Digital technologies are pervading all aspects of our lives. They also have an unprecedented impact on architecture. As with any technology-driven paradigm shifts in the past, the latent changes are viewed by alarmists as sounding the death knell to the discipline, while the starry-eyed technophile welcomes them unconditionally. So there is nothing new here, with the exception that the tidal shifts we are facing in the light of digitalization will probably entail the most profound transformation of architecture for a very long time!
The majority of architects tend to feel threatened by the looming, inevitable shifts, rather than inclined to embrace them to “design” this transformation, as is the case in many other creative domains. This is quite astounding, as any architectural office anywhere on the globe utilizes digital technology in the design process. Or maybe it is exactly because of this. For the last thirty years, architects have readily absorbed digital technologies in their practice, but they have done so primarily to emulate pre-digital processes. Symptomatically, Computer Aided Design has been primarily conceived as replacing the pen with a mouse and the drawing board with a screen. The underlying conception and established logic of the design process remained conveniently unchallenged during this adoption of new technologies, rendering them a mere computerized extension of the well-known.
Across many fields, such computerization has been the usual first step in introducing digital technologies, primarily aiming for increased productivity, efficiency and accuracy, and typically seeking an ever-higher level of automation. However, now that the “magic pen” increasingly gains the capacity to draw – or, dare we say, even learn and think – by itself, architects feel seriously threatened by it. And more often than not the most apocalyptic anxieties are raised by those, who not long ago used to proclaim that the computer surely couldn’t be more than just another tool.
In the words of Sanford Kwinter, “let's begin with the greatest cliché of all, heard frequently in design circles: The computer is nothing but a tool (no different from any other). There are innumerable ways to address the poverty of this formulation.”[1] This deprived understanding of digital technologies, still preached in many schools of architecture today, undermines our creative capacity more than the technology itself. It is this misconception that locks us in discussions about machine versus man, computer versus architect, robot versus craftsman – conversations that are somewhat simplistic, outdated and have been overcome in many other fields. Instead, we should critically engage with new synergies between computation, design and construction that genuinely extend our creative capacity as architects.
In order to do so, we first need to abandon the sweeping generalizations that typically accompany the notion of “digital design” and promote a more differentiated understanding, which also entails distinguishing between the aforementioned computerization and computation. As Kostaz Terzidis aptly stated, “computerization is about automation, mechanization, digitization, and conversion. Generally, it involves the digitization of entities or processes that are preconceived, predetermined and well defined. In contrast, computation is about the exploration of indeterminate, vague, unclear and ill-defined processes; because of its exploratory nature, computation aims at emulating or extending the human intellect.”[2]
While computerized architectural design, with its underlying characteristic of freezing conventions in databases and its natural tendency to strive for automation, causes real concern for me as well, I at the same time muster great confidence in computation, its empowering capacity and the cornucopia of creativity that it can be for architecture. To substantiate my optimism, I will try to explain how computation in our work enables us to critically challenge conventions, to motivate lateral thinking and to question our own doing in regards to disciplinary concerns. I will do so along some selective aspects of one recent project, the Elytra Filament Pavilion in the John Madejski Garden of the Victoria and Albert Museum in London.
One central interest of the project was exploring the impact of computation on the relation between the generation and materialization of form, structure and space, which, more generally, constitutes a focal area of our research.[3] This relation is typically conceived and controlled through geometry, whose primacy has dominated architectural design thinking since the renaissance and has remained largely unchallenged, even in the recent transition from the manually drawn, to the digitally drafted or parametrically generated. The deeply entrenched prioritization of the creation of geometric shape over the processes of materialization imbues most current digital design approaches with a deeply conventional touch from a methodological point of view, although usually well camouflaged in either exuberant free form or functionalist system rationalization. If, in contrast, we view the computational realm not as separate from the physical domain, but instead as inherently related, we begin to question the age-old convention of understanding material merely as the passive receptor of predefined form.
Computation constitutes a critical factor for questioning this convention and reassessing the relation between the generation and the materialization of architectural form and space. On the one hand, computation enables architects to engage facets of the material world that previously lay far outside the designer’s intuition and insight. On the other, it is increasingly understood that – in its broader definition – computation is not limited to processes that operate only in the digital domain. Instead, it has been recognized that material processes also obtain a computational capacity - the ability to physically compute form.[4] In the case of the Elytra Filament Pavilion, this approach was explored through glass- and carbon-fiber composite materials.
