EXPLORING THE USABILITY OF AI PROMPTS FOR MECHANICAL ENGINEERING DESIGN DEVELOPMENT
Year: 2024
Editor: Grierson, Hilary; Bohemia, Erik; Buck, Lyndon
Author: Vuletic, Tijana; Ashcroft, Aaliyah
Series: E&PDE
Institution: University of Glasgow, United Kingdom
Page(s): 461 - 466
DOI number: 10.35199/EPDE.2024.78
ISBN: 978-1-912254-200
ISSN: 3005-4753
Abstract
Computational support for concept exploration has been a research development in the last two decades, and the ascent of AI tools such as Chat GPT and generative design is further expanding it. Mechanical engineering tends to focus on more stringent functional requirements and compliance with regulation, and while the potential of AI developments to support it exists, it also likely requires significant oversight by the human. The levels of helpfulness/need for oversight and checks by humans are explored in terms of where AI could be beneficial in the mechanical engineering design process, how this could be supported, and how reliable can it be in that context. Exploring the use of AI in mechanical engineering design and the level of trust placed on its outcomes is an important question for future engineering design development, especially since AI and machine learning are improving exponentially. Exploration of students’ perception of AI and its objective usefulness will be contrasted, by performing a comparison between designs conducted with and without AI support. AI tools used will focus on product design specification generation, concept image generation and generative design. Then recommendations will be given on tools that are currently considered to be helpful for mechanical design development, highlighting the positives and negatives of the approach using AI and potential for adoption of AI in engineering design education.
Keywords: Mechanical Engineering Design; Education; AI; Generative Design