Board 258: Engineering Design Thinking in the Age of Generative Artificial Intelligence

John Clay; Xingang Li; Molly H Goldstein; H. Onan Demirel; Darya Zabelina; Charles Xie; Zhenghui Sha

Download Paper | Permalink

Conference: 2024 ASEE Annual Conference & Exposition
Location: Portland, Oregon
Publication Date: June 23, 2024
Start Date: June 23, 2024
End Date: June 26, 2024
Conference Session: NSF Grantees Poster Session
Tagged Topic: NSF Grantees Poster Session
Page Count: 8
DOI: 10.18260/1-2--46830
Permanent URL: https://peer.asee.org/46830
Download Count: 157

Paper Authors

biography

John Clay University of Texas at Austin

visit author page

Research Assistant

visit author page

biography

Xingang Li University of Texas at Austin

visit author page

Xingang Li is a Ph.D. candidate working as a Research Assistant in the Walker Department of Mechanical Engineering at the University of Texas at Austin. His research interests include generative design, deep learning for engineering design, and human-AI design collaboration. He received the Philip C. and Linda L. Lewis Foundation Graduate Fellowship in Mechanical Engineering from the Cockrell School of Engineering for 2022-2023.

visit author page

biography

Molly H Goldstein University of Illinois Urbana-Champaign orcid.org/0000-0002-2382-4745

visit author page

Dr. Molly H. Goldstein is a Teaching Assistant Professor and Product Design Lab Director in Industrial and Enterprise Systems Engineering at the Grainger College at the University of Illinois. She is also courtesy faculty in Mechanical Science and Engineering, Curriculum & Instruction (College of Education) and Industrial Design (School of Fine and Applied Arts). Dr. Goldstein’s research focuses on student designers through the study of their design actions and thinking.

visit author page

biography

H. Onan Demirel Oregon State University

visit author page

Onan Demirel is an assistant professor of mechanical engineering at Oregon State University. He received his Ph.D., MS, and BS degrees from Purdue University. Dr. Demirel’s research interests lie at the intersection of human factors engineering, engineering design, and systems engineering.
His research focuses on understanding the human element in the design process, advancing Digital Human Modeling (DHM) theory and practice, and developing multi-disciplinary design frameworks to explore inter-dependencies and co-evolution of the human element in engineering systems. Dr. Demirel’s work includes developing computational and experimental human-centered design theory and methodology that incorporate human factors engineering design principles, particularly in the early phases of the design cycle.

visit author page

author page

Dr. Zhenghui Sha is an Assistant Professor in the Walker Department of Mechanical Engineering at the University of Texas at Austin. Dr. Sha’s research focuses on system science and design science, as well as the intersection between these two areas. Dr. Sha is the recipient of the 2022 Young Engineering Award (YEA) from the Computers & Information in Engineering (CIE) Division of the American Society of Mechanical Engineers (ASME) and received the Best Dissertation of The Year Award in 2017 from the ASME CIE Division.

visit author page

Download Paper | Permalink

Abstract

Generative design (GD) is an artificial intelligence (AI) based design method that uses generative systems and algorithms to automatically create design artifacts by considering objectives, parameter ranges, and constraints defined by human designers. The role of the human designer in the GD process is to translate the design problem so that an AI agent may understand and explore potential solutions in a design space with defined constraints and parameter interactions. The human designer will then evaluate AI-generated designs along the Pareto front and the objective space influencing design synthesis. Therefore, generative designers engage in inverse thinking from the objective space to the design space. However, in traditional design (TD) and parametric design (PD), the human designer is responsible for design space exploration, often via cognitive idea generation and also the evaluation of the human-generated designs, which is a typical forward-thinking direction from design space to objective space. Consequently, the role of designers in the GD process fundamentally differs from the TD and PD processes and thus requires the designer to engage in a different type of design thinking, i.e., the cognitive processes activated during design. Yet, little research has been conducted to outline the cognitive processes activated during GD.

