Behind the Building Information Model Buzz
Last week I was invited to attend a meeting of a few local architects where they discussed the Building Information Model and its relationship to documentation. Before attending I prepared the following document which I distributed amoungst the attendents. In it I aimed to clarify what BIM is (or more importantly what is isn't) and give them an indication of the issues surrounding the concept of BIM from a macro-perspective. Most attendants were very familar with the power of BIM tools such as Autodesk Revit but in general there was a lot of confusion between BIM and parametric modeling. There were also mixed feelings on the capabilities of BIM as a collaboration platform. For the practices involved it had helped internal processes but it was acknowledged that getting disparate information from design team participants into a BIM was a challenging task. Interoperability between various BIM platforms was an issue but there was the greater factor that many people working within the industry still produce relatively 'dumb' (i.e. basic 2D/3D drawings) that need to be practically recreated in BIM if it is to be of use within the model.
What follows is a web version of the printed document I distributed for reference:
Behind the Building Information Model Buzz
The 'Building Information Model' (BIM) is a marketing buzzword that has been heavily promoted by many Architecture, Engineering and Construction (AEC) software vendors. But what exactly is it, how will it effect the way you work and will it solve all of your problems?
Hold on, what is a Building Information Model before I buy one?
The term Building Information Model originated at Autodesk and was unofficially adopted by the rest of the CAD industry after it was 'blessed' in a column entitled “Comparing Pommes and Naranjas” by influential AEC technologist Jerry Laiserin.
BIM as a black hole: A moment of potential
infinite information density within a project
Conceptually a black hole as a useful metaphor for envisaging a Building Information Model. It is a single point within an architectural project where data is (potentially) infinitely dense. The architect sits on the event horizon of this singularity and is able to take in everything at once, so long as what they maybe interested in has been consumed by the BIM. Take note though that the black hole metaphor is used as a metaphor for mass density and not in a foreboding, destructive sense. However parallels can be drawn to the inability of matter to escape a black hole and the similar difficulties faced when allowing different parties to simultaneously modify the same BIM.
The academic definition of what constitutes a Building Information Model however is not precisely defined. The general consensus is that BIM is a single digital database holding all relevant building information, i.e. 2D, 3D and nD (time) data. BIM should not be confused with parametric modelling which can be difficult as the two often coexist together within software products such as Revit. BIM is not necessarily a 3D model or a defined manner of working, at its heart it is a conceptual data structure with an indeterminate number of ways of interrogating and working with the data contained within.
Is this a crazy new idea or a variation on an old theme?
BIM draws its roots from the digital product model concepts within the aerospace, shipbuilding and mass production industries that evolved during the 1970's and 80's.
With the rise of the computer within architecture practice a central focus of AEC software vendors was on the digitisation of conventional methods rather than the creation of completely new ways of working. This ensured a smoother adoption curve but minimised the capability of software to take full advantage of the rich, 3-dimensional representations created by architects.
Product models are highly defined and have
their own visual language called STEP Express-G
AEC researchers identified technologies associated with productised industries (such as aircraft and ships) as being beneficial to the productivity of AEC practitioners. As these industries have relatively closed development cycles, (i.e. one or two large companies collaborating on a limited number of products) software vendors had been able to craft new tools which allowed designers to build inherently intelligent digital models.
As these 'smarter' models facilitated time and financial savings in the architectural process pioneering CAD vendors like Graphisoft (ArchiCAD) began to create the first intelligent building models. Unfortunately for software vendors moving the product model concept into architecture has not been simple because unlike aerospace industry each architectural design is for a specific environment and realised by a team of loosely joined professionals.
Where is BIM today?Currently the complete vision of BIM is yet to be fully realised. The BIM ideal is that the digital model becomes the focus of all documentation for the life of the building. Presently BIM is only now making inroads into architectural design and documentation aspects of the process and has yet to evolve into a set of universal standards for interacting with architectural data. Given the massive investment in conventional CAD tools and entrenched working practices the move to a BIM manner of working has been very slow. However there are more factors at work in this slow adoption than the natural intertia of an industry way of new ways of working.
Barriers to this vision
Time is not the only limiting factor in the adoption of BIM throughout the AEC industry. Currently the industry fragmented by a host of vastly different professional interests (architects, engineers, contractors, quantity surveyors, etc.) and powerful commercial interests formed by the various technology vendors.
An overview of the IFC 2x2 data schema:
Whilst IFCs now cover many facets of the
AEC domain they do so at the expense of
software complexity (click to enlarge)
Accommodating the varied professional demands of the industry within a single digital representation is very demanding. In the early 1990's the International Alliance for Interoperability was established to develop a universal digital model for architectural projects. After fifteen years of continued development their Industry Foundation Classes (IFCs) are yet encompass the informational demands of the AEC domain. The cause of this delay is the human inability to universally agree on complex semantic terms. This is confounded by the fact many architectural terms inherently convey rich cultural histories through subtle physical and tactile differences.
Compounding this problem is the strong commercial divisions which are driven by financial and corporate motives. Large software vendors like Microsoft, Autodesk and Bentley profit most when they can create stacks of integrated products that hinder integration with third parties (data silos). The most recent example of a data silo is Autodesk's TrustedDWG standard introduced in AutoCAD 2007.
So where does the future lie for BIM?
Architectural design projects, like nebulas
come in all sizes and have many points of intensity
Given the fragmented nature of the architecture process, the degree of process change required for BIM introduction and the inconsistent level of technology uptake across the entire industry, it will be difficult for BIM to become the predominant form of architectural information storage as we move into the 21st Century. In fact it is highly likely that whilst BIM will continue to build on its role as a central information repository for the architect it will still play a supporting role alongside traditional documentation, numerous digital models (used for simulation, fabrication and presentation) and messaging services such as email, formal meetings and documentation. Consequently the information behind an architectural project may never take on the pure metaphor of a black hole but rather a nebula, open ended, decentralised and full of interesting things that detract from other interesting things, including the odd black hole.
If such a metaphor is more appropriate for the building design process then what is most significance is how design team members navigate the varying brief requirements, architectural intentions and pragmatic constraints to reach the decisions they made. The artefacts of these decision making processes are the digital models, drawings and documented exchanges generated during the course of a project. Consequently whilst BIM will continue to gain industry adoption it will never constitute the sole outcome of a design, or reach its full potential until it is understood within the greater expanse that is the digital design collaboration process.
Where my thesis begins
The architectural design problem that sets the undertone for the thesis is how to improve the ability for an architectural team members to maintain a digital record of their design processes and decisions in an increasingly distributed operating environment. My thesis begins with the observation that perhaps the concept of the Building Information Model is not the silver bullet to which many AEC professionals are pinning their hopes given it cannot realistically fulfil all the roles destined for it. The thesis then explores the principles of the Internet as a means of moving forward the process of architectural collaboration.
Reasonate: A prototype put together and tested
for my thesis that explored distributed ways
of design development and documentation
As the Building Information Model was based on pre-Internet, product model concepts it cannot flourish within an untrusted, dissimilar and distributed environment such as that present on the Internet. Through interrogating the methods and motives of the Internet my intention is to establish a new set of principles for the Project Information Cloud. The Project Information Cloud is a term of my own creation that describes a method of design collaboration that emphasises distributed working environments loosely coupled together via Internet technologies such as hyperlinks, syndication and evolving classification systems generated within the design team.