This chapter presents the background of AECO digitization, starting with general tendencies and moving on to particular developments in AECO, including BIM. It explains these developments from a historical perspective and outlines the limitations they cause to further digitization and decision making in AECO. 

Private versus business

While in our private lives we are quite digitally minded and data savvy, there is little to suggest that digitization similarly dominates professional activities in AECO. Despite the enthusiastic reception of technological developments, such as 3D printing, digitization has yet to reach a substantial depth or breadth in AECO. We use computer programs like BIM and CAD to draw or spreadsheets to calculate but reality in AECO remains analogue, dominated by information carriers like drawings and other conventional documents on paper: remnants of an era when we did not have the same information processing capacities as today. This is unlike e.g. the music industry, where vinyl, CD and other carriers are just a matter of nostalgia, while the content has become fully digital, or online on-demand services like Netflix or Spotify, which have moreover changed digital attitudes in spectacular ways, practically eliminating music and video piracy.

The probable reason is that AECO generally remains attached to analogue, largely pre-industrial processes that require little if any mediation from digital technologies — much like fishing and hunting, two other industries with a low investment in digitization. These processes cause legacy information solutions, such as paper-based documents, to persist, severely limiting the potential and nature of digitization. Resisting or even ejecting digitization is, of course, justified if there is no reason for it. Regrettably, this is not the case with AECO, given its far from satisfactory performance. It follows that the high contrast with other industries or even private life calls for a closer investigation of the particular circumstances of AECO, towards a clearer identification of underlying causes and resulting problems.

Digital uptake

There is broad consensus that AECO is one of the least digitized sectors.^[1] Everyone seems to be in agreement: on the Internet, in professional and academic publications, in software advertisements. A critical note is that the claim is based on few data, chiefly proxies, and a lot of opinions of people in AECO or digitization, i.e. with vested interests in the deployment of new technologies. Still, the slow digital uptake in AECO seems so plausible that it is widely used as justification for various digital solutions: manifestos by policy makers, standards by professional bodies, new approaches by academic researchers, new software by commercial developers. So, from a vague problem, we jump directly to specific solutions, such as BIM, digital twins, Industry 4.0 etc.: panaceas for all the ills of AECO. The promise of the solutions is invariably deemed so high that the resulting changes in AECO do not just solve the problem; they make it disappear completely.

This poses an interesting conundrum: if the solutions are so readily available and so powerful, there must be at least a significant minority in AECO that adopts them and benefits from observable and convincing improvements in performance. In turn, this should stimulate wider adoption of the solutions in AECO and general advances. In short, things should develop rapidly and smoothly, changing practices and behaviours, as we can see with most digital technologies, from email to satellite navigation. This, however, does not seem to be the case with digitization in AECO. Even CAD and BIM have always been considered primarily with respect to costs and obstacles. This suggests that most of these solutions have little overall effect on the problems of AECO or that they fail to fully utilize the potential of digitization.

The viewpoint advocated in this book is that most solutions do hold some promise for solving real problems in AECO. However, instead of jumping ahead and imposing any solution willy-nilly, we need first to understand the relation between problems and solutions: describe and explain it, so that we can judge if a solution is suitable and feasible. This calls for a closer, more detailed inspection of digitization in AECO and its background, which reveals that more than from slow uptake, digitization in AECO suffers from having a secondary role. Even if investment is low in comparison to other sectors, digitization is clearly present in AECO: drawings are already made with CAD or, increasingly, BIM, while office automation is complete and there are enough crossovers between the two, such as invoicing software that draws data from CAD or BIM. In fact, between 1997 and 2015 investment in digitization among German AECO enterprises more than doubled.

