Black boxing BIM: the public client’s strategy in BIM implementation

Abstract

Building information modelling (BIM) is widely discussed in both industry and academia and studies have argued that public clients play an important role in industry’s implementation of BIM. While public client initiatives have gained much interest in previous research, there is still a lack of knowledge exploring the public clients’ role and choice of strategies when implementing BIM with the aim to change industry. Based on a case study of BIM implementation at the largest infrastructure client in Sweden, this paper applies the theory sociology of translation to explore the role of the public client as a driver for industry change and the strategy used during this process. The case study shows how the public client tries to influence its own organization and the construction industry as a whole. Findings show that the main strategy for enrolling actors to implement BIM is to demand its use in the procurement of all construction projects, thereby attempting to establish the BIM issue as a Black Box. A Black Box, a locked network element, which includes associated inscriptions aimed at prescribing BIM use in projects conducted by this client. The acceptance of this Black Box has however been problematic, resulting in a diverse influence on the demand for BIM in procured construction projects.

Keywords:

• Building information modelling
• BIM
• sociology of translation
• implementation
• actor-network theory

Introduction

Building information modelling (BIM) is often presented as a “new paradigm” in construction and with proper implementation it is expected to enable substantial improvements in the whole industry (Azhar 2011, Elmualim and Gilder 2014). However, widespread BIM implementation has not taken place (Gu and London 2010, Smith 2014). The general perception is that the construction industry is slow in adopting new and improved work practices and implementing innovations (Harty 2008). While some case studies have shown benefits with BIM use (e.g. Azhar 2011), there is also a growing stream of research exploring challenges (Vass and Gustavsson 2017) and providing critical perspectives on BIM implementation (Jung and Joo 2011, Gustavsson et al. 2012, Fox 2014).

To support industry change and the adoption of new technologies such as BIM, the role of client organizations has been highlighted in numerous studies (Azhar 2011, Takim et al. 2013, Lee and Yu 2015). Public client organizations have been argued to be in the right position, and with the needed power to demand change. Thereby, client organizations are found to be key change agents and are thereby suggested to play a vital role in the BIM implementation process (Wong et al. 2010, 2011, Porwal and Hewage 2013). Clients have also been presented as the actor with the greatest potential benefit from an implementation of the technology (Olofsson et al. 2008, Elmualim and Gilder 2014). It has also been suggested that when clients and regulatory authorities start to demand BIM when procuring consultants and contractors, the discussion will no longer be about whether or not to use the technology but about promoting more advanced BIM use (Linderoth 2010).

Following this discussion, a number of government initiatives for BIM implementation have been initiated around the world. One example is the UK where it has been stated that all public contracts awarded from 2014 and onwards should demand project participants to work collaboratively through the use of BIM (Cabinet Cabinet Office 2011). Similar examples can be found in the USA, Finland, Norway, Denmark and Singapore (Wong et al. 2010). However, research in other fields such as information systems (Ciborra 1997, Holmström and Stalder 2001) and organizational studies (Czarniawska and Sevón 1996) has shown that the use of technology, when applied to an organizational setting, often drifts away from the intentions of its developers or the intentions of the actor initiating the implementation. A recent study by Vass and Gustavsson (2017) revealed a mismatch between expectations and beliefs about the client organization’s role in BIM implementation. Still, not many studies have focused on clients’ perspective of the current implementation of BIM and the mission to change the industry.

In this study, BIM is seen as a technology with implications for many actors within the loosely coupled (Dubois and Gadde 2001) construction industry, which consists of inter-organizational projects as its main unit of production. More advanced BIM use is expected to result in substantial changes to work practices, especially in terms of a greater focus on inter-organizational collaboration (e.g. Succar 2009).

The purpose of this paper is to explore the role of public clients in BIM implementation and what strategies they use when trying to change industry by implementing BIM. Following this purpose, the findings from a case study of the BIM implementation at the largest public infrastructure client in Sweden is presented. This case is analysed using the theoretical lens of Sociology of Translation (Callon 1984). Hence, the case presented in this paper constitutes an empirical example of a BIM implementation process at a large public client. Further, this paper elaborates and examines the normative claims that public clients’ demands for BIM will more or less automatically make the industry adopt the technology.

