “what is sent must be the same as what is understood” [72], meaning data exchange must not involve so much data loss or conversion that it becomes unrecognisable. Floros and Ellul [6] converted a 3D model to IFC and then the IFC model to both AIM and GIS for AIM and recorded the information losses. They encountered semantic information losses at all stages and geometry losses for the IFC-AIM and IFC-GIS for AIM conversions. As well as a loss of meaning, there are also costs attributed to data loss between the stages [73].

6.4. Lack of Alignment with FM

The lack of FM alignment with other sectors, particularly the BIM and AEC sectors, may be attributed to a lack of understanding of FM by other sectors. A common misconception of FM held by the AEC sector appears to be that too much information is created during the design and construction phases of an asset. Therefore, the AEC sector assumes it must be reduced to a manageable amount for FM at the handover stage to enable leaner data storage [74]. This mindset may be due to the fact that having surplus information and models of insufficient quality causes functional error when the information is exported from BIM to FM [75], and the easiest fix is to reduce the amount of information. However, the lack of standardisation regarding information requirements for BIM-FM means the resulting models often fail to contain important information whilst simultaneously containing redundant information [64]. In this respect, it is worth remembering that management decisions are not (and should not be) made in isolation. The interrelatedness of management data and decisions is evidenced by the Global Forum on Maintenance and Asset Management [39], who identify thirty-nine subject areas of AM and provide some examples of their (non-constant) interrelationships. The complexity of decisions is also acknowledged by Wang and Piao [76], who used fuzzy Multi-Criteria Decision Making (MCDM) [77] to account for the range of qualitative and semi-quantitative data that they suggest informs maintenance decisions. A practical BIM-FM example is provided by Piaia et al. [42], who created a platform for BIM-FM of a cultural heritage asset for onsite condition assessment and management. To achieve this aim, the BIM-FM system included condition information, financial information, and the impact of potential work on the cultural value. Whilst limited to a single use case, it demonstrates that a single FM activity requires multiple data inputs.

There is abundant information generated at the construction stage, which is theoretically useful for FM. However, the associated high cost of manipulating, storing and managing these data should be justified [6]. Consequently, it appears more appropriate to suggest that the information from the design and construction phase must be filtered to the small selection relevant to FM and supplemented with additional data unique to FM. This implies that the recommendations of works, such as that by Dixit et al., Benn and Stoy and Stride et al. [21,49,59], which suggest that the general lack of information for FM in BIM should be overcome by greater FM involvement in the design and construction phase are once again oversimplifying and misunderstanding the situation. There also needs to be a greater effort to integrate BIM and FM sectors. Tsay et al. [11] found that even with FM involvement at the design stage, for information requirement development, the process was limited by the FM practitioners’ lack of BIM knowledge.

6.5. Lack of Standardisation in BIM-FM

Whilst part of the alignment issue is arguably due to a general misunderstanding of different sectors, it must also be acknowledged that there is also a lack of alignment between the official standards and practices for BIM, FM and AEC sectors. For instance, Lu et al. [25] compared the BSI and ISO standards for asset management and BIM and found that there is no overall framework to link them, and they are often juxtaposed, as evidenced by Kula and Ergen [12], who reviewed a BIM-FM platform for an airport which did not have a BIM created during its construction and found that the BIM team and FM team had contradicting requirements. Furthermore, Sadeghi et al. [78] found that many BIM models for design and construction are developed using organisation-specific naming conventions, which cause issues relating to data extraction and management when they are transferred to another organisation for the FM stage. Likewise, Kula and Ergen and Alnaggar and Pitt [12,66] identified the same issue, and Alnaggar and Pitt [66] go on to claim that even when there were consistent naming conventions, there was a lack of quality control to ensure the procedure was being followed.