In architecture, as well as in the more usual domains of application of fiber-reinforced composites (such as aerospace and automotive industries), they are conceived as amorphic materials, which means intrinsically form-less and thus dependent on an external formwork that quite literally forces a preconceived form on the material. But instead of conceptualizing fibrous form as being obedient to a predefined mold, design computation and robotic fabrication allowed us to tease out the “morphic” character of the fibrous composite system. “Moving fiber into the realm of self-expression”,[5] a motivation more typically associated with fiber art, enables the exploration of the inherent material gestalt, and its related constructional capacity and spatial potential.
Computation here oscillates between the physical and the digital world, and establishes a design feedback between the digital definition of design intent, the simulation of dynamic processes of formation, and the material’s ability to physically compute specific form through the interaction of resin impregnated glass, and carbon-fiber bundles in the process of being robotically wound around a hexagonal scaffold. Using only this one, simple fabrication tool, more than 40 highly differentiated building elements could be produced, each with a specific fiber arrangement, density and orientation that is finely calibrated with the structural force flow. The resulting load-bearing fiber structure is not only unique in its architectural expression; it is also exceptionally effective structurally, with each 5m² large building element on average weighing a mere 45kg.[6]
This remarkable performance stems from the computationally enabled extension of design towards the design of the material structure itself and the intricate, robotically steered, self-forming processes from which it originates. However, as there are hardly any precedents in the history of architecture to refer to in this approach, computation also played a critical role of tapping into a vast pool of constructional principles for fiber structures, which are found in nature. In biology most load-bearing structures are fibrous composites.[7] They share their fundamental characteristics with man-made, technological composites, such as glass or carbon fiber-reinforced plastics used in the project. This similarity in basic composition means that principles of fibrous organization observed in biology can be transferred to the design of novel composite systems in architecture. [8] Computation plays a critical role in interdisciplinary research between architecture, engineering, and natural sciences, both as a common platform for communication, but also as a shared methodological framework. In the research work forming the base for the Elytra Filament Pavilion, we collaborated with biologists to understand the underlying constructional logic of a lightweight fibrous shell in nature, the hardened forewings of flying beetles, which are called elytra.
The astounding structure of the pavilion, together with the striking compactness of the robotic fabrication unit, enables a new model of local production and an ongoing building process, which in turn challenges our conception of a building site, and the related clear-cut transition from architecture under construction and in use. The majority of composite building elements were prefabricated in our laboratory in Stuttgart, but production continued in the V&A’s garden during specific fabrication events. This onsite production, which took place while the pavilion was already being used, resulted in additional building elements for the canopy structure. However, the location and specification of these elements were not predefined, and rather than completing a pre-planned final design through their installation, they were conceived as agents for interactively adapting and evolving the pavilion over time.
As there is no predetermined final state, the canopy needed to be equipped with fiber optical sensors that allow for real time sensing of forces within the structure. This enables the computational monitoring of the changes to the structural systems caused by the further growth and adaptation of the canopy, which is driven by anonymous data on how visitors use the canopy space captured by thermal imaging sensors and interpreted in conjunction with the measurement of environmental parameters such as temperature, radiation, ambient humidity and wind. The real-time sensing combined with the onsite fabrication renders the canopy a learning, cyber-physical system and evolving space that grows and reconfigures over the time. This opens up the possibility to engage computation, and machine learning in particular, to explore the relation between human subject, tectonic object and architectural space anew.
Instead of a predetermined and linear relation between cause and effect – that is human activity and architectural response – here the evolution of space is driven by the habitat probing different configurations and observing the effect on inhabitants’ activities. Of course, the way this behavior unfolds and learns needs to be designed, which extends the role of the architect towards an “educator” that directly and indirectly embeds cultural and social value in this interaction. The ramifications of such a broadened role of the architect are manifold and this piece is too short to even begin elaborating them. But what I hope becomes obvious, though, is that the advance of digital technologies opens up the possibilities of profoundly extending our understanding of design, and thus broadens the scope of the role of the architect rather than delimiting it. It has never been a more exciting time to be an architect.