Literature review

The network perspective of BIM

BIM is often described as a “new paradigm” within the construction industry, which addresses many of the challenges associated with the industry such as low productivity and lack of innovation (Azhar 2011). Examples of benefits from BIM use include reduction in time and cost of projects while simultaneously delivering higher quality (Yan and Damian, 2008). These benefits will be achieved through improved collaboration and decreased fragmentation (Azhar 2011). That is to say, BIM implementation is expected to have an influence on most actors participating in construction projects. This view of BIM influencing a multitude of actors is outlined in a well-cited BIM framework developed by Succar (2009). In this framework, several sequential maturity stages are presented. According to the framework, BIM use matures as actors work more collaboratively. A similar model of BIM maturity progression has been developed in the UK regarding BIM implementation, and other similar maturity models can be found in other normative frameworks, such as the project management maturity models (Crawford 2006). What these models have in common is that they describe a development trajectory which the industry is expected to follow in relation to its adoption of the technology.

BIM implementation, as well as other technology driven change processes, is challenging since work practices and business processes also have to change (e.g. Bryde et al. 2013). Need for such organizational change highlights the difficulties and complexities in the organizational and industry context (Vass and Gustavsson 2017). Substantial implementation difficulties also relate to legal issues, training of practitioners, etc. (Bernstein and Pittman 2004, Tse et al. 2005, Azhar 2011). Challenges also come in terms of technological difficulties: interoperability, standardization and the development of new tools (Azhar 2011, Steel et al. 2012).

The inter-organizational aspects of BIM use along with its collaborative implications, as describe by Bryde et al. (2013), among others, reveal the need for a project focus when studying and assessing BIM implementation. In order to evaluate the benefits from the investments in BIM, a shift in focus from individual actors to the inter-organizational network is needed (Dehlin and Olofsson 2008). It has also been argued that clients are the actor who will benefit the most from BIM implementation, which also emphasizes the value in assessing the results of BIM use on a project level rather than focusing only on single-project participants (Dehlin and Olofsson 2008, Olofsson et al. 2008).

In addition to having the largest potential gain from BIM implementation, clients are also said to play a vital role in demanding BIM. Clients are the actors empowered to specify demands and thereby demand BIM when procuring consultants and contractors (Wong et al. 2011, Khosrowshahi and Arayici 2012, Eadie et al. 2013). When clients formulate such demands, there is no longer the issue whether BIM is to be used or not, but more specifically how BIM should be used in the specific project (Linderoth 2010).

The arguments used to pinpoint the client as the most important change agent for BIM in the construction industry are therefore: potential benefits and their power to set demands when procuring consultants and contractors. However, when viewed as a process of translation, the initial force behind a change process is not considered to be more important than any other force in the change process (Latour 1984). Hence, for change to take place, associations and networks have to be built between all actors that are supposed to be influenced by the change process which requires alignment of the actors influenced by said directions (Latour 1984). Clients must establish networks of actors and try to align their agendas regarding BIM use for change to happen.

Mobilizing an actor-network

The processes in which an actor (or initiator) tries to influence others, as in the implementation of BIM, can according to Callon (1984) be understood as a process of translation. Translation describes how an actor tries to impose themselves and their understanding of a situation onto others (Callon 1984). Sociology of Translation builds upon Actor-Network theory (ANT) and its description of the actors partaking in a translation process. ANT is a well-used method or tool to understand organizational change following the implementation of information systems (Walsham 1997). This is done by the analysis of both human and non-human elements (actors). And by using ANT, it is possible to address the divide between the social and the technical in an implementation process (Tatnall and Gilding 2005). An actor-network perspective thus enables analysis of how different actors are influenced by the implementation of a new technology, and how their roles and relationships change in the actor-network. By viewing the implementation of BIM as the creation of an actor-network, it is possible to map activities taken in order to tie actors to the network as well as the actor’s needs and motives for being enrolled into the network (Linderoth 2010). In this analysis, technology represents itself as a non-human actor which in its involvement in the network influences all other actors in their link to the network.

In BIM-related research, it is commonly stated that BIM adoption will require substantial changes to both business and work practices, along with changes in the culture of how projects are conducted (Mihindu and Arayici 2008, Succar 2009). An analysis of how technology influences, changes and defines roles among the actors in the network is made possible by using an actor-network perspective (Linderoth 2010). It is not a given that actors in the network use the technology in the ways that were initially intended or that they adopt it at all. In order to understand the change process linked to BIM implementation, it is in this paper viewed as a process of translation, a transition where actions are taken to create an actor-network around BIM. Callon (1984) identifies four moments of translation in which the initiating actor tries to establish an actor-network to which other essential actors are tied. These steps are:

•	Problematisation – the process of defining a new idea, a problem or a new opportunity. In this stage, it becomes relevant to identify actors needed to succeed in the sought after change process, finding the indispensable actors. To enrol the actors, there is a need to understand their goals and needs. In this process, it becomes important to find one question, an obligatory passage point (OPP), which is in line with the motives of all identified actors. The identified actors need to be convinced that allying around this question will benefit them all.
•	Interessement – A number of actions by which the identified actors are linked together and given roles within the new actor-network. To interest actors is to isolate them from external actors not involved in the network who want to define their identities in competing ways.
•	Enrolment – a set of strategies orchestrated by the driving actor to make the indispensable actors accept the new roles within the actor-network. Enrolment constitutes negotiations, trials of strength and tricks to stabilize the network and the actor’s roles therein.
•	Mobilization of allies – The methods used to ensure that the spokes persons present in the earlier negotiations truly represent their respective actors and are not betrayed. Only a few individuals were involved in the earlier stages of the translation and in these individuals must ensure that the indispensable actors follow their representatives.