Whilst Lu et al. [25] found that existing AM and BIM standards are often in opposition, it is also true that the existing ISO BIM standard [19] is not yet fully developed for FM. Therefore, it is not unreasonable to suggest that it would be simpler for BIM-FM standards and processes to aim for alignment with FM practices rather than trying to align new construction and FM practices.

6.6. Lack of Data Retention

Marmo et al. [22] suggest that “when no BIM exists yet, it is crucial to do the prior analysis of stakeholders’ information requirements to optimize geometric modelling and information handling effort”. In a study focused on COBie and BIM-FM integration from a Mechanical and Electrical (M&E) contractor perspective, Jang and Collinge [55] stated that “interviewees maintained that early engagement from the client is considered the most important factor to produce accurate COBie information”, asserting that ill-defined information requirements lead to consistent changes and errors in the scope.

However, it is worth noting that even if all the information requirements for the existing assets are known, much of this data can be erroneous or missing, a likelihood which increases with the age of the asset [60]. Furthermore, since FM covers the whole operational life of an asset, it is not unreasonable to assume that new data will need to be added to the system over time to reflect any changes, whilst retaining the historical data. However, this functionality is not yet inherent in all proposed BIM-FM systems since it is not a utility necessarily required in construction, which the majority of BIM authoring tools were created for. For instance, Ergen et al. [79] found that when a model object was majorly updated or deleted, previous information associated with that object was removed from the model and only accessible by querying their FM database. As such, to avoid this data retention problem from recurring, a BIM-FM system should allow new data to be added over time, without compromising existing data, and whilst creating a historic record of any changes.

6.7. Opportunities to Overcome Barriers

It is clear from the findings of the previous sections that defining comprehensive requirements for BIM-FM could help overcome some of the identified barriers. Therein, defining the information requirements would allow exchange formats to be expanded to encompass FM, thus facilitating the integration of other digital FM tools. Moreover, defining functional and modelling requirements would allow future software development to be tailored to actual end user needs and would justify the integration of different technologies. Allied to this is a comprehensive understanding of all requirements, which would facilitate the creation of a new BIM-FM standard, linked with ISO 19650, to encompass the whole asset lifecycle.

Adoption of BIM has been shown to be increased by official mandates [31]. However, BIM level 2, the original requirement of the UK BIM mandate [80], does not extend beyond the handover stage of a construction project. Standardisation of BIM-FM would allow any future mandates to incorporate FM and allow quality control procedures to be mandated, increasing BIM-FM adoption and enhancing the overall output of BIM-FM.

The proceeding two sections of this paper will discuss notable contributions in the literature to the identification and fruition of BIM-FM requirements. Consequently, an overview of the theoretical requirements for BIM-FM (i.e., information requirements, functional requirements and modelling requirements) will be provided (Section 7) as well as a discussion of the most commonly applied approaches and technologies (Section 8).

7. Theoretical Requirements for BIM-FM

7.1. Theoretical Information Requirements

Early work in BIM-FM looked at the recording of asset data at the time of model creation, with little regard to any future works that would/could be undertaken [44]. Hence, beyond a geometric model, there was limited consensus for BIM-FM information requirements. Furthermore, very few works suggested information requirements for a management-focused BIM [73] and those that have tended to focus on either high-level requirements [64], a very specific use case (e.g., Woodward and Heesom [44] who utilised heritage markers to indicate areas identified in the statement of significance for a Grade II listed building) or a small subset of requirements (e.g., Dias and Ergan [74] who identified information requirements specifically for Heating, Ventilation and Air Conditioning (HVAC) systems). This is likely because specific use cases often have multiple data inputs, meaning comprehensive identification of information requirements at all scales is unfeasible. For instance, Ergen et al. [79], who focused only on occupant feedback, generated eight information inputs for a single event. Therefore, to realistically define information requirements for a standard approach to BIM-FM, there is a need to generalise information into discrete sets of data. For instance, using Ergen et al. [79] as an example, the information requirement would be defined as ‘occupant feedback’. This requirement would then consist of additional data inputs, which could be defined on a case-specific base (in their case eight inputs) encompassed within the overarching ‘occupant feedback’ requirements. This is comparable to the concept of Organisational Information Requirements (OIR) defined by ISO 19650, which refers to information required by an organisation to achieve organisational functions or goals [19]. Generalising data sets also enables the implementation of access controls to data, allowing control of which user gets access to what information, e.g., for security reasons and for uncluttering the user experience [52].

