Session: DTM-02-01: Artificial Intelligence in Design

Paper Number: 116576

116576 - Effect of Reflection and Incubation on Artificial-Intelligence-Assisted Design Space Exploration 

Design space exploration is a method for gaining insights into a space of design alternatives to make important design decisions. Current research in design space exploration is studying the use of Artificial Intelligence (AI) agents (called cognitive assistants) to provide cognitive support to human designers and help them obtain insights from large datasets. This work explores the collaborative exploration of design spaces between human designers and AI cognitive assistants. A human subject study was conducted to 1) investigate whether cognitive interventions during a design task, specifically a question-based reflection period or an incubation period, influence the way designers collaborate with cognitive assistants; and 2) characterize the effects of those interventions on the design space exploration process and outcomes (performance and designer learning). The study aims to determine whether cognitive interventions that prompt designers to reflect on their collaboration with the AI agent could improve the quality of their collaboration. The results indicate that the two intervention methods have opposing effects on design performance. Participants who had an incubation period amid design space exploration improved their design performance significantly more than participants who had a self-reflection period amid design space exploration. Participants who had different interventions during the design task exhibited different forms of human-AI collaboration which possibly explain the opposing effects of reflection and incubation periods on their design performance. The findings could serve as a basis for the development of AI-assisted design tools that guide designers toward more effective human-AI collaboration.

Presenting Author: Hyeonik Song Texas A&M University

Presenting Author Biography: Hyeonik is a postdoctoral researcher in the System Engineering, Architecture and Knowledge Lab of the Aerospace Engineering Department at Texas A&M University. He received a B.S. (2016), M.S. (2018), and Ph.D. (2021) in Mechanical Engineering from the Georgia Institute of Technology. His research focuses on computational methods and tools for the conceptual design process with a particular focus on design-by-analogy, human-AI collaboration, and design space exploration. His research seeks to understand designers’ cognitive processes with the goal of developing systematic and reliable design tools that result in innovative and efficient design practices.

Authors:

Hyeonik Song Texas A&M University
Antoni Viros-I-Martin Texas A&M University
Daniel Selva Texas A&M University