This model work as a method for mapping the process in which the driving actor tries to establish an actor-network, linking actors they deem indispensable and trying to ensure that their interpretation of the problem is accepted. By analysing the interplay between the actors within the actor-network and how these actors are linked to the network, a better understanding of what influences the translation process can be achieved.

The stability of the actor-network created in the process of translation is dependent on the ways in which this new network is questioned. The network becomes stable when no more issues are addressed and when there is no need to reconsider the network. Callon and Latour (1981) describe this process as the creation of Black Boxes. “A Black Box contains that which no longer needs to be reconsidered, those things whose contents has become a matter of indifference” (Callon and Latour 1981, p. 285). Black Boxes are locked network elements, not questioned by actors in the network. Such locked network elements can contain habits, forces, objects and models of thought. Black Boxes can also remain even when specific actors in the network are changed or leave the network (Callon and Latour 1981). Therefore, actors can be changed without the need for redefinitions of the network itself (Linderoth 2000).

In the creation of an actor-network the desired programme of actions can be inscribed into specific artefacts linked to the actor-network (Latour 1990). Inscriptions refer to how artefacts can be used to lead actors towards desired programs of use, guiding and restricting the use of said artefact so that the programme is followed (Monteiro 2000). In the case of technical artefacts, which BIM-related tools are examples of, the developer tries to define the potential user and thereby inscribes the requested programme of use into the artefact (Linderoth and Pellegrino 2005). Such inscriptions can, for example, consist of instruction manuals, requests, demands or the design of the artefact itself. The inscriptions are however not absolute and possess varying levels of strength and flexibility. Strong and inflexible inscriptions guide the use of the artefact while weak and flexible inscriptions open for less clear guidance (Linderoth 2000). The flexibility of inscriptions refer how much the inscription influence the patterns of use, where flexible inscriptions only have a limited influence inflexible dictate the patterns of use to a large degree (Hanseth and Monteiro 1997). The strength of an inscription, the degree it must be followed or can be avoided, depends on the level of irreversibility of the actor-network in which it is inscribed (Hanseth and Monteiro 1997).

Method

This paper builds upon a qualitative case study of the BIM implementation process at the largest infrastructure client in Sweden, the Swedish Transport Administration (STA). In this paper, actor-network theory (ANT) is used as a theory giving methodological advice and guidance for the gathering and analysis of empirical findings. This approach offers “a language to describe how, where and to what extent technology influences human behavior” (Monteiro 2000, p. 5). The focus has been on following the actors in their discussions during the implementation of BIM, here viewed as a process of translation (Callon 1984). A translation in which the BIM Initiative at the STA tries to tie together the projects conducted at this client organization with the construction industry and the BIM technology to establish an actor-network around BIM. Analysis of actor-networks also requires zooming in and out in order to identify new limited networks within the larger actor-network (Monteiro 2000). For example, the actor of the construction industry is addressed as a single actor in the translation process; this actor however consists of various sub-actors forming a network within construction projects.

The client that was studied is a large public organization, consisting of several departments and sub-units, each with different roles and responsibilities. In the translation of BIM, some of these departments and sub-units influenced the translation process, thereby establishing themselves as actors in the actor-network. The interactions and negotiations between the actors have been followed by observing meetings where actors have interacted and by interviewing individuals about their perceptions of the translation.

The case study has followed the actions taken in this BIM initiative from the end of 2013 until the end of 2016. The BIM implementation initiative conducted by this client has gone through several stages, each following at the completion of the one before. It started with a BIM Network, continuing with a BIM Initiation Project followed by a BIM Area of Expertise. The different stages the BIM initiative has taken varies in terms of organizational form. However, the general goal has been the same: to lead the translation of BIM at this organization. Therefore, this paper does not distinguish between the different phases, but rather refers to them all as the BIM Initiative.