Only a few works, within the literature reviewed, attempted to identify comprehensive high-level information requirements for FM [53,81], the most detailed of which is found in the work by Khan et al. [53], who identified eight high-level domains of information and thirty-eight sub-domains. Matarneh et al. [81] developed a list of forty-one information requirements by reviewing the literature and standards and validating them with interviews of FM professionals. They were not referring to existing buildings, but their work was, regardless, a comprehensive identification of FM information requirements. The limitation of their work is that the intention was to transfer the information to an FM system at a later date, thus rendering the BIM model obsolete. Di Filippo et al. [54] carried out client interviews to generate a list of information requirements, but only managed to identify a total of five requirements with unclear justification.

A recurrent theme for identified information requirements is maintenance and condition-related information. Jofré-Briceño et al. [40] created an asset management tool for ports, which combined a BIM model with a maintenance severity and urgency scale in Excel to visualise and plan maintenance needs. The authors were able to create a bi-directional link between the excel sheet and BIM using Dynamo. The key benefits of the BIM were the visual filtering capabilities of Revit. Moretti et al. and Marmo et al. [16,22] also included condition assessments in their work, citing similar applications regarding planning proactive maintenance. Similarly, Akcamete et al. and Yoon et al. [18,82] suggest that location data are key to analysing maintenance data and thus enabling efficient planning of future work. However, Yoon et al. [82] carried out validation of their work with FM practitioners who agreed with their proposed system but noted that additional functionalities and information types are required to meet their needs. Akcamete et al. [18] also found the inclusion of location data useful. By visually recording maintenance within BIM, they were able to infer potential causes of maintenance issues by observing where the maintenance was required and if there were any nearby maintenance that may have been caused by the same issue or have caused the issue. Piaia et al. [42] are an outlier and said that, whilst beneficial, the additional effort required to add location information was not comparable to its usefulness. Stride et al. [59] suggest cost information is key to monitoring future and past maintenance.

Table 1. Information requirements for FM identified in the reviewed literature.

Information Category	Specific Information Type	Sources
1. Safety and security information	Fire safety Health and safety Potential threats/risks and vulnerabilities Security	[5,73]
2. Location data	Asset location Maps of surrounding area	[16,42,54,60,73,74,81]
3. Space data	Space usage Occupation limits Space breakdown Room allocations	[22,35,53,54,60,74,81]
4. Maintenance manuals/instructions	Required equipment Minimum level of performance Maintenance to be performed by user Maintenance to be performed by skilled personnel Intervention type Intervention frequency	[16,18,22,35,40,42,44,53,54,73,78,81,82]
5. Information about the fabric of the asset	Materials data (including properties) Architectural data Structure type Condition assessments and surveys	[16,18,22,34,40,42,44,53,54,60,74,81,82]
6. Environmental data	Light levels Relative humidity Temperature	[22,35,53,60,79,83,84]

presents an overview of recurrent information requirements determined from the papers reviewed. The requirements are limited to instances where the given function was FM and the requirements were given by more than one paper. Six categories of information were identified. The most commonly cited requirements (those referring to maintenance (category four) and the fabric of the asset (category five)) related to ‘hard’ FM activities, which refer to physical assets and activities, with a smaller focus on the ‘soft’ FM activities, namely safety and security information (category one). This is consistent with previous research by Pinti et al. [24], identifying a similar research focus on ‘hard’ FM. The table can be taken as an overview of theoretical information requirements for BIM-FM according to the literature.

