Generative Design

The future of CAD

What is missing ?

with 7 comments

I just returned from ”Design the Dynamic”  design conference in Melbourne and would like to share what I heard, felt and learned. This conference was based on Design and  Computational Fluid Dynamics, often referred to as CFD. It was preceded by a 4 day workshop in which impressive progress was made by students of RMIT in prototyping rigging up and analyzing an interesting range of design concepts.

The symposium on the last day had the usual cocktail of presentations from practicing architects showcasing their current work and work processes, academics discussing issues that are relevant to the academic world. Also present were those who connected random thoughts to random words illustrated with equally random images. Noticeably absent were the representatives CAD companies. Perhaps because they are aware of the irrelevance of the rest in a game that is now defined, led and played entirely by them.

The symposium however was interesting and here is what I learned from it.      

Winds can shape form

Streamline of wind flow and pressure

Not only are the dunes of deserts shaped by wind; buildings too can be shaped by winds. Wind can be a generative force. An interesting presentation by engineer Peter Felicetti based on collaborative research with Prof. Mike Xie and JIWu Tang showed how twisted shapes can drive wind upwards and provide an aerodynamic lift that works against gravity. In tall buildings, even though wind forces are significant they  are only a fraction of gravitational forces, still they can help shape them.

Rough & inaccurate tools are still very useful

The results of the 4 day projects that preceded the conference were also presented. Results from various CFD tools of two extreme kinds were compared. Whiles the tools like Vasari reduced the complexity, they were discovered to be less accurate than more advanced analytical tools such as ANSYS that are usually operated by experts.  However, tools like Vasari were found to be useful despite their limitations and misuse by the “Jonny English of CFD” as a presenter described himself, because they can be fixed quickly with a bit of timely expert input. It also seemed that the loss of accuracy was mainly due to Johnny English effect . Those with better understanding of CFD could drastically improve the results over Jonny using the very same tools. So this way, Johnny is in the game. He has his role and the experts have theirs too.

Navigating Geometric search space

Hugh Whitehead from Foster and Partners emphasized the importance of team work and the unsuitability of the optimization methods (especially genetic algorithms) in seeking to optimize multiple criteria design problems where the complex conditions and goals that are constantly evolving.  Kristoffer Josefsson, a young mathematician, now working for Fosters proposed an elegant mathematical way of representing geometric search space. The potential breakthrough is the ability to explore geometric form variations through natural representations of form, rather than the tweaking of individual parameters. Though I did not understand the structuring of this process, I plan to follow it with interest.

A dignified headache

Much of the discussions by mature practitioners were about “Managing Change Propagation” – a dignified term for a perpetual headache caused by ill-structured design processes. There is a known need to continuously stitch together processes between different types of design activity, mainly because of the lack of ways of structuring design intents digitally, except through drawing, drafting and quantity surveying tools developed separately by CAD vendors. CAD companies are now busy stitching them all together and are happy to announce significant improvements.

Scripting moves to center stage

No more presentations on Shape Grammar or the virtues of ancient geometries. The glorification of Euclidean geometry within Euclidean CAD systems through the rationalization of ancient grammatical rules seems to be thankfully over – at least in this new continent. Scripting seemed to have entered the center stage. Unnoticed by academics, it entered academia through the back doors for student usage, thanks to tools such as Grasshopper. Perhaps because it had no known association to any ancient ideas of orders, it remained unnoticed by academics. Its now ready to be baptized belatedly, given its current dominance.

Scripting happens to be the nature’s natural  choice for designs exploration –

because she designs with code and not geometry. In her language, geometry is the result of code and not vice verso.

Architects are now increasingly addicted to the richness of form authored by code. Scripting is most definitely here to stay.

From the presentations, it appears that Grasshopper is becoming the preferred platform for early stage digital design exploration, due to its fluidity and connectivity to various other plug-ins that are now available and the creative energies of the vibrant community that produces it. It is increasingly becoming the medium through which early stage designs are expressed in exploratory and generative form.

Externalization of a thought process

Modelling was interestingly described as “Externalization of a thought process” – which I thought to be a powerful way of thinking about modelling and central to much of the discussions that we need to have.