As the main deliverable from the BIM Initiative, new BIM-related documents have been studied. The documents represent the main channel through which the BIM Initiative try to inscribe BIM use into the project process at the STA. Documents produced throughout the BIM implementation process have therefore been analysed to identify the inscriptions aimed at influencing actors linked to the BIM translation. In this analysis, particular focus has been placed on identifying new demands and guidelines influencing the management and procurement process in construction projects. The studied documents include: the STA BIM strategy, four versions of BIM-related guidance documents and three versions of updated procurement templates relating to BIM. Additionally, the document analysis includes a referral procedure where comments have been collected on an early version of a guidance document. This referral procedure comprises 314 comments from various departments at the STA.

A total of 11 semi-structured interviews have been conducted during this period of 2.5 years: 6 with BIM experts developing and driving the BIM implementation at STA, 3 with project managers at STA currently using BIM in their construction projects, and 2 with project managers managing the two different BIM implementation projects at project conducting departments: Investment and Large Projects. The interviews were between 1 and 2 h long. The interviewees were selected based on their involvement in the BIM implementation process and the development of BIM-related documents.

The interviews were conducted in two series. The first took place in 2014 with individuals having formal roles in the BIM Initiation Project and with individuals employed as project managers working with BIM in their respective pilot project. These interviews aimed at understanding the BIM implementation process and the formal strategies used by STA to support BIM implementation. The second series was conducted in early 2016 as BIM was being implemented on a wider front at STA. These interviews aimed at mapping different BIM actors’ interests, motives and agendas, and thereby enabling a better understanding of their interplay. In conjunction with the second interview series, a formal meeting with spokespersons from all internal STA actors related to the BIM implementation process was observed. The aim of the meeting was to compare the different actors’ views of BIM, and to coordinate their understanding in order to formulate formal demands of BIM models in procurement. This meeting was held in November of 2016.

The empirical material has been interpreted using the theoretical framework of Sociology of Translation as described by Callon (1984). Additionally, ANT centric concepts have been used to analyse the findings. The findings have also been related to previous research on BIM implementation with a focus on clients and their role in supporting BIM implementation.

Delimitations

The case study does not constitute a comprehensive ANT analysis of the BIM implementation process at STA; rather inspiration from ANT has supported the interpretation. Despite the minimalistic use of ANT, relevant findings have been collected and the study presents an understanding of the change process at STA in which BIM plays an important part.

The translation is still ongoing at STA, thus findings do not include any final accepted results of this process. The question: if and how BIM is used in construction projects at the STA, is not included in this paper. Instead, focus is on the translation process and how it has evolved.

BIM at the Swedish transport administration

The BIM Initiative was started at the STA by enthusiastic project managers in 2010. These project managers were inspired by the buzz around BIM and took the initiative to try BIM to various degrees in their construction projects. The project managers created what they called a BIM Network with the purpose of sharing experiences and develop their BIM skills. The BIM Network was an informal forum where project managers could exchange experiences. Project managers of some of the largest infrastructure projects in Sweden were involved in this network. At this time, in 2010, the STA went through a major reorganization. The STA was established in 2010 and was a merger of two prior organizations: the Swedish Road Administration and the Swedish Railway Administration. According to a project manager partaking in the BIM Network, the leadership was unclear during this reorganization. This situation provided project managers with the possibility to take action and start BIM implementation initiatives.

The Swedish government authorized a committee to analyse the public clients’ ability to increase productivity and innovation in the infrastructure and construction industry. This committee started their work in 2009 and resulted in a Swedish Government Official Report in 2012 (SOU 2012:39). This report explicitly suggests STA to work towards the implementation of BIM in the infrastructure sector. Representatives from the BIM Network were directly involved in writing this report. Based on the report, the General Director of STA made a formal decision to: “in a coordinated and controlled way implement BIM, Building Information Modelling/Model in the whole Swedish Transport Administration”. This decision was made in early 2013 and can be seen as a formal starting point for the implementation of BIM at STA.

Problematizing the BIM issue

The initial steps towards BIM implementation were taken by members in the BIM Network. In this network the main focus was to take advantage of the opportunities with the new technology. After the General Director decision, the BIM Initiative gained legitimacy. In the end of 2013, the BIM Network was reorganized into a BIM Initiation Project. In contrast to the BIM Network, this project had a project specification with goals, time schedule and budget. The project was given the task of paving way for BIM implementation at STA. Many of the members of the BIM Network continued working with this project. They participated either directly with the BIM Initiation Project or as project managers in pilot projects where BIM was used. Within the boundary of this project, a BIM strategy for STA was developed. In this document, BIM is described as: “the use of information models in a continuous flow through the main processes connected with a constructed facility”. It is also expressed that object-oriented information can be used for multiple purposes, such as: clash control, analysis of different design alternatives, cost calculations and time scheduling. When combined the benefits were expected to result in a more efficient construction project process.