7.2. Theoretical Functional Requirements

From the reviewed literature, two key functionalities of BIM-FM can be identified: the ability to access and input data in the field [15] and the ability to record previous work and create intelligent maintenance schedules [54].

7.2.1. Inputting and Accessing Data in the Field

One simple approach for locating corresponding BIM objects in the field involves the use of physical markers like QR codes [41], a tool already utilised by FM practitioners. Gouda Mohamed and Mousa [56] incorporated pre-existing QR asset tags within Navisworks to set up a BIM for an Egyptian University building. The QR codes allowed pre-existing information associated with a specific object, stored in an Excel format, to be automatically populated within the corresponding BIM object. Whilst the case study given was for the population of a new BIM, it is reasonable to assume that the QR codes could be used to populate new information over time. It is currently unclear if QR codes could be widely applied in a heritage context especially if the listed building status has to be abided by.

Moretti et al. [16] suggest that identifying objects via other location data is more reliable than tags or QR codes because physical markers can be moved. This suggestion relies on an asset having a fixed position in space, so would exclude moveable assets such as furniture, whose location/quantity is often required to inform FM decisions. Alternatively, both Ma et al. [85] and Ergen et al. [79] used Bluetooth Low Energy Beacons [86] to attribute real-world locations to a digital model. Having a shorter range than conventional Bluetooth, the beacons were able to locate the room/zone where the worker was located. Another approach involves RFID integration, which has been shown to significantly reduce the time taken to carry out normal FM tasks [50].

An emerging alternative to physical markers is automatic feature recognition. However, whilst specifically trying to locate a worker in the field, El Ammari and Hammad [87] state that feature recognition is not sufficient indoors due to repetitiveness and the uniform nature of some buildings (particularly office buildings). They chose to use a combination of marker tracking (a similar approach to RFID) and feature tracking. Chung et al., Wang and Piao and El Ammari and Hammad [41,76,87] also employed a virtual solution, utilising AR to allow field personnel to associate digital objects with their real-word equivalent and collaborate with remote workers.

For the population of occupant feedback on facilities, Ergen et al. [79] allowed occupants to pick the exact issue location on a 2D or 3D model. This assumed a degree of comprehension of the 2D drawing, which may not be true for all users and locations, an issue later identified as part of the location validation stage. It is evident that users inputting incorrect information would likely have adverse effects on subsequent remedial work, so a greater location accuracy of technology-based solutions may be desirable.

7.2.2. Intelligent Maintenance Schedules

An acknowledged issue with current FM practices is the tendency to undertake reactive maintenance rather than proactive maintenance [47], which can be an inefficient and often expensive use of resources. To enable proactive maintenance BIM-FM needs to perform two functions: monitor future maintenance requirements, and record previous maintenance activities.

Several works have employed BIM-FM to actively track the urgency level of future maintenance requirements, often by colour-coding objects according to the urgency/severity [40,44]. Setting up this approach generally requires good prior knowledge of maintenance programmes, but ongoing programmes may then be updated according to the newly inputted data. Most existing case studies are in the implementation stage of this function, so do not discuss its long-term success.

There is evidence to support the argument that storing previous maintenance records within BIM enables proactive maintenance. For example, Akcamete et al. [18] found that storing previous work orders in a BIM format enabled them to identify the cause of previous issues by providing a visual correlation between incidents originally considered unrelated. However, as previously discussed, Ergen et al. [79] found that when a model element was majorly updated or deleted and replaced to maintain visual accuracy, previous information associated with that model was removed. Whilst the data were still accessible within the FM database, their availability would rely on the personal experience of the user, so would not be viable in the long term.

