AU2013 Day Two: Computational Design Symposium

The Computational Design Symposium occupied this morning. Organised by Matt Jezyk, the theme for this year was ‘computational BIM’. This has been a trend emerging at this year’s AU, and the symposium’s timing felt perfectly poised to capture, catalyse, and reiterate some of that building energy. 
Computation is clearly something Autodesk is anticipating will become an even bigger part of their market. This was underscored by the Autodesk CEO, Carl Bass, giving a shout-out in the keynote and then introducing today's session and staying to hear some of the presentations.
For someone that runs a billion dollar software company, Carl Bass is impressively up-to date with the nuances of what must be an incredibly small part of his overall responsibilities. Coming from a programming background, Bass seems very comfortable talking about computation and, by reports I’ve heard, actually uses computational software like Dynamo on his personal projects. I’m a hard critic to win over, but I was fairly awestruck by Bass and by what he had to say.
From Autodesk’s perspective, computational design is one area (probably the only area) where architects are leading the other design industries. Bass hinted that Autodesk were interested in applying to other products and industries what architects were doing. In this sense, computational design perhaps isn’t an incredibly small part of the company, but rather a major direction in the coming years. 
In the architecture industry, we are starting to see computational design become less about curvaceous shapes and more about the business of getting work done. The selection of speakers Matt Jezyk curated all fell into this category, with computation being applied on actual projects and in a much more diverse sense than typically discussed. Nathan Miller and I spoke about projects happening at CASE, but leaving us aside, I think the curation of speakers, the quality of speakers, and the format, was the best I’ve seen at any computational design conference – and I’ve seen a lot of them! Below if a brief summary of all seven talks from the morning.
Jonathan Schumacher, Thornton Tomasetti
Jonathan kicked off the session talking about a number of projects at Thornton Tomasetti. The discussion focused on how computation was changing their design process. Perhaps most impressive was a section titled ‘service as a software’ where he discussed the possibility of parallel analysis rather than the typical linear exchange of data backwards and forwards. Much of this was powered by a cloud-based tools Thornton Tomasetti had developed that allowed their clients to upload model data and receive analysis almost immediately.
Greig Paterson, Aedas R&D / UCL
Greig is a PhD candidate at UCL and a researcher at Aedas. He started off with this remarkable graph of how analogue data has rapidly grown in the last thirty years, and how digital data has exploded in the last ten years. Greig made the argument that we now had a lot of data associated with environmental performance, but that we were still terrible at predicting the energy use of buildings. Greig calls this gap between predicted performance and actual performance the ‘performance gap’. By his estimations, the gap is as large as 50%, meaning a building might use twice as much energy as predicted. He offered a number of reasons for this. They generally came down to the analysis software making bad assumptions about how the building would be used. In the questions afterwards, Greig also suggested that architects could tweak the parameters to simulations to get desirable predictions (important for satisfying clients or regulatory bodies). To counter this, Greig is in the process of designing early stage analysis software that takes the performance data from built buildings and uses it to predict the performance of future buildings. 
Antony Viola, Adrian Smith + Gorden Gill
'Form Follows Performance' was the title of Anthony’s presentation. He showed a number of super-tall buildings and the analysis software needed to keep them standing up. A subject that came up in Anthony’s presentation, and a number of other presentations today, was analytic dashboards. More and more we are seeing these high level summations of data used to help inform design decisions and make tradeoffs in the design process. At Adrian Smith + Gorden Gill there were a number of dashboards used to help interdisciplinary teams optimise and hit performance targets.
Nathan Miller and Daniel Davis, CASE 
The positioning of Nate and my talk was quite fortuitous. The three speakers that had gone before us had all discussed how computation was making their design process more messy and more iterative. Nate and I expanded on this theme, arguing that the typical conception of BIM as a linear process (from big details to small) was being challenged by the inclusion of computation in the process. We showed a number of projects that CASE has worked on where the linear process either bifurcated into a branching of design explorations, or centred around measurement of design options, or twisted back upon itself to become iterative, or spread through a mesh of interoperability. In doing so we hopefully demonstrated how computation was not about building a good model but about facilitating a process that leads to good decisions that result in a better buildings. 
Scott Crowford, LMN
Scott showed off a number of acoustic tools LMN have used on recent projects. Many of these can be found on the LMN blog. Scott argued that if the projects are always changing, and the requirements always changing, the only place to achieve rigour was in the process. It was a good demonstration of how computation can be applied in other fields. For acoustic designers, their role has traditionally focused on analysis, and Scott’s talk showed how the inclusion of computation shifts the discussion to be much more about discovering possibilities. I expect Scott’s talk is prototypical of many others we will see in the coming years as more fields are folded into the computational design world. 
Jeff Vaglio, Enclos Studio
Jeff stepped through a stunning project, the Fulton Street Transit Centre. The most interesting part of this was how the parametric model was used to improve the project’s constructability. Jeff explained how the foundations in the model were flexible, so they could be precisely positioned once the surrounding structure was constructed. With the foundations in place and the form finalised, they then used the model to produce an extremely detailed animation of the project being constructed, which helped explain the process of construction to the contractor and the people tasked with doing the work. It was an interesting application that I’d never considered before but that has immediate benefits. 
Enric Ruiz-Geli, Cloud 9 
The session was concluded by Enric Ruiz-Geli. I saw Enric present at Smart Geometry two years ago and I am a huge fan of his work. He has a number of amazing project and a provocative outlook on the industry. He challenged architects to spend money on research rather than steel, arguing that the investment in research on the Media-ICT building recouped the cost by using 25% less steel than comparable buildings. He said many people at AU were talking about budgets, but none talked about the economy. With construction making up a largest portion of the world’s economy, it was important to take seriously the potentials for savings in an industry he clearly saw as wasteful. Enric walks the talk, and achieves impressive results. Enric concluded by saying the Carl Bass wasn’t the boss, the tree (and nature) is the boss, which brought the session full circle as Carl had spent some time discussing the possibilities designing architecture that grows like a seed. All exciting possibilities that left me wanting more. 
- Daniel Davis of CASE