COMPARING UNDERSTANDABILITY OF HAND SKETCHES VERSUS AI-GENERATED RENDERS FOR PRODUCT DESIGN

DS 131: Proceedings of the International Conference on Engineering and Product Design Education (E&PDE 2024)

Year: 2024
Editor: Grierson, Hilary; Bohemia, Erik; Buck, Lyndon
Author: Bartlett, Kristin Alicia; Mills, Jonathan Edward; Hagins, Carly
Series: E&PDE
Institution: University of Kentucky, United States of America
Page(s): 199 - 204
DOI number: 10.35199/EPDE.2024.34
ISBN: 978-1-912254-200
ISSN: 3005-4753

Abstract

Sketching to communicate design ideas is an important step in the design process. Image-generative artificial intelligence (AI) tools are increasing in prevalence and popularity, and these tools are being explored as aids in the design process. In this paper, we describe an evaluation of an image-generative AI tool, Vizcom, which uses a sketch-based input alongside a text prompt. We explored the use of this tool in a course with undergraduate product design students who were working on medical device concepts with teams of biomedical engineering students. We wanted to explore the possibility that using Vizcom might help facilitate communication between collaborators from different disciplines. In the course, students each drew five concept sketches by hand, and then used those same hand sketches as Vizcom inputs, resulting in Vizcom renders. Our analysis of these sketch/render pairs indicated that there was no significant difference in understandability between the hand sketches and the AI-generated renders. The characteristics of the hand sketches: line quality, proportionality, and understandability, were all positively correlated with the proportionality and understandability of the AI-generated renders. Our results suggest that the use of Vizcom did not reduce the need for strong hand-sketching skills in communicating design concepts, but Vizcom may offer some communication benefits. Results may be different for other types of design concepts, as novel medical devices are likely less represented in the datasets used to train Vizcom and other generative AI models.

Keywords: Artifical Intelligence, Product Design, Design Education, Medical Device Design, Design Sketching

Download

Please sign in to your account

This site uses cookies and other tracking technologies to assist with navigation and your ability to provide feedback, analyse your use of our products and services, assist with our promotional and marketing efforts, and provide content from third parties. Privacy Policy.