Construction is, by definition, a local business. The vast majority of AEC firms operate from a single office and bid for projects within an hour or so's drive. Their competition is other local firms. Globalization may worry textile firms but why should small architects and their employees worry abut competition from China?

However, for signature architects like Fosters or NBBJ, engineering contractors like Bechtel or Flour, it is all about being global. The majority of their work involves multi-office design with final construction some other place on the globe. Technology to help communicate effectively is a vital part of their competitive edge. They invest in software tools that help them to be more effective in the instant consortia that they form to execute projects.

Most vendors of AEC software tools for design now propose a future design process around a building model as the way to create designs and communicate them to stakeholders. But take up, though increasing, is limited mainly to the big firms. Are the needs of local and global architects firms the same? Will the building model design paradigm trickle down from signature architects to the single office firm working on a local school?

At the recent Autodesk University in Las Vegas, Autodesk's Building Solutions Division focused attention on Revit, their building information modeling tool. The vast majority of AutoCAD users are small firms, creating working drawings with AutoCAD in 2D. Autodesk is bundling Revit with AutoCAD 2004. They believe this will help users move from a traditional CAD workflow to the new design paradigm by providing several workflow options to customers of the bundled products without incurring license costs.

Of course, many of Autodesk's competitors have offered building modeling for some time. It has always been at the heart of Graphisoft's product suite. Graphisoft was a pioneer in creating GDL, a parametric building component description language. Bentley has TriForma. Nemetscheck's ALL suite also focuses on building models as the link from design to construction. All these companies are trying to hide the complexity of modeling with intelligent user interfaces.

Building modeling is not a new idea. In the 1970's, well before Computer Aided Design and Drafting tools from Autodesk hit the market, MIT and Cambridge University produced research exploring the idea. An early Computer Aided Design tool was Building Design System, from Applied Research of Cambridge, UK. Note the difference between CADD and CAD. No drawings!

All architectural practices regardless of size, have the same two fundamental business drivers. First, they must win new projects to design. Secondly, when the construction company executes the project, they must hit the cost and timescale targets in the bid. The architectural firms fee is small in comparison with cost of the construction phase. Firms tend not to compete on fees. They compete on customer satisfaction and one of the key inputs to that is meeting cost and timescale expectations. Locally, architects get a word of mouth reputation for producing designs that exceed the initial cost estimate to complete. Globally, certain signature architects get reputations for buildings that win prizes but break the bank in construction.

That is not to say that the cost of the building must always be minimized. For some clients, owning a distinctive building is more important than cost. Guggenheim's in Bilbao Museum, Swiss Re's Gherkin in the City of London are examples. The complex geometric shapes and what they cost met clients' expectations. However, the clients wanted to know what they were getting into.

However, everyone in the construction business remembers the Sydney Opera House saga. A brilliant design, based on very basic, almost diagrammatic concepts, won the competition for the young architect Jørn Utzon. However, the initial concept of free form concrete shells was not structurally practical. It took Ove Arup 350,000 engineering hours over three years before a solution was found. The roof shells were described as the "most difficult task of prefabricated assembly in the history of building." It ran over time, over budget. The various stakeholders fell out. It took ten years before the Opera House was ready. For the full story on this saga see: http://www.sydneyoperahouse.com/h/m_story_fs2.html. In case you think these problems are a thing of the past consider the new Scottish Parliament building in Edinburgh.

Certainly building modeling might help some of these stakeholder miscommunications. But today, another factor comes into play. Historically, building design has had a strong "lego" approach, where designers assemble building blocks from building component catalogues. But this can stifle creativity. Building shapes tend to be rectilinear. At local level, repetitive choice of particular building components often defines the style of a particular architect or firm. There is, of course, a set of customers who specifically want this repetitive style, often nationally or globally. McDonald's diners, Toys'R'Us stores, BP gas stations all use repetitive elements to project their brand to consumers.

Distinctive buildings tend to use specially designed and engineered components. This can increase costs dramatically. Instead of amortizing engineering costs over thousands of units they have to be covered by a single unit. The process on site can be difficult to control when a large number of similar but distinct components have to be assembled together.

With mass-customization, distinctive buildings can use single instance components that are designed and manufactured as parametric variants of a generic component. However, designers need specialist tools to define such components.

One example is Bentley's GenerativeComponents, a new technology to define individual building component instances from a set of design rules and limits. They can then be manufactured and erected at the much the same cost as standard components. It is particularly suited to complex structural support and enclosures. Different, but already validated, design options can be explored at the conceptual stage.

Another example is Frank Gehry, famous for his flowing metal structures. He wanted a method of digital fabrication to ensure correct construction at a reasonable cost. His answer was to form Gehry Technologies to create DigitalProject™. It is built on top of CATIA, better known for its role in the design of automobiles and aircraft. The concept is of a "3D virtual building model" as a gateway to information for all the project's stakeholders.

Frank Gehry believes that what he has learned in the realization of his designs will be applicable to the broader community of designers and builders, regardless of the architectural style of the projects they create. He will now market his variant of CATIA to other studios' projects.

The client is the pressure point in the construction industry chain. Clients forgive a lot. However, the one thing, large or small they never forgive is the building that was promised for occupation at 70 dollars a square foot in a years time but actually cost 140 dollars a square foot and wasn't ready for eighteen months.

A design process that minimizes project creep and removes the hiding place for cost and time over-runs is going to be good for all sizes of firm. The trickle down effect may introduce building modeling software tools into the design process for all sizes of firm. But the big trickle down gain might be from changes in building components. A move from mass production to mass customization could deliver distinctive looking buildings at a reasonable price. It's another case of efficiencies in the construction industry being driven by changes in the way we think about the design process.

Mike Evans

mike.evans@cambashi.com

A version of this article was first published in the January 2003 issue of AEC Automation.

Other Cambashi articles that may be of interest:

Who will pay for the Building Information model?
Is PLM applicable to AEC?
A-E-C Systems 2002 review

back to top