In Part II of ‘Beneath the Surfaces: The Hidden Elegance in Parametric Design’ architect Don Chong continues to discuss the fuss over ‘parametric design’ and ‘computational design’, based on his observations at the Bentley SmartGeometry conference in San Francisco recently. His attention turns to design data and the development of design tools in real-time.
Building as Instrument: Real-time, Raw Data
The process of design as a beneficiary of ‘dynamic feedback’, is intriguing to say the least. How inspiring and promising, then, it is to see ‘design’ truly rooted in the actual criteria of performance versus a predisposed idea of, say, just composition or aesthetics. A building’s micro-climate, for instance, cannot be so decidedly deferred and assumed to just ‘fit into the building afterward’.
For the many that subscribe to this ‘dynamic feedback’ approach, maybe that’s not a compliment (though it’s trying to be), but just a given. I suppose these folks have already known that these endeavours have always gone way beyond mere visual or formal choices.
This is about front-line, investigative evaluation in which the ‘analog instrument’ secures its role as a provider for real-time assistance – by way of the built form being a resultant of the careful integration of digital datasets, parameters and modelling.
{sidebar id=359}
{sidebar id=360}
Enter the research project ‘Adaptive Pneus: Inflatable Adaptable Façade Project ‘ by Mehran Gharleghi and Amin Sadeghy for a delightfully sophisticated urban building located in Iran. The project, developed while at the Emergent Technologies and Design programme at the Architectural Association in London, struck a resounding chord with the SmartGeometry audience this year.
By deftly ‘harvesting’, among other things, conditional characteristics of natural daylight and temperatures across the surface of their proposed building – in an already challenging, tight urban condition – Gharleghi & Sadeghyi structured an astute and credible logic to create a form at once pleasing, yet anchored in the know-how of building science efficiencies. Gharleghi & Sadeghyi proposed a convincing architectural project whereby the skin of the building was a physically functioning, organically-mechanized system configured to manipulate itself to changing micro-climactic patterns.
The building employed adaptable pneumatically-powered openings configured to increase the ‘quality of spaces’, as Gharleghi & Sadeghyi puts it, as the weather turns throughout the day. With local sensing and decision-making, there is a collective and distributed intelligence within the building.
{sidebar id=361}
Gharleghi & Sadeghyi’s big leap here might be their vision in seeing a building as essentially a working local ‘data-harvester’. By this, I mean a building which continually monitors and re-calibrates itself to optimize both internal and external conditions – with a bonus. The actual openings themselves, at the specific locations of the particular ‘data point’, would fluctuate its physical shape in accordance with the optimal collective surface of the building.
Here’s the major advance: that the building’s requisite infrastructural system takes a purposeful change, or local adaptive shift, at the specific locale(s). Basically, ‘change’ where it counts. It’s beautifully simple in that it fundamentally believes that if a building can now have the means to ‘learn’ at selected points, right where things are changing, then a building should also directly ‘address’ the issues at those same selected points.
Gone is the notion that the overall building would ‘flick a switch’ to perform a more generalized, blanketed response, such as providing more heated forced air, or increasing the fan strength, etc. Remember, this would be more than just predetermined ‘zoning’ or a menu of ‘programmes’ punched in mere months before the opening of the building. It’s raw-data shaped into real-time response.
{sidebar id=362}
Gharleghi & Sadeghyi is proposing a system whereby various ‘pillow’-like louvres are inflatable and ‘dilatable’. The form begins to resemble, curiously though elegantly, like a sea urchin with its cellular look. Their choice to use a ‘smoke-injection’ system provides the necessary gradations of translucency to regulate the incoming light; light which is at times welcomed, and at other times preferrably avoided to maintain the ‘quality’ of the spaces behind.
The system claims to be inspired by movements of muscle in nature, which makes sense. The real action taking place occurs beneath the surface, and at individualized points. In actuality, it is the air within these louvres which would increase depending on sunlight by virtue of being connected to compressors – as typical with normal pneumatic pressurization balancing systems.