In order to achieve the benefits outlined in the BIM strategy, the BIM Initiation Project formulated questions on how BIM should be implemented, addressing implementation barriers like interoperability and other technical aspects of adoption. Apart from those issues, the main question was how to support projects to use BIM by influencing them to adopt new possibilities.

Identification of actors

The BIM Initiation Project identified several actors vital for BIM implementation at STA. The actors are addressed in discussions and in documents linked to the BIM Initiative. The actors were: Construction projects conducted by the STA, The construction industry and the BIM technology. In the problematisation stage the actors identified to be linked to the BIM implementation initiative according to Figure 1.

•	Construction projects performed by STA – It is in construction projects that the actual changes have to be made. If BIM is going to be used, it is going to be used in construction projects. The projects at STA are not a homogenous. Instead they vary between small and medium projects at the Investment department and large and more complex projects at the Large Projects department. Despite the fact that projects at these two departments have different preconditions, their project managers are addressed as a single group in the BIM translation process.
•	The construction industry – Contractors, designers, consultants and other actors active in construction projects conducted by the STA. A prerequisite for using BIM at STA is new ways of delivering project related information. It is also the ambition of the BIM implementation to influence these actors to change towards a BIM centric work practice and thereby change the construction industry.
•	BIM technology – BIM technology was tested in several pilot projects, however it can be used differently. It has different requirements depending on how it is used. More advanced collaborative uses of BIM requires more advanced models and higher levels of interoperability. The level to which BIM should be used in projects will have to be addressed in the implementation process. Infrastructure has not been in focus in the development of BIM tools and currently there are problems in modelling elongated objects such as roads and railways. When it comes to standardization, infrastructure has not come as far as buildings. Especially road alignment has been identified as a problem in relation to developing models based on open formats.

Figure 1. BIM network, identified actors.

BIM as an obligatory passage point

The perceived low increase in productivity is the often referred to reason behind the interest for BIM. How BIM can address this problem through the new possibilities it provides is extensively described both in the Swedish government official report (SOU 2012:39) and in the STA BIM strategy. In this way, the BIM Initiative tries to make BIM an obligatory passage point (OPP) for how productivity can be increased. Indispensable actors who are not already convinced of BIM as the OPP for increasing productivity are forced or persuaded into the network during the translation. However, to influence the actors to be enrolled into the actor-network, the BIM strategy states the importance of evaluating the extent to which BIM should be used. In other words, know how the technology will influence work practices among actors in construction projects. Based on this understanding, corresponding demands on BIM can be inscribed into guidance documents influencing the actors deemed indispensable. With this strategy, BIM was supposed to be established as the OPP for how the industry’s productivity could be increased.

Establishing interest for BIM

After the start of the BIM Initiation Project, construction projects previously connected to the BIM Network were referred to as BIM pilot projects. In addition to these projects, several new BIM pilot projects were started. However, the level to which BIM were used is these new projects were very limited and when contacted the project managers their BIM use as “We have not yet started with BIM, therefore it has no direct influence on the project”. Two of the largest projects conducted by the STA had earlier been connected with the BIM Network and continued as pilot projects for BIM, and provided much material to the BIM initiative. Especially one of the projects, the Stockholm bypass, had major influence on the outcomes of the BIM Initiation Project. It was in this project that the first draft of the guidelines and the demand for BIM was developed. These documents would later be revised and implemented in the whole organization.

When establishing the BIM Initiation Project, the understanding of BIM expressed in the BIM Network became formalized and accepted as a general view of BIM at the STA. In this way the issue of BIM was not only relevant for a small group in the BIM Network but rather as something that was relevant for the whole organization. The BIM issue was also presented to the different sub-units in the organization on a series of so called “BIM days”. In this process the BIM strategy was presented to other parts of the organization and linked to possibilities with the new technology. In several presentations given by the BIM Initiation Project the technology was described as a natural development and a certainty of the construction industry of tomorrow. The BIM maturity stages were prominent, showing the expected progression towards more advanced BIM use.

Enrolling actors

After major rework following a referral process, 32 documents were approved and implemented in the management system by mid-2015. Among the documents were the demands on Object Oriented Information Model (OIM) which describes how object oriented information should be produced and shared within projects. The procurement templates at this client organization refers directly to this document, making demands for BIM a natural part of all newly procured projects. As the main deliverable, the documents make up the central part in the enrolment of the actors identified in the problematisation stage. It is in these documents that the specific use of BIM, as defined by the BIM initiative, is inscribed into the project process. However, in order for the documents to be accepted, some sections were deleted and others changed as a result of substantial critical feedback in the referral procedure. The project manager for the BIM implementation at the Investment department stated that “We were unable to procure projects using the specification in the first versions of the BIM demands documents, the demands stated there were not in line with other guidelines on how demands should be specified in procurements”. The changes made in response to this feedback limited demands on the use of models, instead focusing on the model’s delivery. By the implementation of the new guidance and demand documents all the actors identified by the BIM initiative were addressed, see Figure 2:

•	Construction projects conducted by STA – By making changes to guidance documents and procurement templates, object oriented models (BIM models) were introduced as the preferred choice of medium for project related information. When accepted, the documents specify that deliverables from contractors and consultants should be handed over in coordinated models instead of traditional drawings. How models should be produced and managed is also specified in the documents. A majority of changes to existing documents were however only semantic, i.e. a change from “drawing” to “model”. This change was made in order for STA to be consistent in the demand of models. In this way the BIM initiative tried to inscribe the establishment of demand for BIM models among project managers procuring projects at this organization. However, the new guidance documents do not describe how the internal projects process could make use of models and how internal work practices can be changed. Information regarding how models should be handed over to the facilities management department after the project’s completion was also lacking in the new documents. Instead, the focus in this process is external, i.e. inscribing demand for models among the industry actors.
•	The construction industry – This actor, which is not homogenous but rather a multitude of actors, is enrolled by the new demand for BIM as referred to in the new procurement templates. Details of how the procured project participants should use BIM are not defined. This might be in conflict with other STA initiatives and was therefore left out. Instead, the BIM initiative inscribed a demand for coordinated models which should be delivered by the main contractor. That is to say, models containing information from all disciplines should be handed over to the STA at the end of the construction project. This demand is aimed at providing the main contractor with incentive to work towards collaborative BIM use among project participants. Collaborative BIM use would make the creation of coordinated models easier, limit the need for the main contractor to create them after the project’s completion.
•	The BIM technology – How models should be produced and managed was identified as a major issue in the problematisation stage of the translation. Problems with interoperability and other software issues were identified as problematic in the pilot projects. To address this issue, the guidance documents refer to .dwg/.dgn file formats. These file formats are far from the object-oriented information initially aimed for, however such object orientation could not at the time be achieved in the infrastructure sector.

Figure 2. BIM actor-network.

Competing networks

In parallel with the BIM implementation, STA worked with an initiative to become a “Professional Client Organization”. This initiative strives to focus on STA’s client role, not concentrating on details in construction projects but rather emphasizing on the procurement of projects, allowing contractors and consultants to be more innovative when developing design and construction solutions. This way the STA aims to establish an environment suitable for supplier led innovation. As a part of this initiative the STA limit requirements on specific work practices and instead only specify the function of produced infrastructure. The work with these two change initiatives, Professional Client Organization and BIM implementation, occasionally became contradictory. BIM-usage is generally viewed as new work practices, focusing on collaboration and the joint development and use of models. Because of this possible contradiction the General Director’s decision to adopt BIM specifically states that the BIM implementation should be in line with the Professional Client Organization concept. How these two initiatives have interacted has been described as problematic by individuals in the BIM Initiative. Hence, policies from both initiatives influence construction projects conducted by STA and therefore the Actor-Network around BIM see Figure 2.

During the referral procedure of the new guidance documents, representatives from the BIM Initiative expressed how demands on the use of BIM are essential to influence the industry to change towards a more collaborative BIM use. However, several referral responses from individuals from the Investment department voiced concerns with these demands, stating how they were not in line with other policies at the STA. The discussion of how to prioritize between the two initiatives was seen as a reason to why the referral period took longer than first expected. According to the individual responsible for collecting referral comments, conflicts existed even after changes had been made as per suggestions given in the referral.

Mobilization of allies

When the BIM Initiation Project was finished in late 2014, two local BIM implementation projects started, one at the Investment department, and one at the Large Projects department. The Investment department handles the main bulk of projects, everything from small improvement projects to larger road or railway construction. This department works relatively standardized with projects managers conducting multiple projects simultaneously. The Large Projects department handles the largest and most complicated construction projects at STA. In contrast to Investment, projects at this department are managed more independently and have larger internal support structures. These two BIM implementation projects were subjected to continue the translation, and implement the new guidance documents into their respective departments, i.e. to make BIM use happen in practice. When implemented, the new documents were supposed to establish a high degree of irreversibility in regard to the demand for BIM in projects, thereby mobilizing construction project managers.