7.3. Theoretical Modelling Requirements

For FM purposes, a geometrically simplified BIM is usually considered sufficient due to the importance placed on the descriptive detail associated with objects [34,52]. In their review of the purposes of BIM for FM in Corporate Real Estate Management (CREM), Benn and Stoy [49] found that BIM-FM users established visualisation to be the lowest priority in BIM for FM development. Furthermore, Munir et al. [46] conducted a study investigating a company with a high level of both BIM maturity and practical application and found that the company switched from 3D models to 2D CAD objects due to a lack of any real benefits of a 3D model. Another justification for model simplification is that FM managers are unlikely to have access to sophisticated computers, so geometrically complex models will result in a non-optimal user experience, as Dlesk et al. [52] found when they tested models of three different complexities on two computers of differing quality (i.e., processing power and graphics capabilities). Kula and Ergen [12] also found that complicated model objects reduced the performance of the model. They opted to use simplified generic models with some resemblance to the real condition, as opposed to low-complexity symbolic models, so that objects could be recognised more easily.

Photography is a key resource for FM as much necessary information can be inferred from an image. This may explain why simple models are often considered sufficient. However, accessing and organising this information is exceedingly difficult especially when an asset has a long service life or is complex in nature [85]. Theoretically, a point cloud could provide the same information in a more understandable format as evidenced in Marzouk and Ahmed [88], who designed a BIM-FM system for a wastewater treatment plant. They collected data using laser scanning and then separated the points referring to each water pump into regions in Autodesk Recap 360. There were no apparent issues reported with this as all the pumps could be visually distinguished, and the required maintenance information was linked to each colour-coded region. The integration of accurate point clouds with simplified geometric models could serve as an effective compromise for BIM-FM, whilst acknowledging the likelihood that sophisticated software systems may not be available.

Whilst acknowledging that complicated models experienced system performance implications, Dias and Ergan [74] carried out focus groups with FM practitioners and found that they required a minimum Level of Development (LOD) of LOD 350 with some participants requesting LOD 400, both of which require highly precise geometric representations [89]. The authors were specifically investigating HVAC systems, where the use of high geometric accuracy models has been previously established as a requirement by the BIM community [90], so this may not be indicative of all building elements. However, there are other notable arguments against model simplification for FM. For instance, to utilise quantity take-offs for cost estimation, the geometry must be reasonably accurate [59]. Moretti et al. [61] discussed the need to aggregate individual model elements from the design stage into larger, simplified components during the asset management stage. However, this is contradicted by [42], who state “one would conduct very different activities if a defect was identified for a wall finish than for a wall’s structural layer”; thus, there is a need for some granularity of data and additional semantic meaning.

Based on the evidence, it seems to the authors of this current paper that the tendency to utilise simplified BIM models for FM may be due to the difficulties of creating and utilising a complex model, as opposed to the actual need of BIM-FM end users.

The previously apparent simplified modelling requirements [34,52] for FM may also explain the lack of research into advanced model creation in BIM-FM. Notable exceptions include Usmani et al. and Romero-Jarén and Arranz [58,91], who investigated scan-to-BIM, and [92], who discussed laser scanning processes. In addition, Romero-Jarén and Arranz [91] utilised machine learning to create an algorithm capable of automatically segmenting point clouds into planar objects (walls, columns, ceilings and floors) and non-planar objects (cars, etc.). Whilst they found their approach to be successful, they noted that it experienced issues if the objects assumed to be planar (walls and floors, etc.) were not. This arguably limits their algorithm to newer constructions as older structures, particularly heritage ones, are characterised by complex and heterogeneous geometries.

Whilst technological considerations are justified, it is arguable that the rapid development of BIM technologies and machine learning techniques, particularly prevalent in the field of HBIM [27], will render some concerns obsolete in a relatively short time. Thus, any future definition of a BIM-FM standard should also consider in greater detail the modelling requirements of end users without implicitly imposing limits due to currently available technology.