Our goal is to define GDT as the most recent evolution of engineering design thinking by systematically reviewing the concepts central to the proposed Evolving Design Thinking (EDT) model. The EDT model is a meta-representation of the evolution of design thinking from traditional design thinking (TDT) to parametric design thinking (PDT) and to Generative design thinking (GDT) across three levels: design thinking, design technologies, and design cognition. The EDT model guides our review and motivates two research questions (RQs). RQ1: What cognitive processes are activated by designers using traditional/parametric design? To answer RQ1, we review the technologies and cognitive processes activated during traditional and parametric design (i.e., TDT and PDT) and highlight how technologies shape design cognition. RQ2: What cognitive processes are activated by designers using generative design? To answer RQ2, we systematically review the technologies available to generative designers and consider the influence of technology on cognition shown through RQ1. Finally, a definition of GDT will be synthesized by considering the evolution from traditional to parametric to generative design and by considering the cognitive processes of TDT and PDT.

A definition of GDT that highlights the GD-relevant cognitive processes is expected to generate significant impacts on design education and research. For example, the research outcomes can potentially improve future generative designers’ education, as current GD curricula have been developed without insights into how generative designers think. The systematic review will also allow future research to leverage our insights and apply neurocognitive methods to further explore GDT, for example, by conducting experimental research to verify that the identified cognitive processes are carried out by generative designers, by devising methods for facilitating these cognitive processes and developing psychometric tests to measure GDT efficacy.

Citation
Format

Clay, J., & Li, X., & Goldstein, M. H., & Demirel, H. O., & Zabelina, D., & Xie, C., & Sha, Z. (2024, June), Board 258: Engineering Design Thinking in the Age of Generative Artificial Intelligence Paper presented at 2024 ASEE Annual Conference & Exposition, Portland, Oregon. 10.18260/1-2--46830

TY - CPAPER
AB - Generative design (GD) is an artificial intelligence (AI) based design method that uses generative systems and algorithms to automatically create design artifacts by considering objectives, parameter ranges, and constraints defined by human designers. The role of the human designer in the GD process is to translate the design problem so that an AI agent may understand and explore potential solutions in a design space with defined constraints and parameter interactions. The human designer will then evaluate AI-generated designs along the Pareto front and the objective space influencing design synthesis. Therefore, generative designers engage in inverse thinking from the objective space to the design space. However, in traditional design (TD) and parametric design (PD), the human designer is responsible for design space exploration, often via cognitive idea generation and also the evaluation of the human-generated designs, which is a typical forward-thinking direction from design space to objective space. Consequently, the role of designers in the GD process fundamentally differs from the TD and PD processes and thus requires the designer to engage in a different type of design thinking, i.e., the cognitive processes activated during design. Yet, little research has been conducted to outline the cognitive processes activated during GD.

Our goal is to define GDT as the most recent evolution of engineering design thinking by systematically reviewing the concepts central to the proposed Evolving Design Thinking (EDT) model. The EDT model is a meta-representation of the evolution of design thinking from traditional design thinking (TDT) to parametric design thinking (PDT) and to Generative design thinking (GDT) across three levels: design thinking, design technologies, and design cognition. The EDT model guides our review and motivates two research questions (RQs). RQ1: What cognitive processes are activated by designers using traditional/parametric design? To answer RQ1, we review the technologies and cognitive processes activated during traditional and parametric design (i.e., TDT and PDT) and highlight how technologies shape design cognition. RQ2: What cognitive processes are activated by designers using generative design? To answer RQ2, we systematically review the technologies available to generative designers and consider the influence of technology on cognition shown through RQ1. Finally, a definition of GDT will be synthesized by considering the evolution from traditional to parametric to generative design and by considering the cognitive processes of TDT and PDT.

A definition of GDT that highlights the GD-relevant cognitive processes is expected to generate significant impacts on design education and research. For example, the research outcomes can potentially improve future generative designers’ education, as current GD curricula have been developed without insights into how generative designers think. The systematic review will also allow future research to leverage our insights and apply neurocognitive methods to further explore GDT, for example, by conducting experimental research to verify that the identified cognitive processes are carried out by generative designers, by devising methods for facilitating these cognitive processes and developing psychometric tests to measure GDT efficacy.
AU - John Clay
AU - Xingang Li
AU - Molly H Goldstein
AU - H. Onan Demirel
AU - Darya Zabelina
AU - Charles Xie
AU - Zhenghui Sha
CY - Portland, Oregon
DA - 2024/06/23
PB - ASEE Conferences
TI - Board 258: Engineering Design Thinking in the Age of Generative Artificial Intelligence
UR - https://peer.asee.org/46830
DO - 10.18260/1-2--46830
ER -