Presence, however, is not enough because digitization remains too far in the background of AECO decision and production processes. Digital technologies are mostly found at the office, where they used to produce conventional analogue documents, for use in outdated decision processes and arguably more significantly in largely manual production processes: building construction still relies more on cheap labour than on digital means, such as productive robotization. AECO appears to have limited investment to basic digitization, such as CAD and electronic invoicing. More advanced and domain-specific technologies, from 3D scanning to robotics, are rare, despite their acknowledged potential for competitiveness, innovation and productivity. The reason for that may be that there is little incentive in advanced technologies that are unrelated or conflicting with current practices: why invest in 3D-scanning precision if the tolerances in building construction remain high? This affects even basic digitization, such as CAD and BIM: why invest in well-structured, precise models if the sole purpose of the software is to produce drawings on paper? It is enough that these drawings look correct.

INFORMATION EXPLOSION IN AECO

Despite the slow, limited uptake of digital technologies, there is ample evidence of the explosive growth of digital information in AECO. On one end of the spectrum, we have new information sources that produce big data, such as smartphones and sensors. These tell us a lot about users and conditions in the built environment, and so promise a huge potential for the analysis and improvement of building performance, but also require substantial investment in technologies and organization. Predictably, there is limited interest for this end, despite the appeal of subjects like prop-tech and smart buildings.

At the other end of the spectrum, we encounter general-purpose technologies (basic digitization) that have already become commonplace and ubiquitous, hence also in AECO. Office automation has taken over the production and dissemination of memos, reports, calculations and presentations. Email, for instance, dominates communication and information exchange by offering a digital equivalent to analogue practices like letter writing. A main characteristic of these technologies is the replication of fragmented analogue practices , to the detriment of integrated, domain-specific technologies. For example, communicating on issues in a BIM-based project via email and reports produced with text processors and spreadsheets is redundant because most BIM software includes facilities for reporting issues and making calculations in direct connection with the model.

Domain-specific technologies, which attempt to structure AECO processes and knowledge, exist in the diffuse zone between the two ends of the spectrum. These try to offer more relevant alternatives to general-purpose technologies, as well as connections to the abundance of digital data. Currently paramount among them is BIM, an integrated approach that is usually justified with respect to performance.^[2] Performance improvement through BIM requires intensive and extensive collaboration, which adds to both the importance and the burden of information. Integration in BIM and return on investment also require coverage of most aspects of a project and put emphasis on larger projects. Both comprehensive digitization and larger projects, however, come against interoperability, capacity and coordination problems, making BIM deployment even harder and often haphazard.

The end result is that AECO still resides in the mentality of information overload. In a 2015 survey,^[3] 70% of AECO professionals claim that project information deluge actually impedes effective collaboration, while 42% feel unable to integrate new digital tools in their organizations. We have no reason to assume that the problems have been alleviated since then. As information needs in AECO have changed little since the 1980s, when digitization was in its infancy, this suggests that the problem lies primarily not with the unchanged quantities of information but with the way information is accessed through the new, digital means. Therefore, the resulting dissatisfaction with digitization cannot be dismissed as a teething issue. If digitization approaches in AECO were successful, any such issue would have been resolved long ago. Its persistence suggests fundamental misunderstandings that impede the deployment of real solutions to AECO information needs. AECO consequently appears to share many of the problems of the digital information explosion without enjoying adequate benefits from the information-processing opportunities of the digital era.

ORIGINS AND OUTCOMES

To identify and explain these misunderstandings, we have to go back in history and look at the origins of AECO digitization. AECO has always been an intensive producer and consumer of information. In fact, most of its disciplines produce information on buildings rather than buildings, primarily documents that specify what should be constructed and how. Especially drawings have been a major commodity in AECO, both as a widely accessible isomorphic representation of buildings and as a basis for conceptualizing designs through geometry. Throughout the history of AECO, drawings have been ubiquitous in all forms of specification and communication, as well as quite effective in supporting all kinds of decision making.