8. Integrating Other Technologies

A study by Mannino et al. [26] noted that it has been frequently suggested that BIM alone is not enough to meet all the requirements of FM [26] and that BIM should be integrated with the Internet of Things (IoT). However, they observed that such an integration is still in its infancy, partially due to a lack of existing studies which encourage FM to utilise IoT and a lack of interoperability between IoT technology and BIM [26]. A one-way integration between IoT and BIM would constitute a Digital Shadow [93], a virtual replica of an asset informed by real-time data.

Another potential area of integration is a Digital Twin [25]. The term ‘Digital Twin’ is often incorrectly used interchangeably with BIM, but the main difference is a Digital Twin is a bi-directional simulation environment where the model state changes based on external data input [94] (e.g., from a sensor) and can be used to influence the real condition. Correa [94] suggested that the easiest stage to apply a Digital Twin is FM because it is much easier to track than construction. However, a current limitation of Digital Twins for BIM-FM is that static data and real-time data (dynamic) are managed differently (e.g., IFC is designed for static data only), so it is often hard to find a system that can look after both [20,61].

Consequently, there have been a number of attempts at sensor integration for BIM-FM with varying levels of success. Quinn et al. [20] attempted to use Dynamo to link the two data stores. Their work was somewhat successful but only mapped a single element at a time and contained some errors. It also was not tested by facilities’ users, so it is unclear how effective it would be in practice. Su and Kensek [83] created a new Revit plug-in capable of integrating sensor data related to thermal comfort into Revit. The plug-in was able to visually highlight if a sensor had failed or was reporting values out of an acceptable range. Moreno et al. [35] proposed a system for the integration of sensors to BIM using a series of pre-existing software and BIM plug-ins. Likewise, Kazado et al. [95] investigated integrating sensors with BIM using three different approaches: Revit, Navisworks and Navisworks with an API add-in. They found that Navisworks was preferable to Revit as users cannot accidentally edit the model elements (but the model can still be updated within Revit and reloaded). The API add-in was the best solution as it stored historical data, hence allowing the visualisation of trends. The study by Kuo et al. [84] is one of the only works to successfully extend COBie to integrate dynamic data. These examples suggest that, whilst not an automatic function, sensor integration with BIM is feasible with minor development.

Moreno et al. [35] do acknowledge that there are some issues with regards to installing sensors in historically significant structures. Furthermore, Ergen et al. [79] claim that real-time occupancy feedback is as important as sensor data as some comfort requirements cannot be measured by a sensor (e.g., how comfortable furniture is). It can be concluded that while the integration of Digital Twin/Shadow technology with BIM is feasible for BIM-FM, it should not be limited to sensor data alone, which would risk limiting its worth.

9. A New Approach to BIM-FM

Key barriers to BIM-FM implementation discussed in Section 6.4 and Section 6.5 included a lack of aligned procedures and standards between FM and other sectors. The authors believe that conceptions of BIM-FM previously outlined in Figure 4 do not adequately account for this misalignment. There remains an implicit assumption that BIM tools and procedures, originally developed for AEC purposes, can be expanded to incorporate FM needs/functions. However, expansion does not account for misaligned, sometimes, contradictory requirements, nor does it account for the vastness of FM needs. Furthermore, as previously established by a number of authors, the ability to pass on a BIM from the construction stage to the operation stage does not automatically make that model useable [7].

The authors would therefore like to propose a new conception of BIM-FM, as summarised in Figure 5. Within this new perception, the following apply:

• The ‘AEC-BIM’ serves as an input for the BIM-FM system. This is the BIM created for new capital projects. This would require minimal to no changes to established BIM procedures, standards (e.g., ISO 19650) and tools within the AEC sector. The input of an AEC-BIM is envisioned as an optional step that would reduce BIM-FM setup time without excluding existing assets with no AEC-BIM.
• The BIM-FM system is here treated as a new, separate entity. It should be a single system (although likely to consist of interconnected subsystems) that contains all information, and performs all functions required for FM and which maintains the visualisation benefits of the AEC-BIM by including a 3D model suitable to defined FM requirements. It is conceived as a system which will be populated and grow over time as new data is added, whilst retaining historical data.
• The BIM-FM system should be able to output an AEC-ready BIM able to be used for new capital projects. This is particularly important given the growing trend towards retrofitting existing assets as opposed to new construction projects.