In this way, this proposition make it more than a monolothically pressurized system of, for instance, a covered stadium or a tennis dome. The modular components also have the ability at times to regulate openings for natural ventilation as well, while also creating a favourably textured patterns along the skin of the building, too.
No pun intended, this clairvoyant project pushes the envelope where we begin to see an emergent role for an ‘architecture as apparatus’ – an architecture which operates with a clear sense of local building responsiveness within a more global building system.
It allows for an associative response for the building, as it evaluates itself through the course of its day, week or, frankly, life. The building begins to respond the way nature itself lets plants gradually shift in their phototropic ways. For Gharleghi & Sadeghyi, the building would take action in an iterative fashion. Talk about building performance.
It’s hard not to be reminded of what is now a ‘classic’ precedent in late 20th-century modern architecture, the Arab World Institute designed by Jean Nouvel. The institution’s most notable feature is a system of dilating apertures across its facade, where wafer-thin mechanisms using camera aperture technology provides a pleasing if not inspired pattern of openings.
These openings would adjust according the travelling sun and its light impact on the skin of the building though, as its date of completion in 1987 would suggest, the apertures operate on perhaps in a more generalized level than what we are seeing as possible two decades later. The marvel of this project is the completeness in which the architecture finds a resolute identity in the craft and in the technology.
It’s a project that has arguably been waiting for ‘software structuring’ to be equally as sophisticated. And an ‘upgrade’ towards a dynamically-systemetized data model, such as Gharleghi & Sadeghyi is putting forth here, could be just the kind of direction to go in…
{sidebar id=363}
Isn’t it high time that the architecture that we produce not only maneuvers in a sensitive fashion during the design process, but perhaps afterwards, too? Could we not expect to see the entire building as a larger mechanism that is indeed ‘alive’, something akin to a habitable instrument?
Maybe with Gharleghi & Sadeghyi’s rather ambitious project there is something eloquently simple or comfortably practical here. Perhaps they see the need for the direct questioning, or the revisiting, of architecture as a means to provide its own ‘self-awareness’, its own ‘self-checking’.
Suddenly, the building is not a static idea. It is not, anymore, a constructed mass of additive and indifferent components – but rather a breathing, living mechanism that can have the ability to self-evaluate, learn and mature. Or more rightly, grow.
Design as Logic Tool: Elegance Beneath the Surface
A fascinating take on computational design, “Metro Mechanics: Based Rule Generative Subway Station Designs,” was presented by Stig Anton Nielsen’s vision for a new approach to tackle Copenhagen’s growing need for new subway station designs.
Nielsen opted to use a parameter-based logic based on known and measurable criteria which characterizes potential sites for subway station locations. As one would expect, the daunting task of mapping the various infrastructural events throughout the metropolitan region along with traffic systems, water systems, power lines, etc. is followed by the ever-more daunting task of looking for the range of opportunities for clearances, access and critical adjacencies.
The positioning of access/egress requirements for subway pedestrians coupled to the thoroughfare access of the subway lines in Copenhagen incurs considerable study for multiple conditions in multiple layers. What Nielsen thoughtfully demonstrated was how a computational method would not only seek out critical constants and variables, but would help create new expression and opportunities for an ultimately ‘satisfying’ and well-earned architectural form – all purely derived from the otherwise incidental (or, likely, hard-to-digest) conditions on each site.
From an engineering standpoint, this is significant. It allows the key parameters, as well as linkages or adjacencies remain intact within given angular ranges. This would be based off of allowable code minimums as well as minimum standards for comfort, sightlines and overall traffic patterns.
{sidebar id=364 align=right}
In essence, Nielsen sets up rules and key restraints. They are based on real conditions including passenger entry points, the capacity along side the tracks, the concourse level amenities, the available ceiling plenum spaces for key services and structure, or the various means of egress. Some elements are known and fixed. Others maintain a minimum width or height, while others are freeform.