The history of digitization in AECO starts quite early, already in the 1960s, but with disparate ambitions. Some researchers were interested in automating design (even to the extent of replacing human designers with computers), while others were keen to computerize drawing. In the end, the two ambitions coexisted in the scientific area of CAAD, where design automation was generally treated as the real goal. 3D modelling was acceptable, especially if directly linked to design processes, while computerized drawing was largely left to software companies. With the popularization of computers in the 1990s, however, it was computerized drawing (CAD) that dominated AECO digitization in practice.

As with other software, the original use of CAD was the production of analogue documents: conventional drawings like floor plans and bills of materials on paper. For many years, the advantages of computerized drawing were presented in terms of efficiency improvement over drawing by hand on paper: faster production of drawings, easier modification and compact storage. Even after the popularization of the Internet, the emphasis on conventional documents remained. The only difference was that, rather than working with paper-based documents only, one could also produce and exchange digital files like PDFs.

In this manner, AECO information remained firmly entrenched in conventional, document-based practices. While analogue documents like telephone directories were being replaced by online information systems and people adapted to having their day planners and address lists on mobile phones or using navigation apps instead of maps, AECO stubbornly stuck to analogue practices and documents, prolonging their life into the digital era. This is evident even in BIM, which has stronger relations to design automation than drawing computerization but still retains drawings not only as the main output but also as the primary interface with the information contained in a model.

A further consequence is that the digital AECO information comes in huge amounts, with many and often large files that are poorly connected to each other. The content of these files is accessible through separate, usually proprietary software (as opposed to e.g. browsers that can access all information on the Internet) and involves human interaction and interpretation. The user remains the centre as well as the main actor in information processing, which further increases the number of documents, as users tend to summarize and combine sources. This reveals the biggest problems of this file-inundated information landscape: more than the amounts of information, file sizes and inefficient software, they are redundancy (multiple files covering the same subjects with considerable overlaps), lack of coherence (poor conceptual and operational connections between these files) and low consistency (different descriptions of the same aspects in various files and different descriptions of related aspects).

BIM: RADICAL INTENTIONS

The latest big chapter in the history of AECO digitization concerns BIM. Drawing from product modelling, BIM emerged as a radical improvement of computerized drawing that could provide a closer relation to design. The difference with earlier attempts at design automation was that it did not offer prescriptive means for generating a design but descriptive support to designing: structured representation of buildings, collaboration between AECO disciplines, integration of aspects and smooth transition between phases. By doing so, it shifted attention from drawings to the information they contained. At least, this is the popular perception of BIM. Behind it, lies something more fundamental that forms a recurring theme in this book: meaningful symbolic representation.

The wide acceptance of BIM is unprecedented in AECO computerization. Earlier attempts were often met with reluctance, not in the least for the cost of hardware, software and training they required. By contrast, the reception of BIM was much more positive, even though BIM is more demanding than its predecessors in terms of cost (an issue that nevertheless resurfaced after the initial euphoria). Arguably more than its attention to information or collaboration, it was its apparent simplicity (a Lego-like assembly of a building) that made BIM appealing, especially to non-technical stakeholders. The arcane conventions and practices of analogue drawing no longer seemed necessary or relevant.

Still, BIM remained rooted in these conventions. It may have moved from the graphic to the symbolic but it did so through interfaces laden with graphic conventions. For example, entering a wall in BIM is normally done in a floor plan projection, in a fashion that largely replicates analogue drawing: the user selects the wall type and then draws a line to indicate its axis. As soon as the axis is drawn, the wall symbol appears fully detailed according to the wall type that has been chosen: lines, hatches and other graphic elements indicating the wall materials. The axis is not among the normally visible graphic elements. Such attachment to convention impedes users from understanding that they are actually entering a symbol in the model rather than generating a drawing.

More on such matters follows later in the book. For the moment, it suffices to note that BIM signifies a step forward in AECO digitization but remains a transitional technology that may confuse or obscure fundamental information issues. Even so, as the currently best option for AECO, it deserves particular attention and therefore constitutes the main information environment in this book: representation and IM are discussed in the framework of BIM. Future technologies are expected to follow the symbolic character of BIM, so any strategies developed with respect to BIM will probably remain applicable. It is telling that current proposals on digital twins (representations that capture not only the form and structure of buildings but also their behaviour, as reported in real time by sensors in the real thing) generally depart from BIM-like models.