The BIM implementation project at the Investment department was given tight guidelines in regard to the implementation of BIM. The guidelines stated that BIM implementation was not allowed to add any extra costs to projects: no new roles were allowed which would expand the internal project organization. It was further specified that the BIM implementation had to be in line with the Professional Client Organization initiative. Implementing BIM and simultaneously following these guidelines became problematic. In order for the BIM documents to be accepted at the Investment department large changes had to be made, the project manager for the BIM implementation at this department described it as: “The comments we collected in the referral was very critical, the view were mostly that we were unable to implement the documents and they had to be scrapped and redone”. Based on the referral comments and in order to make the guidance documents in line with other directives only parts of the documents were actually implemented; additionally both procurement templates and guidance documents had to be modified. The template for mission statements in the procurement process was changed in June 2015 (UB-mall 7.0). This document refers to parts of the new BIM guidance documents and specifies how project information should be produced, exchanged and delivered within projects. This date can therefore be seen as the date from which all newly procured projects at the Investment department used BIM. However, in practice this change had little impact.

The BIM implementation process at the Large Projects department was different. Construction projects at this department have extensive support structures and are managed in a relatively autonomous way when compared to projects at the Investment department. This made BIM use more diverse. Some projects had previously acted as BIM pilot projects for the BIM Initiation Project while others had not. Pilot projects served as role models for BIM use and they actively tried to influence project participants to adopt BIM-supported work practices. Generally, projects at Large Projects are managed based on their specific circumstances and therefore have a more flexible view of change initiatives such as BIM or Professional Client Organization.

To support these two implementation projects, the BIM Initiation Project was reorganized into a BIM Area of Expertise. This was a new organizational context but many of the individuals working in the BIM Initiation Project continued and therefore established a link from the earlier stages in the BIM Initiative at the STA. This BIM Area of Expertise was tasked with managing BIM related documents; continue the BIM development and further supporting BIM implementation at the Investment and Large Project departments.

Discussion

The story of BIM implementation at STA is a story of how an informal network of early BIM adopters and other BIM enthusiasts translate their understanding of the technology to create an Actor-Network around BIM. It was the project managers who tried BIM first and started the BIM initiative, thereby initiating the BIM translation. As the change process continued, this translation has been developed in various ways. The General Director for the STA made the formal decision to implement BIM in the whole organization; this decision was however heavily influenced by the project managers in the BIM initiative.

The understanding of BIM, as expressed by the project managers in the early BIM initiative, is well in line with earlier research on the clients’ role in relation to BIM implementation. The view among the members of the BIM Initiative is that of clients’ as change agents due to their power to demand BIM when procuring consultants and contractors but this view is also influenced by how clients are expected to benefit from this change (compare with Porwal and Hewage 2013). The belief that actors’ acceptance of BIM is dependent on clients’ demands for BIM use is dominating in the BIM Initiative. In line with previous research by Linderoth (2010), focus is expected to shift from whether or not to use BIM to how BIM should be used when client’s demand for BIM is established. This problematisation of BIM directly influenced how the BIM Initiative identified the indispensable actors for BIM: Construction projects conducted at the STA, the construction industry and the BIM technology. This normative understanding of BIM also influenced how the actors were approached in order to enrol them into the Actor-Network.

In the efforts to enrol the identified actors, both internally and externally, changes were made to the guidance and demand documents at the STA. By implementing new guidance documents, a formal demand for BIM was inscribed into both the project management process and procurement processes. The details in the demands for BIM were however limited by the Professional Client Initiative. This resulted in a situation where the demands on how models should be produced and managed were in focus, not how they should be used in practice. Consequently, only an open demand for BIM was inscribed into the procurement templates and this was viewed as a limitation among many individuals in the BIM Initiative. This strategy tried to increase the strength of the inscriptions as they were more in line with other competing initiatives, thus making them harder to ignore. However, as the inscriptions were very flexible, project managers were given large potential to interpret them freely, making the inscriptions flexible.

By trying to establish BIM as something given, and by focusing on the technical aspects of the technology, the translation process tried to steer the whole issue of BIM into a Black Box. Black Boxes, as described by Callon and Latour (1981) are network elements that no longer need to be reconsidered. This way, project managers at the Investment department and Large Project department would not question if BIM should be demanded or not. Instead focus would be placed on how the use of BIM models could be supported and developed. By Black Boxing BIM the BIM Initiative tried to give the change process a large degree of irreversibility.

As described by Succar (2009), the major changes to work practices following the implementation of BIM need to take place between project participants. Therefore to implement BIM is to influence work practices and relationships among and between industry actors. Therefore, there seems to be a conflict between the two change initiatives: BIM Implementation and Professional Client Organization, as the second limits the degree to which the client should influence the work practices. The BIM Initiative strived to distance itself from this conflict, trying to limit expressed demands for particular work practices. However, instead of fully adapting the BIM translation to the Professional Client Organization initiative, actors were influenced to prioritize BIM issues as disregard other competing initiatives during the BIM translation. This was done by arguing for the importance of BIM and the need to influence external project participants in their use of BIM to ensure the desired results. This is a concreate example of how the BIM Initiative tried to interest actors to the BIM Actor-Network and as Callon (1984, p. 9) states it: “build devices which can be placed between them and all other entities who want to define their identities otherwise”.