From an actual scripting standpoint, Nielsen uses as the ‘primary input for the components’, as he puts it, the following: a primary plaza, a secondary plaza, four directions from incoming traffic and ‘passengerflow’ data. One can see that the elements are effectively ‘body parts’ articulated in a way where some joints are locked and others are hinging. The skeletal variations would be next step.
{sidebar id=365}
As Nielsen showed in his real-time presentation, he was able to intuitively rotate and move the collection of key segments or components of the proposed station. In an eloquently responsive way, one quickly saw the resultant interferences and allowances of the subway station in situ. Instantly, the spectre of limitations and restraints became somewhat ‘game-like’ in its interactivity, and tactility.
It raises a good question about the issues related to the GUI (graphic user interface) experience. Even Jonathan Rabagliati and PC called this out directly in their enjoyable presentation, in how the feedback aspect in computational design is absolutely critical. In their mind, why not learn from the culture of gaming.
Make it fun, make it memorable. In many ways, it seems as if the power, affordability and speed of current workstations are now up to par with the software technology to work out realtime ‘navigation’. Whether one ‘relaxes’ a mesh surface to find optimal structural members, or is simply seeking to optimize tangential surfaces, it seems natural to let the experience become far more dexterous.
It would be something special if we could take Jeff Han’s incredibly responsive full-screen experience developed through Perspective Pixel, to meet a parametrically-based project challenge with a deep dataset model. One could manipulate conditions, slide variables and sift through options in a highly visual manner. That would be something that to me is a natural progression with the advances in digital technology, and hopefully is merely only a matter of time.
{sidebar id=366 align=left}
To be fair, however, this is already happening Nielsen’s work, though with more normative graphic interface conditions. He effectively built a tool so that he could ‘read’ the possibilities. The profound nature of the work is that its operational elegance is completely beneath the surface, in more ways than one.
The beauty lies in the robustness of the mechanism to respond to the changing ‘mechanics’ of the local infrastructure, on its way providing numerous opportunities and configurations. (It recalls some of the work of architect Stephen Holl, whose ‘logic’ or elegant ‘configurations’ in architecture often precedes the actual articulation of the final form; the iterative studies of the interwoven paths and circulatory conditions often, in my mind, ‘anchor’ much of Holl’s more memorable projects.)
{sidebar id=367}
Ultimately, and what I admire in Nielsen’s project, is that it ‘gets it’: it knows that we are, after all human. And the final decision is all the more likely be a human decision. The computational iterations allow the possibilities to be ‘pinned-up’ on the wall to then fully review. This would allow for the inevitably more qualitative judgement processes which would always occur. It’s certainly a sustainable mindset to have a tool that doesn’t pretend to go all the way on its own.
It merely sets the table, and lets the human skillsets take over, complete with all the usual trimmings of objectivity and subjectivity. And for a computational design process to be remain engaging, possibilities render good solutions. Ultimately it is about generating or, more aptly, ‘mining’ for options – not the final answers.
{sidebar id=368}
So there you have it. My take on an engaging and enlightening conference in San Francisco. It was a rich tapestry of voices, all woven together with a collective interest in showing that computational design is, ultimately, not about the computer but the designer or team of designers, and the process which is forged.
It had become apparent to me, that parametric design truly has the depth which I had always been hoping for. I would go as far as to suggest that aspects of parametric design may well be the true basis for ‘design as it’s meant to be’. In other words, with parameters, we understand that the design of a building is often an act of coalescing the many layers of relationships between events, items, and conditions.
And these are often not part of the project ‘proper’, but in fact influencing characteristics. Now that the earth and the economy are calling upon all those engaged in building, we are afforded a unique opportunity to see once again how the ‘tools’ we want to use are beginning to resemble the ‘design process’ we aspire to: namely, by both being effective, efficient and elegant.
The first part of this article is available here.
————————————————————————————–
Don Chong is a V1 Magazine columnist and architect based in Toronto. His work and studio information are available at www.donaldchongstudio.com
