The task of material selection plays a pivotal role in a large number of industries, from manufacturing to construction. Material selection is usually carried out after several cycles of conceptual design, during which designers iteratively refine the design solution and the intended manufacturing approach. In design research, material selection is typically treated as an optimization problem with one correct answer, and restricted to specific types of objects or design functions. In this work, we instead treat materials selection as a problem with multiple answers with varying feasibility. We collect a dataset of experts' material preferences for nine material categories across 16 diverse design scenarios, and report some trends in the data.
Recent advancements in the natural language processing field around transformer-based architectures, leading to the development of a multitude of pre-trained Large Language Models (LLMs) capable of understanding human language and exhibiting reasoning behaviors, raise some questions about the potential use of this technology for automating the material selection, especially during early the early conceptual design phase. In this work, we identify the degree to which material recommendations provided by an LLM align with those of expert human designers. The results underscore the potential of LLMs to inform material selection.
📦 LLM-for-Material-Selection
├─ Comparison Notebooks
├─ Dataset
└─ Scripts