Agile conceptual design and validation based on multi-source product data and large language models: a review, framework, and outlook

The ability to rapidly and accurately identify high-value requirements and complete the conceptual design and performance validation of iterative products is crucial for companies to achieve swift product upgrades and enhance market competitiveness. This paper provides a comprehensive review of the latest methods and practices reported in the literature, covering various aspects such as requirement mining, scheme generation, and scheme validation in conceptual design. Furthermore, it proposes a framework for agile conceptual design and validation based on multi-source product data and Large Language Models (LLMs). The specific research focus areas include requirement mining using product usage data, AI-based conceptual scheme generation, and rapid conceptual scheme validation facilitated by physics-informed data-driven approaches. Finally, the paper discusses the challenges and future research prospects of agile conceptual design based on multi-source product data and Large Language Models.