Computational Design and Evaluation Methods for Empowering Non-Experts in Digital Fabrication

TL;DR

This paper introduces computational methods to enable non-experts to design and evaluate complex 3D shapes in digital fabrication through generative modeling, physics-based optimization, and crowdsourcing techniques.

Contribution

It presents novel algorithms for shape synthesis, functional design optimization, and crowdsourced assessment to support non-expert digital fabrication.

Findings

Generative shape modeling facilitates user-driven modifications.

Physics-based shape optimization improves functional compliance.

Crowdsourcing effectively assesses complex design problems.

Abstract

Despite the increasing availability of personal fabrication hardware and services, the true potential of digital fabrication remains unrealized due to lack of computational techniques that can support 3D shape design by non-experts. This work develops computational methods that address two key aspects of content creation:(1) Function-driven design synthesis, (2) Design assessment. For design synthesis, a generative shape modeling algorithm that facilitates automatic geometry synthesis and user-driven modification for non-experts is introduced. A critical observation that arises from this study is that the most geometrical specifications are dictated by functional requirements. To support design by high-level functional prescriptions, a physics based shape optimization method for compliant coupling behavior design has been developed. In line with this idea, producing complex 3D surfaces from flat 2D sheets by exploiting the concept of buckling beams has also been explored. Effective design assessment, the second key aspect, becomes critical for problems in which computational solutions do not exist. For these problems, this work proposes crowdsourcing as a way to empower non-experts in esoteric design domains that traditionally require expertise and specialized knowledge.