Computer Graphics Laboratory ETH Zurich

ETH
CGL » Research » Computational Design

Computational Design

Introduction

Our goal is to bridge the gap between the virtual and the real world, and to unleash the full potential of modern fabrication technology with the help of computational methods. We investigate novel algorithms and approaches for acquiring, simulating, and fabricating artifacts, structures, and robots. We also focus on the representation and intuitive editing of material properties as well as higher-level computational design tools, allowing to intuitively and quickly create custom content.

Topics

Robotic Creatures and Mechanical Characters

Animated mechanical characters have intrigued mankind for millennia, though it took the skill of an ingenious designer and engineer to create them in the past. With the advent of advanced manufacturing techniques and easy-to-use microcontrollers, the means for creating personalized mechanical toys, artifacts, or even robots has become widely available. However, without the assistance of specialized software, designing such mechanisms is still a virtually impossible task for casual users. We investigate the use of simulation and optimization to support the creation of individualized mechanical characters and robots.


Deformable Materials

How objects deform can be an essential part of their perception and function. Part of our research focuses on controlling or reproducing the deformation behavior of physical objects. We employ specialized optimization approaches and state-of-the-art fabrication methods, with a focus on 3D printing, to create multi-material objects or metamaterials with specific deformation properties.


Computational Design

Many traditional design processes involve of substantial amount of trial and error iterations, making the process slow and expensive, and discouraging complexity. Computational design approaches aim to alleviate these problems through computational methods. Visualization, simulation, and optimization are used to highlight design problems, predict the real-world behavior of objects without the need to build them, and automatically improve the design, so the designer can focus on the creative part of the design process. We have applied this computational design paradigm to a wide variety of design problems, from balloons and inflatables to furniture design and architecture.

Publications





© CGL, ETH Zurich