What is SmartGeometry and where should it go? (part 2)
In my last article on this site “What is Smartgeometry and where should it go? (part 1)” I promised to give some of my views what SmartGeometry has achieved so far, what we are currently exploring and where interesting fields for the future might lie. The first I’ll cover in this article, the last two will be covered in the next parts of this series.
If I try to look at the past SmartGeometry workshops and conferences and try to abstract this in a general overview, what we have explored so far with SmartGeometry is the idea that computation, more specifically parametric and associative software and design strategies, could have great benefits in design, because it is able to model (part of) our design logic and knowledge rather than only the artefact of the design. This is an evolution of traditional CAD system where a representation of the artefact (building usually in our case) is modelled in 2D drawings or 3D models. These models only contained the result of a design process: a representation of the geometry of the building. Parametric and associative systems model a generative process which represents the logic of the designer in an explicit manner, close to mathematics and programming, and derives the model of the artefact from this logic by generation.
At SmartGeometry, we have experimented with this concept, explored its possibilities and sought a new language it could create in design, which can be seen in the many workshop results on http://smartgeometry.designscience.ca/, http://www.smartgeometry2007.com, http://www.smartgeometry2008.com and first buildings recently built or under construction.
On the one hand I have to say that parametric and associative has been done and we should seek for the next step. To be honest, and I expressed this at the last alumni day, I start to observe repetition in the results of the workshops which might be a sign that we are finished with this concept and we need to look for a new one to explore. However, you could also see this as a challenge for new attendees of the SmartGeometry workshops to study all the material already produced really well and not try to recreate any of that material, but try to come up with something really new and exciting, a viewpoint that no one has taken before or a geometrical paradigm no one has ever seen before. Perhaps this could revitalise the results and completely change my opinion of the workshop next year.
On the other hand I see that we have only worked on the “first iteration” of explorationof parametric and associative design. We have briefly explored a lot of different (mainly geometrical) possibilities in architecture and a tiny little bit of engineering, but I sense that we could do more iterations to really start to use parametric and associative design software to its full extend in practice. What controls and concepts would we need if we really would want to make use of a parametric design system from early start to end? What do we need if we want to start sharing these models? What do we need if we want to collaborate with these models? In my research practice I currently see lots of interesting developments and I hope to see some new ones at the next SmartGeometry workshop.
Like last time, I’ll conclude with inviting everybody to discuss these or other ideas. Discussion will enhance our direction and revitalise us for the future.
Jeroen Coenders
- Jeroen Coenders's blog
- Login or register to post comments
7 comments
Who's online
Upwards and backwards?
As a new reader of SmartGeometry, I'm mildly astonished that we seem to have abrogated the design process to ever-more-clever software. Yeah, yeah, yeah, CAD vs. BIM blah blah blah. I'd like to postulate that the fundamental function of any computer system is to GET OUT OF THE WAY of the design process.
I feel particularly at odds with Mr. Coender's statement that "Parametric and associative systems model a generative process which represents the logic of the designer in an explicit manner, close to mathematics and programming, and derives the model of the artefact from this logic by generation."
I'd like to make the following points:
Inspiration is not based on logic. Yes, I agree that a building must have a program, but iconic buildings, like the Sydney Opera House and the Bilbao Art Museum seem to me to reflect a process more akin to sculpture and art in general than to any "connect the dots" process-based methodology. It's the conceptual leap into the darkness that counted here - and in so many other brilliant buildings - not the following of a prescribed path.
We seem to see a widening split in the practice of design, and in particular in the practice of Architecture and Engineering. The vast majority of practitioners today earn their daily bread reiterating the tried and true; from the merely pedestrian to the frankly boring, never mind the out-and-out crap we see all too often. Design for these poor creatures lies in taking directions from their fearful and ultra-conservative clients (Developers and Government alike!) and churning out yet another (gag, hack!) "traditional" or "classic" pastiche.
The ever-widening gap between this group, and the ever-diminishing number of committed and thoughtful designers, Architects and Engineers, who are willing to take both pains and risks in their search to raise their own personal levels of achievement, speaks volumes of the direction of the professions as a whole.
To all out there! Please! Please! Design is a fundamentally human process, informed and supported by education, experience and personal committment. Tell me - how is this to be programmed into the software?
Colin Cargill
The ease of technology
Dear Colin,
I totally agree with you saying that design is a human process, and that inspiration is not based on logic. Totally true, but....
I don't think all this development of tools, software and systems is something that blocks the road for a iconic design, unless you want to make it all handcrafted from day 1.
In my opinion, technology is not there to replace human creativity, but to support it. Where the creation of stunning design is merely bound by ones creativity, in the real world one is bound by real restraints and boundary condition (knowledge, ability and cost, and probably more). This technology helps us to translate creative designs and visions in something logical and manufacturable. The development pushes our knowledge and abilities to realise what we come up with.
In short, these developments are aids, not replacements. The skill with which the most beautiful project have been made in the past, is still needed to make them in the future... with the aid of more advanced tools and software.
This needs not to be programmed.
Janwillem Breider
Janwillem Breider
Arup
cutting edge starts to bleed
I suppose as the 'real world' starts to adopt associative/parametric methods it's pretty much guaranteed to start to look 'done'.
I suppose the big question is always 'what's next?'
As we tidy the blood that comes of the back of the sword into new recipes for balck pudding, somewhere else, someone is making a new cut.