The BIM-FM system would be defined by a new standard specific to the requirements of FM. Initial work for the definition of these requirements is included in Section 7. The standard should also incorporate new developments such as Digital Twin technology (Section 8). Whilst not confined by the existing BIM standards, the new system should utilise the established BIM exchange formats (e.g., IFC) to facilitate AEC-BIMs to be input/output to the system. A full definition of BIM-FM requirements would engender this expansion.

Assuming the existence of a full definition of BIM-FM requirements and the technological capability to realise the system, an asset owner will need to carry out the following procedure in order to utilise the new approach to BIM-FM:

• Define which of the BIM-FM requirements are applicable to their organisation;
• Create the BIM-FM system architecture to facilitate their defined requirements;
• Mandate compliance with their defined requirements for all internal and external actors.

The first stage is comparable to the creation of a BIM Execution Plan (BEP), the preliminary stage of AEC-BIM implementation [19]. Once these steps have been realised, asset owners will be able to populate the BIM-FM system with existing asset information. For new constructions, provided the BIM-FM system is compliant with established data exchange formats and external actors comply with the mandated requirements, the project team should be able to easily transfer any required information from the AEC-BIM to the BIM-FM system.

Whilst the perception of BIM-FM given in Figure 5 has not been formally proposed or implemented elsewhere, preliminary justification for the proposals made can be evidenced by the University of Birmingham (UoB) Estates department. UoB Estates is currently developing a comprehensive digital record of their entire estate, compiling both AEC-BIM models of newly constructed assets and BIM models created retroactively for existing assets and assets of historical significance. Their latest development, the creation of a “Macro-Twin” of the UoB campus using Autodesk Tandem, combines BIM information and BIM models with live environment data. Whilst implementation is ongoing, preliminary usage suggests it has facilitated more efficient decision-making. Further details are provided by Autodesk [96].

10. Future Work and Conclusions

This paper carried out a semi-structured literature review to evaluate the current implementation of BIM-FM for existing assets. Two approaches to BIM-FM, current and developing, were outlined. It found that the current practical application of BIM-FM is minimal, but that application is greatest when the value provided by BIM-FM is understood. Consequently, potential areas where BIM-FM provides value were discussed. However, it was found that the majority of identified values were theoretical and there was minimal quantifiable evidence to justify them due to a lack of practical use. Future work should aim to provide evidence to support these values by carrying out quantitative evaluation of BIM-FM case studies.

Subsequently, the barriers to BIM-FM implementation were discussed and key barriers were found to be a lack of exchange formats, a lack of FM alignment to other industries, and a lack of standardisation. It was theorised that defined requirements for BIM-FM could serve as an opportunity to mitigate these barriers. As such, the requirements for BIM-FM and the methods for encompassing them discussed in the reviewed literature were identified. This included information requirements (six information categories were identified), functional requirements (two functional requirements were discussed) and modelling requirements. It was determined that the modelling requirements for BIM-FM should be further investigated, as the apparent trend for simplified BIM-FM models seems to be primarily caused by software limitations as opposed to end user needs. Future work should also seek to validate the theoretical requirements identified herein. The incorporation of other technologies (primarily Digital Twins) to meet the requirements was discussed and found to be an area of increasing prevalence within research.

The limitations of the two existing approaches to BIM-FM were discussed. Consequently, a new conception of BIM-FM was proposed. This paper subsequently suggests that new FM-specific BIM standards and procedures should be developed.

The authors are involved in ongoing research to evaluate the applicability of the theoretical requirements, proposed herein, to BIM-FM applied to assets of historical significance. A survey for the wider heritage community has been created to facilitate the validation of the proposed requirements and the identification of further heritage-specific requirements.