The conflict between the BIM Initiative and the Professional Client Organization resulted in an opportunity for construction project managers to pick and choose between the two initiatives. This substantially decreased the strength of the inscriptions in the BIM demand documents. How project managers perceived the strength of the inscriptions in the Black Box depended on how well they were enrolled into the BIM Actor-Network. The Large Projects department was more enrolled into the BIM Actor-Network when compared to the Investment department. Thereby the Large Projects department perceived the inscriptions as strong. However, as projects at this department works relatively autonomously, they tend to view inscriptions in guidance documents as flexible and adapt them to suit their specific construction projects. The Investment department, on the other hand, rejected many of the inscriptions in the Black Box. Instead this department modified the guidance documents and only used limited parts of the inscriptions. As this department has a more standardized project process inscriptions in guidance documents are viewed as inflexible. Altogether, this has resulted in a situation where construction projects who generally view inscriptions as flexible has accepted them, while those who view inscriptions as inflexible has not. Thereby there is currently no standardized way BIM is used by STA and project participants receive different guidelines in different projects.

When the implementation of BIM is viewed as a process of translation it becomes clear that having power to initiate a change process is not enough to ensure its success. In line with Latour (1984) this case shows how all actors involved in a translation influences its outcome. When one actor takes over the change process and continues the translation they will adapt it based on their own preferences. To assure obedience to the BIM programme within the network requires alignment of all actors’ agendas (Latour 1984). Change towards BIM thereby requires all actors to be enrolled into the BIM network. Thus, the normative claims of a natural progression through sequential BIM maturity stages (e.g. Succar 2009) following the client organizations formulation of demands for BIM (e.g. Porwal and Hewage 2013) can be question.

Conclusion

Clients are frequently pinpointed as important change agents, needed to achieve inter-organizational change in the construction industry. In this role, clients are expected to use their power and influence to demand change among project participants in projects they conduct. In this paper a case study of such a change process has been presented. An empirical example of how a large public client within infrastructure implements BIM to both change both its own ways of working, and at the same time influence the construction industry. The main strategy for enrolling actors into the BIM Actor-Network has been to demand models instead of traditional drawings. The use of models was thereby inscribed into the project management and procurement processes, trying to establish the BIM-issue as a Black Box. However, Due to other competing policy initiatives the translation has been problematic. In order to make the Black Box’s inscriptions stronger, only limited demands on the use of the models were possible. Hence, this strategy only inscribed an open demand for BIM models, not how BIM should be used. Simultaneously, the BIM technology is not as developed for infrastructure projects as for other types of construction which have also influenced the inscribed demands. The project managers at the client who were supposed to be influenced by the inscriptions also had several conflicting directives to take into consideration. This conflict made their enrolment into the BIM Actor-Network unsure.

Implementing an innovation into the loosely coupled (Dubois and Gadde 2001) construction industry is not a new phenomenon; it has been studied for decades. Except for some recent studies taking more critical and organizational perspectives, this knowledge is seldom related to the challenges faced when implementing BIM. Instead, BIM is still mainly studied using rational perspectives.

This case study has found similarities between the implementation of BIM and the implementation of other innovations to the construction industry. The way of inscribing BIM use found in this case study disregards the characteristics of relative boundedness (Harty 2008) often found in the construction industry. In order for inscriptions to be influential they need to have enough strength to be accepted and enough inflexibility to be followed as intended. An industry characterized by relative boundedness presents even higher demands on the strength and inflexibility of inscriptions. In the case study presented in this paper, several problems were found in establishing BIM as a Black Box with strong inflexible inscriptions. The view of clients as the single most essential change agent for BIM-adoption can thereby be questioned. The Actor-Network around the BIM issue includes a multitude of actors who need to be enrolled in order to achieve collaborative BIM use. Collaborative BIM use impacts many aspects of the project and procurement processes and thereby interferes with many programs of actions among the actor within a construction project. In line with Latour (1984) this paper shows the importance of aligning actor’s different agendas in the network in which change is sought. The understanding of BIM implementation as a process of translation also highlights how all actors in the Actor-Network will influence the change process.

To conclude, this paper shows how the view of clients as change agents within the construction industry disregard the knowledge of translations processes and the characteristics of relative boundedness in the industry. Instead it can be argued that the role of the client is more complex and how the strategy of Black Boxing BIM in this case was too limited.

Disclosure statement

No potential conflict of interest was reported by the author.