I suck at parametric design

Daniel Davis – 24 August 2010

I have spent the past month revising this project, and have come to the conclusion: I suck at parametric design. I set off with intentions of making the entire model parametric, but four weeks later I am left with a shambolic set of simi-parametric tools linked together with long manual processes and discarded ambition for the complex parts of the project.

Everything that I wanted to do with a parametric model could be achieved with a parametric model; this claim is the origin of the rhetoric for the flexibility of parametric models. In practice however, I did very little of the project with a parametric model, I did a bit of the project manually and I modified the the rest of the project so it was achievable. The result is a design that is distorted towards the strengths and away from the weaknesses of parametric tools. An example of this is the curves along the surface (illustrated below). The the curves are an explicit list of points that are wrapped onto the surface with a UV curve passed between them. Most of the time this curve looks straight, but at the ends of the surface (shown in the illustration in red) the UV curve starts to bend. The solution is to use a geodesic curve (the green one), but this breaks the relationships to the other curves. Given enough time, it would be possible to use straight geodesic curves but, given how long it would take, it is far easier to use UV curves and modify the design to allow the bend of the curve.

Bent UV curve shown in red, straight geodesic shown in green

Bent UV curve shown in red, straight geodesic shown in green

It is a trade off: will the 12 hours I spend fixing the parametric model be noticeable in the final design.  And in some cases: is it faster to get the parametric model to conform to the real world, or is it faster to get the real world to conform to the parametric model. People mistakenly associate this decision with choice and blame the architect for being a luddite. But the decision is loaded and biases the design towards things that are easy to do parametrically and away from things that are hard to do parametrically. In this way, just like any other CAD tool, parametric designs can be read as being in the language of parametricisim –  dare I quote Schumacher – biased towards the easy solutions.

So parametric flexibility is more than being able to do something, it is the ability to actually do it. For me, what I am able to do and what I actually do is quite different. I suspect the difficulty in doing some of these tasks is related using mathematics as the intermediate language between designer and parametric tool, but that is a topic for another post. Have you encountered a similar phenomena in your models?



Join the mailing list to receive an email whenever a new post comes out. Otherwise, you can follow me on Twitter, or add the RSS feed to your reader.


  • Kris 25 August 2010 at 10:58 pm

    I generally see these kinds of problems in hindsight. I tend to get really invested in solving a particular problem, spend hours working on it, get frustrated, feel a great sense of triumph when I finally crack it and then wonder, “Could I have fudged that in a few minutes?”

    This is often followed by a little self-satisfied internal voice saying, “Well, I’ve done it the *right* way and the next time I use this particular parametric widget it will really work.”

    I’m dubious of this little voice. I have been using these tools such a short time that I haven’t really reused anything (at least not consciously) and its possible it could save time in the future, but who knows?

  • Daniel 27 August 2010 at 7:57 pm

    Yea I definitely get that as well. There is something aesthetic about scripting / parametric modelling and often I spend time perfecting something that doesn’t need to be perfected.

    Interesting that you have not reused may of your definitions. I am surprised by how little I reuse my definitions – or even reuse other peoples definitions. I think this is tied to the readability of the parametric model; what makes sense at the moment of creation often ceases to make sense in three months time.

  • Sivam Krish 28 August 2010 at 11:49 pm

    We are at very early stages in understanding how to model parametrically. Most parametric models seem to be single stage play. You can only do very limited variations with this approach.

    On the other hand if you think of your parametric model to be based on hidden skeletons you will be able to build complex geometry on top of simpler geometry. So you will have layers of geometry that you can parametrically manipulate. It is best to effect the parametric changes in these skeletons rather than in the last stage.

    You need to first build the geometric logic. Parametric play without geometric logic management will result in noise.

  • Daniel 30 August 2010 at 1:39 pm

    Hi Sivam,

    I definitely agree parametric modelling is in its early stages. I seems what you are doing with open generative design, and I am doing with this PhD business, is trying to anticipate how the later stages will play out.

    This project was essentially using the skeleton approach, in that the parametric tools that were cobbled together were first generating a surface, then adding points, then adding lines and finally adding and detailing the beams. I guess part of the problem I was having was connecting the skeleton to the flesh gracefully rather than manually. These manual junctions slow down the fluidity of the modelling process and limit the potential for bi-directional relationship between the tools – the detailing tool could not tell the point tool that a point was in an impossible location and should be moved.

    Then again, if the whole thing was totally linked together in one monolithic tool, I would worry about the flexibility and speed of the tool. So perhaps there is no winning here.


  • Philip 31 August 2010 at 6:35 pm

    Definitely agree that the ‘flesh’ part of the model is where the parametric approach seems to fall down. I’ve found that resolving detail in a parametric model seems to require much more effort and computer power to resolve properly, leading it to be neglected more often than not.

  • Daniel 1 September 2010 at 1:47 pm

    Hey Philip,

    It seems ironic that detailing/fleshing should be hard in parametric modelling when it is set up for variational design – although I have to agree with you. Digital Project/ CATIA is the only software I have really seen deal with this issue, basically by freezing the model and allowing you to draw geometry over the top. Even still, Gehry Tech has a team of ‘automators’ who manually write scripts to automatically add the flesh to the models.

  • Shuying 22 September 2010 at 4:57 pm


    the parametric model looks a little alike that wind-powered pavillion in the link , although they may focus on totally different things …

  • Mentioned 14 March 2011 at 8:07 pm

    […] pretty down moment (after preparing the project for an upcoming publication) I wrote a post on how I suck at parametric design. Well luckily those issues were sorted out, along with how to control the bend of timber and […]

  • Mentioned 6 August 2013 at 11:24 pm

    […] pretty down moment (after preparing the project for an upcoming publication) I wrote a post on how I suck at parametric design. Well luckily those issues were sorted out, along with the issues of how to control the bend of […]

Leave a comment