Smart Geometry's show and tell

Amid the gilded grandeur of the Palace Hotel in downtown San Francisco more than 200 parametric design practitioners from all over the world convened to share their experiences of the use of parametric design software within architectural practice.

The event is an annual itinerant flocking of an otherwise widely dispersed collection of individual practicing architects, academics and students who fuel their interrogation of computational design with a heady mix of skill, theory enthusiasm and hard work.

Born of a “Let’s make it happen!” email in 2002 from Lars Hesselgren of KPF to former colleagues J Parrish of Arup Sport and Hugh Whitehead of Foster & Partners, Smart Geometry continues to flourish under their leadership. It is built on the premise that openly sharing knowledge is the best way to unlock the full potential of the parametric design software they exploit in their own work — as demonstrated by their involvement in projects such as London’s Pinnacle (KPF), the 2008 Beijing Olympics stadium (Arup Sport) and Beijing Airport (Foster’s).

Four workshop days explored the use of Bentley’s GenerativeComponents software (originally devised by Robert Aish). In the workshops, students and practitioners worked on individual projects in a collaborative, tutored environment. The focus was on providing a 3D scriptable modelling and analysis environment flexible enough to offer opportunities to use the software in ways that were never envisaged by its creators.

Bentley not only sponsors the Smart Geometry conference, by involving its own software development team in the event it also supports participants and responds to their demands to develop the product.

The alumni day allows past and current participants to present the evolution of their thoughts on computational design. Highlights included presentations by Brady Peter, Stig Anton Nielson and Martin Tamke of the Royal Danish Academy of Fine Arts, which combined pragmatic problem-solving in the fields of acoustic analysis and planning of underground stations, through to research into materials and construction techniques that explore the use of cad/cam to foster digital craftsmanship rather than industrialisation.

Martha Tsigkari of Foster’s demonstrated with boundless enthusiasm how the use of GenerativeComponents has been integrated into the work of design and production information. The Valery Gergiev Cultural Centre project illustrated the use of a single Generative Components model to generate four related but individual buildings (see illustration).

Foster's used Generative Components to propose four related but individual buildings for its Valery Gergiev Cultural Centre design. Clockwise from top left: philharmonic concert hall, multi-purpose concert hall, school and youth centre.

Jenny Sabin of Lab Studio, University of Pennsylvania, presented work showing how biological structures at both the level of individual cells and of larger organisms can be used as form makers for use in computer-aided design.

The final conference day saw Smart Geometry cast its net wider to invite guest speakers who lead the field in computational design in architecture, structural design or other industries. Architectural Association director Brett Steele set a theme for the day with a presentation on the impact of digital technologies on collaborative working, using the AA school as an example in the particular, then expanding it to what the architecture profession might learn from it in general.

In case any of the audience were in any doubt, both Parrish and Whitehead gave presentations on how parametric design software had been instrumental on major projects in which they had been involved, the fundamental point being that parametric design software had assisted in delivering significant international projects to tight deadlines and had allowed the designers to manipulate and control complex geometries without an endless series of manual iterations.

The aim is to offer opportunities to use the software in ways that were never envisaged

Judit Kimpian and Josh Mason of Aedas together with Jeronen Coenders of Arup presented the results of a collaborative project undertaken with other consultants to produce an interactive design tool for tall buildings, combining the parameters of form, floor areas, cost and environmental performance to generate 3D computer sketch models in response to design choices.

Architecture has always lagged behind other design professions in its use of parametric software. David Kasik of Boeing described how fundamental 3D computer modelling is to the development of the design and assembly sequencing of new aircraft. He described how 3D model-viewing software had been improved so that members of a 2,000-strong design team could now view and manipulate on a standard laptop complex 3D models containing more than 1.5 billion polygons.

To emphasise that parametric design software software isn’t just for hi-tech projects, Mark Burry of RMIT University illustrated how he has been deploying it to assist in resolving the parabolic geometries of the stonework of the Sagrada Familia in Barcelona, while Chris Williams of Bath University demonstrated software he had written for manipulating arched forms made from whole tree trunks.

Overall, the conference was striking in that it eschewed aggrandisement and self-promotion in favour of a collaborative spirit. Rather than any sense of individual or corporate endeavour, the atmosphere was reminiscent of a celebration of shared success on a notional open source project.

Does parametric design create a new aesthetic for architecture?

Lars Hesselgren: It genuinely helps to create new forms. You can draw anything by hand or with cad, but you can’t do iterative processes. It gives you enormous opportunity to do things that are simply not feasible otherwise. That changes people’s attitudes to how you design. With the first building we did using GenerativeComponents, the speed with which we were reacting to design decisions was extraordinary. If we had been modelling by hand in a complex situation, it might have taken us a fortnight but when we ran it in GC it would take five or ten minutes. People haven’t really grasped what you can do with iterative processes.

J Parrish: Each of us uses it in a different way. For me, parametrics have meant it’s possible to do things much more effectively. Before, maybe we’d have a couple of goes and then run out of time — the project waits for no man. I see it as something that’s defining a particular form of design. It’s a very powerful set of tools that allows you to try things that you wouldn’t otherwise. It’s also immensely helpful in dealing with the mundane and the repetitive. Once you’ve identified the rules of design in a particular field and codified those, you can concentrate on the things that are going to make a difference to your design and try out many more options. The end result is without doubt a better designed building.

Left to right: Lars Hesselgren, Jay Parrish and Hugh Whitehead.

Hugh Whitehead: We’ve all been dabbling with parametrics for longer that we care to admit. What’s emerged now is a wonderful choice in different modelling strategies. Most design is developed using some kind of model strategy. It’s much more like the way people tackle musical composition these days. We’ve got midi recording systems and synthesisers. It’s all out there, and it offers a lot of new territory. But it does mean you have to put in the time building up fluency.

The current generation are very energetic and keen to develop the skills, but putting them together in a meaningful workflow takes a lot of practice. We’re seeing an amazing amount of technical fluency, but we’re keen to see it directly applied to more obvious advantage. Let’s experiment and see what happens when you have a design focus.

J Parrish: Let’s be clear: it is not a substitute for design skill, experience and ability. It‘s something to empower them, and not a substitute in any shape or form.