Autodesk DesignScript released into the wild (Autodesk Labs) October 5

A few days back Robert Aish’s team at Autodesk finally released into the Autodesk Labs environment the scripting software now known as DesignScript. Check it out at Autodesk Labs. Some initial thoughts. 

I recall the buzz at the Autodesk University in 2008 when Robert arranged a Computational Design Symposium which I was fortunate to attend. At the conclusion of the symposium, he showed a proof-of-concept of something he tongue in cheek referred to as D-sharp (D# - Design-# , a tipping of the hat to C# - the programming language). An insight into how this demo was compiled in the short time since he had moved from the Bentley Generative Components (GC) team can be found here 

And a video interview with Robert from AU 2008  is available here. 

Most interesting to me was the presence for quite long periods during the day of the Computational Design symposium of Karl Bass the Autodesk CEO... It seemed to signal a degree of significance for the project from the senior management. It is almost four years later, and a working version is out.

I recall being surprised in 2008 at the choice of AutoCAD as the D# / Design Script platform. The origins of Aish's efforts with GC had been architectural. In an inspirational talk in 2000 or 2001 when he visited Wellington Robert Aish differentiated his research from Parametric Modelling of widgets such as was very familiar from 3D modelling within industrial design. He argued cogently for a language of design that permitted Architectural principles to be worked with explicitly. He showed how a major tour de force piece of architecture - the Waterloo Station terminus of the Eurostar high speed trains - might have been designed simply and efficiently had it been 'architecturally parametric'. The award winning Nicholas Grimshaw and Partners building snakes through London (though it has been empty since 2007):

http://en.wikipedia.org/wiki/File:London_Waterloo_former_Eurostar_station.JPG

http://en.wikipedia.org/wiki/File:Waterloo_International_after_closure_front.JPG

http://en.wikipedia.org/wiki/File:Waterloo_from_London_Eye.jpg

With the major promotion at AU 2008 of Revit as a BIM product (some over-zealous types even then seemed to be of the view that Revit=BIM), the Design Script platform being Autocad was a definite surprise.

The intriguing aspect of this new product, apart from the curiosity factor, is the proposed integration with other Design Analysis / Simulation tools:

• Robot (Structural Analysis)

• SMART Form (form finding) integration with DesignScript implemented by Buro Happold

• Optimisation Strategies

• Ecotect environmental analysis

• Green Building Studio (Building performance analysis and simulation)

Oddly, the Autodesk Labs web site links the 2009 AU interview with Robert Aish:

It also includes a brief 'White Paper' and a short tutorial. Not a lot to go on as yet. Disappointingly most of the tutorial is about running a program that produces no graphical output. I cannot see the folks who find Grasshopper intriguing because it hides this code from them being at all excited by this tutorial. The Design Script brochure follows the same approach. It spends much of its first page on a picture of a script and the following:

"Imperative programming is characterized by explicit ‘flow control’ using for loops (for iteration) and if statements (for conditionals) as found in familiar scripting and programming languages such as Python.
On the other hand, Associative programming uses the concept of graph dependencies to establish ‘flow control.’ Changes to ‘upstream’ variables are automatically propagated to ‘downstream’ variables." REF

Only a few rare designers are going to be interested in this...? 

Finally, I'm left with wondering where Vasari,  fits in this? Vasari has 'graduated' from Autodesk Labs to Beta release. It appears to be an integration of some aspects of GBS and Ecotect into Revit...? More exploration on our part?

The questions for R & D over the next 5 years will be: how might these all be integrated with other leading edge analytical developments (3DS Max Design, Radiance and Daysim climate based daylight analysis? IESVE Thermal, Lighting and Ventilation analysis? OpenStudio/EnergyPlus or just EnergyPlus? CATT Acoustic / EASE / Odeon ?)

And who will do the validation? 

• Vasari has a great 'virtual wind tunnel' feature, but it does not model the effect of the turbulent boundary layer that leads to downwash causing danger / discomfort at the base of poorly designed tall buildings. Wind tunnels were validated against measured data. I live in a really windy city with lots of opportunities for comparison of reality and virtual environments, but it took a long time for the wind tunnel testing back in the 1980s to collect enough data to demonstrate credible 'virtual wind' environments.
• Ecotect is essentially an extremely powerful simulation analysis results viewer coupled to hand calculations developed decades ago. How might these DesignScript or Vasari options be linked to and then validated with modern simulation software that addresses the physics more directly? And how might these analytical tools be applied at a time in the design cycle of a building when they might make a difference?
• 100 years or so in the case of Sabine's Reverberation Time in the Acoustics module
• 30-40 years ago in the case of the Admittance method for thermal performance analysis which assumes essentially that climate trends (temperature and sun) can be described as sine wave 'excitations' responded to by a building in such a way that the interior (temperature) environment  also follows a sine curve, somewhat delayed in time of the peak and reduced in terms of size of the peak by the amount of mass inside. 
• 50+ years ago in the case of Daylighting the BRE Split Flux Method was developed as a simple way of calculating how much light arrived directly from the sky and how much was reflected off the external environment: reflected light in the room is less well handled so complex lighting environments such as atria are not well covered...