Translating between the virtual and the real: Interview with artist Scott Kildall

Michael Ang

February 13th, 2015

Scott Kildall is an artist whose work often explores themes of future-thinking and translation between the virtual and the real. His latest projects use physical data visualization - the transformation of data sources into physical objects via computer algorithms. We're currently collaborating on the Polygon Construction Kit, a software toolkit for building physical polygon structures from wireframe 3D models. I caught up with Scott to ask him about his work and how 3D printing and digital fabrication are changing the production of artwork.

How would you describe your work?
I write computer algorithms that generate physical sculptures from various datasets. This has been a recent shift in my art practice. Just five years ago, digital fabrication techniques, 3D printing, CNC machinery and other forms of advanced fabrication were simply too expensive for artists. Specifically, I've been diving into what the media ominously calls "big data," which entails thousands upon thousands of data points ranging from city infrastructure data to biometric feedback. From various datasets, I have been generating 3D-printed sculptures.

Water Works - Imaginary Drinking Hydrants (2014) 3D-Printed Sculpture with laser-etched wood map

What are some of the tools that you use?
I write my own software code from the ground up to both optimize the production process, and create a unique look for my work. My weapon of choice is openFrameworks, a C++ toolkit that is relatively easy to use for a seasoned applications programmer. The other open source tool I use is Processing, which is a quick and dirty way to prototype ideas. Python, my new favorite language, is excellent for transforming and cleaning datasets, which is the not-so-glamorous side of making "data art".

You've just completed some residencies and fellowships, can you tell us about those?
In 2014, I was an Artist In Residence at Autodesk in San Francisco, where I live. Autodesk has an amazing shop facility including 6 state-of-the art Objet 500 printers. The resulting prints are resin-based and capture accurate details. During a several month period, I was able to iteratively experiment with 3D printing at a rate that was much faster than maintaining my own extrusion 3D printer.

Data Crystals (2014) Incidents of Crime Data from San Francisco

The first project I worked on is called Data Crystals, which uses public data sets from the city government of San Francisco, which anyone can download from the data portal at The city's open data includes all sorts of goodies such as geolocated points for incidents of crime and every parking meter in the city. I mapped various data points on an x-y plane using the latitude and longitude coordinates. The z-plane was then a dimension of time or space. To generate the "Crime Data" crystal, I worked with over 30,000 data points. My code represented each data points as a simple cube with the size being proportional to the severity of the crime. I then ran clustering algorithms to create one cohesive object, which I call a "crystal", like a synthetic rock that a data miner might find.

In a sense you're mining an abstract data source into a physical object...
It was more like finding a concrete data source and then turning it into an abstract physical object. With conventional 2D data visualizations you can clearly see where the hotspots of crime, or other data points, might be on a map. However, the Data Crystals favor aesthetics over legibility.

The central question I wanted to answer was "what does data look like?" When people create screen-based data visualizations, they focus on what story to tell. I was intrigued by the abstract data itself and so made art objects which you could look at from different vantage points.

What is it about having the data occupy a physical space that's important to you?
When data occupies a physical space, it is static, like a snapshot in time. Rather than controlling time, like you would on a slider with a screen-based visualization, you can examine the minutiae of the physical characteristics. The data itself invites a level of curiosity that you don't get with a mediated screen-based interaction. Real objects tap into the power of conventional perception, which is innate to how our brains interact with the world.

Tell us a bit about your series of sculptures that were created in the virtual world of Second Life and then physically created using Pepakura software and papercraft techniques.
An earlier project that I worked on, in collaboration with my partner Victoria Scott, is No Matter (2008). This project was instrumental in my development about how to transform the imaginary into the real. For this project, we worked with the concept of imaginary objects, which are things that have never physically been built, but exist in our shared imagination. They include items from mythology like the Holy Grail or the Trojan Horse, from fiction like the Maltese Falcon or the Yellow submarine, or impossible objects/thought experiments like the Time Machine or Schrodinger's Cat. We constructed these objects in the imaginary world of Second Life and then extracted them as "digital plunder" then rebuilt them as paper sculptures in real space.