Does anyone care to venture a guess as to where that cut is being made?
what next?
Massive optimisation. Optimisation is a well-proven technique, but it is limited by combinatorial explosions.
The capture of design process is precisely where intelligent appeal to optimisation can be created and controlled.
Architetects must learn to formulate questions more precisely. It is the putting the question that will be the new skill.
The first obvious application is performance-based architecture. Once we understand that the edge WILL bleed.
Lars Hesselgren
Rip it open!
I agree with Lars, addressing performance is the next obvious thing.
Technically speaking we have been there for quite some time already. The mindset of architects and consultants alike is the bigger problem. This mindset is nurtured by the rigid way in which contractual constraints prevent many practitioners from sharing their models and from exploring performance across disciplines.
I’ve been looking at parametrics combined with optimization for quite some time now. I have created manifold parametric variations on the ‘Rectangular Pitch Stadium’ in Melbourne, Australia together with Arup to optimize the overall roof geometry and its structural member-sizes concurrently. As much as we were able to create a tight link between GC and Arup's structural optimisation software, the GC file was not always bounced back and forth between architects and engineers – there was too much concern about ‘losing control’ over the geometry.
Fortunately this mentality is changing – It is a change that is needed in order for us to dip into the full potential of what ‘smart’ geometry is able to offer. At Arup we are currently introducing the term ‘recipe’ in our jargon to describe the parametric setup of geometrical components in unmistakable terms so they can be re-created by others (such as the archtitects) without running risk of losing control about the underlying (rule-based) principles. I understand that Hugh refers to something similar as: ‘geometry method statement’ at Foster’s.
I believe we need to learn to work with these sort of ‘recipes’ more and more often; we need to learn how to use them to everyone’s advantage - be it the engineer, the architect, or even the QS. If we do so, we can start to fully explore the ‘aesthetics of performance’ through optioneering processes that will present us with an array of informed design alternatives. These alternatives will offer us solutions that go far beyond the formally-driven parametric experiments we have learned to cherish in the past few years.
Let's rip open some conventional boundaries and let the parameters bleed out to all disciplines!
Dominik Holzer
Integration
In spite of being a fairly new face, i would still like to react to the above.
First, I agree that at first sight parametrics seem the be done. Using a clever set-up, one can parametrize almost anything these days. I do believe that that is the way to deal with the more complex projects of today and in the future.
I think the next question is how to use this. More specifically, how will this approach translate in real buildings being build in a more efficient and intelligent way.
On that point I can find myself in Dominik's approach with the recipes. These recipes could be very useful to clarify the communication between various parties. But I would like to see the use of such recipes being stretched not only between parties, but also through the development of a design process. In different stages of a project there is the need for different types of work to be performed. These can (nearly) all be approached with parametrics, as could be seen at the last conference(s) (I can only speak about the last one in Munich) where the workshop attendees had projects varying from very ealry design tools to tools meant for the design and optimisation of individual elements and details.
But how to go from one stage to another (and back)? Here the recipes could play an important role as a 'backbone' to guide the project through the process of refinement and re-design, without having to start with new models or creating the next custom made interface to translate from one tool to the next. In Arup, I already see work being performed on that issue.
I also think that optimisation is the way of making the best benefit of the computer-crunching-power we have at our hands. This optimisation can be performed in any stage of the design.
I hope that in this way the various parametric and performance based design tools can be combined to be able to make a multi-disciplinary project from first idea to realisation.
We have the stones, let's make the necklace.
Janwillem Breider
Janwillem Breider
Arup
Fully Integrated Approach From Day 1
I agree whole-heartedly with Janwillem regarding the need for a fully integrated approach to design that, at the building scale, involves architects, structural engineers, and mechanical engineers. To further the thought, the need for fully integrated project teams at the urban scale is needed as well. It is not enough for one player in the project team to have the virtuous or bold idea that is then rationalized by the other players. The need is here to have everyone sitting down at the same table from day 1. Only then can a truly integrated desgin be achieved.
So, my question becomes what is holding this back? As architects and engineers are we all so used to the architect/consultant model that the work-flow systems in place for dealing with multi-disciplinary team prohibit it? Are ego's waging a battle for ownership of the "Big Idea?" Are developers operating under some false pretence that they don't need to be concerned with who the structural engineer is or who the mechanical engineer is?
Perhaps issues of workflow in a multi-disciplinary project team need to be addressed in schools. Maybe architects need to learn a whole lot more about structural and mechanical engineering and vice versa. Slowly the paradigm can shift and the lines btw architect, structural engineer and mechanical engineer can be blurred. I can see there being a new breed of designer that combines all three. One that occupies the middle ground and skillfully mediates btw all 3.
I think Smart Geometry can help to faciliate this integrated multi-disciplinary approach and I, personally, would like to see more information on work-flow discussed. Whether it is through the nature of the projects undertaken during the workshops, or the publication of a directory of all professional and academic entities interested in a full-collaboration effort, or the involvement of Smart Geometry in the curriculum in schools, I think that Smart Geometry can have a very positive influence on the professionals that produce high-performace sustainable cities and architecture and I don't think that seeing recurring projects from conference to conference is a bad thing, unless it were done by the same person. Evereyone brings their own interests, motivations and sensibilities into a project, computational or otherwise.
David LeFevre
Smith + Gill