Engineering design - does AI change the path of evolution in methods & teaching?

Woolman, Tim (2024) Engineering design - does AI change the path of evolution in methods & teaching? Bohemia, Erik, Bohemia, Erik, Buck, Lyndon, Grierson, Hilary, Kim, Jisun, Storer, Ian and Whitehead, Timothy (eds.) In Proceedings of the 26th International Conference on Engineering and Product Design Education: Rise of the Machines: Design Education in the Generative AI Era, E and PDE 2024. The Design Society. pp. 675-680 . (doi:10.35199/epde.2024.114).

Abstract

Engineering design tools are evolving, where AI can enhance speed and quality. Which of the current and future developments will change the teaching of engineering design? Text-to-image tools appear to be a precursor to text-to-design tools which will offer sensible-looking outputs in areas where many design precedents enable interpolation. However, where novelty is the goal, especially in new design spaces, human understanding of multivariate decisions, including where empathy for welfare and delight are considerations, is hard to express as merit functions for machines to learn. AI is already improving productivity in collating information relevant to discovering and defining performance requirements, and information on design precedents and available supplies. It can also improve the validity of simulation during the creation and refinement of design solutions and automate the application of engineering drawing language to a final design specification. We suggest engineering students' future selves will value their, i) learning to use human interaction to characterise performance requirements, ii) gaining knowledge as to how (well) and why existing engineering designs work, iii) practicing sketching as a means to visually communicate design ideas, iv) using CAD to experiment with state of the art modelling and simulation functionality while creating their own system designs, v) developing proficiency in engineering drawing language, and, vi) learning to make, break and tinker in workshops and to explore the potential & limitations of production. A familiar path, in which to add AI experiments.

This record has no associated files available for download.

Contributors

Download statistics