Limitations and necessities

The current digitization tendencies in AECO are dangerously confusing. While digitization invites us to interpret and even experience the world as information, AECO is still entrenched in analogue practices that keep information implicit. This means that we miss the opportunity to develop new conceptual models of reality, which are a prerequisite to digitization and information processing by machines. Instead, we use the old and arguably outdated analogue practices as the domain of discourse (the stuff that should be digitized).

Equally limiting is that digitization in AECO still calls for human interpretation, which runs contrary to the general tendency to remove ourselves from the centre of the information world. As a result, the explosively increasing amounts of digital information become a burden rather than an opportunity: we still focus on the availability of information for human consumption instead of on the information-processing capacities of machines that can support us in reliable, meaningful ways.

Even worse, the very availability of information may be underplayed. While digitization in general makes increasingly difficult to claim ignorance of anything, in AECO a project can be an isolated microworld that fails to acknowledge what exists beyond its scope. Learning and generalizing from precedents remains unsupported by AECO information technologies but even within a project many silos persist. The brief and budget, for example, are practically never integrated in the setup of a model in BIM, thereby leaving powerful options for design guidance and automation severely underutilized.

Such limitations do not merely affect IM; they also undermine decision making. As we shall see in the chapter on decisions and information, there is strong evidence that human thinking comprises two kinds of processes. The first kind (Type 1) is fast, automatic, effortless and nonconscious, while the second (Type 2) is slow, effortful, conscious and controlled. Type 1 thinking dominates daily life and allows us to be quite efficient in many common tasks but it also regularly leads to errors, especially in complex tasks. Regrettably, we tend to rely too much on the economical Type 1 processes and accept their products, even in situations that clearly call for Type 2 thinking. For example, we tend to make judgements on the basis of the limited information available in our memory at a given moment (e.g. news stories of the past few weeks), instead of taking the trouble to collect all relevant data and analyse them properly before reaching a decision.

This type of thinking occurs only too frequently with respect to the built environment: we become concerned about fire safety only after a publicized disaster and then go into a frenzy of activity that nevertheless soon subsides, especially if there is no similar disaster to rekindle our interest or if a disaster of a different kind occurs, even though the probability and risks of building fires remain the same. Moreover, we do not exhibit the same concern about stair safety, despite the fact that annually there are more victims of stair falls than of building fires, probably because each stair fall usually involves only one person, while a single building fire can have tens of victims.

That such problems are not restricted to AECO is not a consolation but a further danger: studies of human decision making reveal that people take decisions intuitively, on the basis of readily available rather than necessary, well-structured information, even in sensitive, high-risk and high-gain areas like finance. Share trading, for instance, is usually presented as a highly skilled business but performance is not consistent: it seems more a game of luck than one of skill. It is therefore important to take such failures into account also when we try to learn from other areas, especially with respect to management.

In addition to acknowledging and controlling our biases, so as to use Type 2 processes more frequently and purposely, we must take care that we always have access to the right information for these processes. This information, structured in transparent and operational descriptions of a task and its context, is the real goal for digitization in any AECO project: it returns human-computer partnerships, where machines support human decision making through extensive data collection, analysis and representation. Note that this does not imply a lessening role for humans in decision making. On the contrary, it adds to the capacities of humans by facilitating Type 2 thinking through explicit information, as well as by freeing resources for Type 2 processes.

The general conclusion is that AECO digitization is in urgent need of substantial improvement but this improvement is not merely a matter of importing new technologies as panaceas. The prerequisite to any change is a thorough understanding of building information and how it relates to our cognitive and social processes. As we shall see in the following chapters, once this is achieved, all goals, including IM and decision support, become clear and fundamentally feasible.