Foil: A Three-Dimensional Design Sketchbook

Introduction

Three-dimensional design tools are not for the faint of heart. In professional design environments such as Autodesk’s Maya, menu items carry cryptic names like “Jiggle Disk Cache Attributes” and even the most basic functions require a three-button mouse or multiple-keystroke commands. I have used tools of this kind for more than a decade, but have never found myself able to utilize them in the formation of my ideas. I generally work a design concept out on paper through a series of interrelated diagrams and only turn to software when I am ready to execute a final model. This experience led me to the notion that most computer-aided design software is geared towards the execution of an idea rather than the development of one.

To address this idea, I began a yearlong process of developing a tool called Foil for my thesis work at New York University’s Interactive Telecommunications Program (ITP). Foil is a three-dimensional design tool inspired by the freedoms of a paper notebook and enhanced by an arsenal of intelligent aids, which help the user to easily and elegantly develop rough conceptual sketches and turn these into a finished model.

Background & Research

The blank page is a tool of inestimable power - it is open-ended, a surface adaptable to all manners of thought and expression. In a sketch book, the artist or designer may tack on detail views, side notes and visual markers of his or her own invention, which help to illustrate the ideas under development. Despite its many virtues, this medium is limited in its ability to enable its user to reapply the same components in divergent variations. A digital tool, on the other hand, allows its user to save numerous versions of a project without harming the original, but tends to enforce a particular way of working.

Today, there are approximately four and a half million CAD users worldwide. This is less than half the number of Photoshop users. In the last two decades, digital image manipulation has become a commonplace activity of the mainstream computing culture. Three-dimensional design, on the other hand, has largely remained a niche activity. This is in part due to the fact that three-dimensional design tools are difficult to learn. Due to the conceptual complexities of three-dimensional geometry and the ambiguities associated with its representation in two dimensions, it is easy to understand why the development of a full-featured 3D design platform can become a protracted endeavor.

Design Software - A Place to Think

In developing Foil, I sought inspiration from my own sketchbooks, from my childhood and more recent years, as well as from the notebooks and manuscripts of some of my intellectual heros. This latter group included the technical notebooks of Leonardo da Vinci and Thomas Edison as well as the manuscript drafts of artists from various disciplines such as Beethoven and Allen Ginsberg. One of the most important themes I found in these works was the tendency to organize the space around a central diagram or version of the work and then use lines, arrows and other such notations to connect an aspect of this central version with an auxiliary space in which some key detail or divergent concept could be further explored in temporary isolation.

Every block of stone has a statue inside it and it is the task of the sculptor to discover it.

— Michelangelo

User Interface

Foil’s user interface attempts to bring the best features of paper notebooks and computer-aided design environments into one easy-to-use software platform. It is meant to serve the user’s own creative process rather than circumvent it. A Foil project has four main interface elements: text notes, two-dimensional diagram views, three-dimensional diagram views (also called “detail views”) and the main project view.

I developed Foil in C++ for Mac OS X and the iPad. Due to the novelty of some of Foil’s interface elements and because it is of great importance for me to understand every aspect of my software, I chose to write Foil entirely from scratch with no third-party libraries with the two exceptions of OpenGL and Cinder for multi-platform display windowing and rendering purposes only.

Machine Assistance Tools

In recent years, computer scientists have begun to apply powerful machine-learning algorithms to computer-aided design processes. One of the most noteworthy examples of this is NASA Evolvable Systems Research Group’s work on the use of genetic algorithms in the design optimization of radio frequency antennas. Foil takes a unique perspective on the role of machine-learning algorithms in design software. Its purpose within Foil is twofold: it is designed to assist the user in keeping track of the various components, diagrams and versions of a design as well as to quickly generate variations of the user’s designs in order to offer the user a simple and streamlined way of refining a diagram.

Personal Statement / Prior Work

I came to ITP with a lifelong interest in virtual reality and three-dimensional graphics. At the age of eight, I began producing a series of aluminum foil mock-ups for virtual reality headsets. This project continued for several years and across numerous design iterations, each one pointed at overcoming a design limitation of the previous.

My thesis work on Foil has been a direct extension of my thinking about the DEcomp and Digitalis projects. I find forced perspective to be a fascinating concept and the relationship between DEcomp and Digitalis afforded me an important opportunity in the development of my thinking. Ultimately though, I came to believe that forced perspective, while quite interesting to me, was too narrow an application to warrant a full-featured three-dimensional environment. With Foil, I wanted to apply some of DEcomp’s core concepts to more general problems in design.

Conclusion

Through a thoughtfully designed user-interface and with the aid of machine-intelligence techniques, I believe it is possible to distill three-dimensional digital design into a creative and enjoyable process that is accessible to everyone. This project represents the latest iteration of my lifelong interest in three-dimensional representation and in systems that aid people in understanding the world around them and the possible worlds they might imagine.

Acknowledgements

This project would not have been possible without my thesis advisors Nancy Hechinger and Marina Zurkow, my ITP advisor Daniel Shiffman, Danny Rozin, Dan O’Sullivan, Eric Rosenthal, Jared Schiffman, David Nolen, Douglas Rushkoff, Heather Dewey-Hagborg, my fellow students at ITP and my family.