GEORGE KATODRYTIS / STUDIONOVA ARCHITECTS :: article :: Texts on DIGITAL FORM :: NEWS :: ASCAAD 2006

April 2006: George Katodrytis made a presentation and chaired a session during The ASCAAD (Arab Society of Computer Architectural Aided Design) Conference, Sharjah, 25-27 April, 2006

THE AUTOPOIESIS AND MIMESIS OF ARCHITECTURE
Algorithms and the breeding of digital forms (Extract)

One technique by which mimesis is be constructed is by Algorithms. Algorithms may be defined as a detailed sequence of procedures to solve a problem. As such algorithms may be programmed to execute a series of mimetic tasks. Genetic algorithms constitute a class of search algorithms especially suited to solving complex optimization problems. In addition to parametric optimization, genetic algorithms are used in creative design, such as combining components in a novel and inventive way and ultimately creating a new complexity of language and form. Genetic algorithms transpose the notions of evolution in nature to computers by imitating natural evolution. They find solutions to a problem by maintaining a population of possibilities according to the ‘survival of the fittest’ principle. Because of this ability to “search” algorithmic scripts or codes generate form, which is precise and complex and which would be impossible to have conceived using the basic software interface and tools. Inevitably, a new universe of formal and compositional possibilities opens up and techniques for digital form-finding.

A common approach is to define a building envelope in terms of a series of parametrically defined elements such as the structural ribs. Some forms are curvilinear, non-planar and irregularly shaped, yet precise. Furthermore, some approaches that involve rule structures seek to generate designs via various forms of growth and/or repetition algorithms. Additionally, there are approaches that seemingly abandon any kind of formal approach to shape generation but that seek to allow designers to “discover” meaningful shapes that exist within more complex geometrical patterns. Most of the approaches using formal shape-identification algorithms require specially written computational algorithms. Architects can now use advanced software to breed new forms rather than specifically design them. As De Landa notes, “ …only if what results shocks or at least surprises, can genetic algorithms be considered useful visualization tools.”

Algorithms are based on non-linear wave function that through parametric differentiation organizers vectors of density. Can we then talk about fabricating of dense and large cityscapes, employing poly-directional structural networks?

On another level, the role of design has now been transformed into that of breeding fit and beautiful forms. There is clearly is an aesthetic component: the “sculpting” of beauty and the development of a personal artistic style. As with any socio-technological revolution throughout history, architecture inevitably invents a new formal language.

Ultimately, if traditional architectural representation has been based in resembling and describing the appearance of 
the architectural object, through its use the algorithm architectural notation has become operational; to design the choreographing of the transformation process. The architectural object is transformed into event and performance, either by understanding architecture as the dynamics of spatial conditions, or by the object being understood as the actualization of built-up potentials.





Projects by STUDIONOVA ARCHITECTS / GEORGE KATODRYTIS using advanced modeling techniques and algorithms (Alias Maya MEL script)

Intuitive vs. systematic (Extract)

Through this highly refined algorithmic process, both intuitive and logical form-making processes are necessary. A grid-like structure transforms gradually into a convoluted scheme with numerous overlapping coils or folds revealing complex patterns. As the new architectural forms become complex, interconnected nodes and parameters change and, in a chain reaction, affect the outcome. Algorithms offer a degree of rationality therefore allow the designer to become a free form maker.

One assumes the role of the artist, not of computer programmer. The outcome requires being visual, and it is about inventing form than about solving problems in a systematic way. The resulting exuberance of formal composition should be judged as any architectural proposal, i.e. first on its visual impact, and the potential to become a building.

Architects may start thinking less in terms of typology and more about topology and variations. One way to introduce this thinking process is by contrasting the results. More variations are necessary in order to keep the “search” exciting. One of the design goals of the generative process is not only to support both the “designer’s” (intuitive) view and the “programmer’s” (formal) view, but also to reconcile these two views and create appropriate mechanisms as an interchange between them.

New possibilities may be the introduction of a new kind, a heterogeneous structure, as layers of form generated by different codes. Complex behavior comes from the interaction of simple parts. The basic principle of Generic Algorithms is searching for optimum balance, and the eventual survival of the fittest. Evolutionary Algorithms can be both a structural making device and a search mechanism. The structurally oriented form-finding approach becomes a useful tool for composing space through repetition. Structural members define both the surface and the volume of forms.

Filed under: Texts-DIGITAL-FORM+ NEWS