No Matter (2008) Second Life Installation at Ars Virtua

No Matter (2008) Yellow Submarine paper sculpture

Because they were paper sculptures that were physically fabricated, there were physical constraints such that the forms themselves had to be vastly simplified to smaller faceted objects. The collision between this faceted proto-object with beautiful high resolution prints resonate with most viewers on an aesthetic level. Working with that kind of virtual space led me to thinking about the question of data -- could this intangible "thing" also be represented materiality?

No Matter (2008) Installation at Huret & Spector Gallery

What are you working on now?
I'm developing a new project called Machine Data Dreams, which examines the question of "how do machines think?" To make computers function, humans program code in languages such as JavaScript or Python or C++. There are whole sets of people that are literate with these machine languages, while most others in the world don't "speak" them. Knowing how to code gives you power and money, though usually not prestige. However, understanding how machines process language will be increasingly important as they will undoubtedly be increasingly integrated with human biology.

My proposition is to create a room-based installation that reflects the structure of language and how machines might view the world through a datasets representing machine syntax. What I will be doing is taking machine language (e.g. Javascript, Python, C++) and translating that to a language-based data sets. From these, I will algorithmically generate a cavelike 3D model with triangulated faces.

The Polycon Construction Kit -- which you developed -- will be instrumental in making this happen. Last year, I had sketched out ideas in my notebook about creating custom 3D-printed connectors for large-scale sculptural installations, and then I found out that you had already been working on this technology. So, thank you for inviting me to collaborate on Polycon! I'm grateful to be figuring out how to do this with a trusted colleague.

What are some of the trends or new technologies that you're excited about?
There's so much happening in field of digital fabrication! For example, 3D printing technology has so much to offer now, and we're at a pioneering stage, akin to Photoshop 1.0. Artists are experimenting with the medium and can shape the dialog of what the medium itself is about. It's clear to me that 3D printing / 3D fabrication will be a larger part of our economy and our universe. It fundamentally changes the way materials are produced.

Many have made this observation, but it is still understated. 3D printing is redefining the paradigm of material production, which also affects art production and factory production. This is how capitalist-based economies will operate in the next 20 or 30 years.

From an artistic standpoint, working with digital fabrication technology has changed the way I think about sculpture. For example, I can create something with code that I never thought was even possible. I don't even know what the forms will look like when I write the code. The code generates forms and I find unexpected results ranging from the amazing to the mediocre to the crappy. Then I can tweak the algorithms to focus on what works best.

You've had a chance to work with some of the most advanced technologies through the Autodesk residency. Now you have your own 3D printer in your garage. Do you see a difference in your creative process using the really high-end machines versus something you can have in your garage?

Working with the high-end machines is incredible because it gives you access to making things as perfect as you can make them. I went from working with printers that probably cost a half a million dollars to a printer that I got for $450.

Like half a million to half a thousand?
Yes! The Autodesk printers are three orders of magnitude more expensive. The two types of printers have vastly different results. With my garage setup, I have the familiar tales: messed up 3D prints and the aches and pains of limit switches, belts and stepper motors. I've been interested in exploiting the glitches and mistakes. What happens when you get bad data? What happens when you get glitchy material errata?

My small extrusion printer gives me a lot more appreciation for fixing things, but when making something precise I'd much rather work with the high-quality 3D printers.

Where can we see more of your work?
All my work is on my website at My blog at has my current thought processes. You can follow me on Twitter at @kildall

About the Author:

Michael Ang is a Berlin-based artist and engineer working at the intersection of art, engineering, and the natural world. He is the creator of the Polygon Construction Kit, a toolkit for bridging the virtual and physical worlds by translating simple 3D models into physical structures.

comments powered by Disqus