Review of current and potential uses of large language models in engineering

Abstract

Background Large Language Models (LLMs) have emerged as transformative tools in engineering, offering capabilities that streamline complex processes and support decision-making across diverse disciplines. Despite notable advancements in applications such as robotics task planning, autonomous driving, program repair, and technical documentation, challenges persist concerning ethical considerations, transparency, and accountability in safety-critical systems. Purpose This study aims to conduct a systematic literature review (SLR) on the applications, challenges, and ethical implications of LLMs in engineering. The objective is to synthesize existing knowledge and identify research gaps to guide future investigations. Method A comprehensive review of peer-reviewed publications from 2014 to 2024 was conducted, resulting in the selection of 23 relevant articles. These articles were classified into five thematic categories: automation of complex engineering tasks, knowledge generation and discovery, enhancing engineering education, ethical considerations and challenges, and integration with real-world engineering practices. Results The review highlighted (i) increasing interest in LLM applications across multiple engineering domains, (ii) a growing emphasis on ethical and regulatory concerns related to LLM adoption, (iii) significant potential for enhancing productivity and fostering innovation, and (iv) a critical need for interdisciplinary collaboration to address reliability and scalability challenges. Conclusions LLMs hold considerable promises for advancing engineering practices by automating tasks, facilitating knowledge discovery, and supporting education. However, ensuring ethical deployment, transparency, and model reliability remains essential. Future research should focus on developing frameworks for responsible AI adoption and fostering interdisciplinary efforts to overcome existing limitations.