Prof. Tom Kvan discussed the problems of implicit and explicit representations. Design is described as a process that stretches between the implicit and the explicit. While a sketch is an implicit design model, BIM is an explicit construction model. What was felt to be  missing was ways of creating appropriate digital representations which are capable of transitioning from implicit to explicit representations. I would guess that they would have the unfinished quality of sketches, encapsulating multiple possibilities and high levels of uncertainties.

A noticeable gap

In conclusion, I felt that the discussions revolved around a single dilemma: the dilemma of early stage design. How to represent ideas and concepts in digital form – before the fancy forms and details of design processes take over. Disappointingly, except the lecture by Peter Felicetti and Prof. John Frazer, none of the lectures or workshop presentations went beyond sophisticated tool use – making me conclude that,

We really don’t know how to use these increasingly powerful tools of analysis to shape architectural forms in their conception.

Now, that is a challenge for generative design.


Written by Sivam Krish

November 26, 2011 at 1:12 pm

7 Responses

Subscribe to comments with RSS.

  1. Sivam,

    Thank you so much for documenting your review of the conference. There are hundreds of conferences like this one happening every year. I always wonder what useful information I am missing. I wonder a little less about that conference because of your work. Thank you, again.

    Stan Carroll

    Stan Carroll

    November 28, 2011 at 1:27 am

    • Hi Stan, I rarely go to conferences. Often there are only a few events that move you beyond the known edge. I am glad to have gone to this one because it gave me the pulse of where things are moving.

      Sivam Krish

      November 29, 2011 at 8:52 pm

  2. Sivam,

    Thank you for coming to the symposium and taking the time to compile this = it is very useful to have the detailed feedback on what you took away from the day. We hope that for the next event, you might also be able to join a workshop to explore some of these open ended questions hands-on.

    A few comments to place your reactions in context:

    The workshop and symposium were about dynamic feedback systems in early design, with subtopics such as nested systems.

    Sailing and air movement in relation to the built environment were taken as particularly challenging contexts in which to experiment and understand more about tackling this topical design problem.

    Reciprocity and design and analysis conversations that move between analog and digital models was an important site of experimentation in the workshop. Computational Fluid Mechanical analysis is highly complex field and rather challenging to incorporate meaningfully in rapid early design. Nevertheless this is a challenge that is currently of interest in built environment design modelling to all of design practice, the academy and software developers.

    I note that the workshop participants were a broad mix of architectural practitioners, post graduate students and academics from a range of practices, universities and Australasian cities, not solely RMIT. They included architects, computer scientists, engineers, mathematicians, sailors etc.

    I note that the whole event was led by one of the world’s leading architectural practitioners in area of computation in design who also set the workshop challenge. In this way we took our starting points in the “real world” and trust that some of the outcomes will find their back there.

    In response to the charge that CAD company representatives were absent, I would note that there was a presentation from Memko Pty Ltd on Dassault products at the symposium and that several of the Autodesk developers of the beta software Vasari were closely engaged in the conversation in the workshop cluster exploring the representation of wind behaviour in cities in early iterative design.

    I am surprised that you felt it revolved around a single dilemma and about representing early design ideas in digital form. Certainly for me it was about being catholic in the ways that we find to integrate the best of digital representation and analysis with what we can gather and re represent and share effectively through analog means. This is an important point with respect to the efforts to construct design decision making environments that provide feedback that people can really usefully assimilate and use in their design process.


    November 29, 2011 at 1:16 pm

    • Hi Jane,

      I forgot to mention that I attended only the symposium held in the last day. I popped in for some beer at the end of the day to where the workshop was held. Met some RMIT grad students there and was very impressed with what they had rigged up in 4 days. But as I mentioned these approaches fall well within standard, ideation, design and test processes.

      FEA and CFD tools have now been used in engineering for decades for design development – though they are now becoming increasingly easier to use and are therefore used more in the earlier stages of design development. This is certainly a significant and meaningful improvement around which I believe the event was centered.

      As far as design process development is concerned it does not matter if the engineering analysis is structural, aerodynamic or thermal. The issues related to using analytical tools in early stage design are somewhat similar, in that they are carried out with hi-level of uncertainties around designs that are still under evolution. The problem domain was well chosen boats and buildings share a lot of common issues.

      I agree that there world’s leading architectural practitioners involved in the workshop. But I was making the point that no new approaches in the use of design technology was reported. By this I mean the use of computers to explore design possibilities (without being driven by the designer) which is the central concern of this blog. What I saw was the good use of analytical tools in slightly earlier stages of design.

      Yes the only CAD industry presentation was from Memko and that too was a generic presentation. Yes folks from Autodesk attended the event and I was sitting next to one of them. Since the CAD industry has moved so far so quickly in recent years, they are often able to make impressive presentations on the impressive stuff they have developed. Thankfully this was not one of those events.

      Coming back to what I still feel as the central challenge is in the representation of early conceptual design in digital form – this is where I saw no real progress. This happens to be the area of interest to me and the readers of this blog. Stated briefly – “ How to use computers not as tools but as creative exploration engines “. Hence my report on this event was based from this perspective – to share what I learned from this event.

      I did not see any promising approaches other than those I mentioned which I repeat here. Which I shall qualify.

      1) A possible new way of representing geometric space as combination of various instances instead of driving them with design parameters by Kristoffer Josefsson.

      2) The approach taken by Peter Felicetti Prof. Mike Xie and JIWu Tang who presented the use of combination of CFD analysis and a structural optimization method known as Evolutionary Structural Optimization (ESO).

      3) The presentation of John Frazer – one the pioneers of generative design. He presented some mature generated designs derived from what appeared to have originated from generative cells by his former Phd student Patrick Janssen. I have met Patrick sometime ago and could not get a clear idea of how the cellular arrangements are translated into complex built form. I am looking forward to learn more about this from John when I interview him on this blog to which he has kindly agreed.

      The use of digital tools in early stages of design presents significant conceptual and implementation difficulties – which I believe are poorly addressed in most forums and happens to be one of the reasons for starting this blog. From the kind of traffic we are seeing in this blog, I am convinced that there is some real interest in this.

      This blog article was not meant to be a general review. It was written primary for those who are watching the evolution of generative design technology. Hence issues are discussed from that perspective. Overall I enjoyed the symposium. It was very well organized. I look forward to your next event and will be happy to participate.

      Sivam Krish

      November 29, 2011 at 10:03 pm

  3. Thanks Sivam for taking this proactive approach to keep all of us who missed such a valuable discussion to be engaged in the subject. Much appreciated.

    Design sensitivity is at the heart of this question. You really don’t want those tools to “engineer” the design at the early stages while you are blending the wide range of design variables. Design goes beyond the forces that naturally shape it. There is a poetic dimension that distinguishes the creative work from merely a work that is generated as a result of natural forces. These forces can provide a general guide to the main form proportions or orientation, but after that we need inspiration to blend those forces into a creative output. Here where design begins.

    Zayad Motlib

    November 29, 2011 at 2:39 pm

    • Zayad, you point to some very relevant issues.

      It is possible to derive architectural form from purely technical considerations. Architects like Nervi, Cadella, Ferri Otto and Bucky did it well from a structural point of view. We see less and less of that and more and more structurally inappropriate and inefficient forms that are eye catching. So aligning form to be sympathetic to material qualities, structural issues and environmental issues is where I believe generative design will play a redefining role.

      The conference touched upon the good use of tools but fell short of exploiting the computers inherent creative capabilities except in ways that are know in engineering optimization. Perhaps generative design was not within the scope of this conference.

      Now, what you propose “we need inspiration to blend those forces into a creative output” is a much much greater challenge. And this is where generative design parts ways with optimization – which is numerical end goal driven. I believe that generative design will be driven by human designers and not by numerical criteria. In its best form, generative design will be a tool that supports human inspiration.

      We are a long long way from that.

      Sivam Krish

      November 29, 2011 at 10:28 pm

  4. […] Quotes via, and Read More at:What is missing ? � Generative Design […]

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: