2025 ASEE Annual Conference & Exposition

The recent breakthrough in generative artificial intelligence (AI) exemplified by ChatGPT has
pushed scientific and education community to gradually adopt AI-based tools to augment and
partially automate research. While debates still exist on the scope and role that AI plays in the
research activities, there is a consensus that we should try to leverage the potential of AI
tools.

A literature review is essential in scientific research because it helps researchers understand the
current state of knowledge in their field, identify gaps or inconsistencies in previous studies, and
build on existing findings. It provides context and justification for the research by highlighting the
relevance of the topic and supporting the need for further investigation. Additionally, a thorough
literature review ensures that researchers avoid duplication of work, stay informed about
methodological advances, and frame their research within a broader scientific conversation. In the
past, the review is done manually and the volume of the literature needed to produce quality
summary is significant. This is a particularly daunting task for students or working professionals
with minimum experience, but it is also an area that is prime for AI applications.

As powerful as the AI tools, there are still imperfections and hurdles. The first is how to conduct
the review in the most accurate way, i.e., how to instruct the tool to sift through the literature and
extract the most relevant information. The second is the reliability of the AI returns. In this
work-in-progress paper we will present a case study on guiding groups of fresh engineering and
engineering technology students and professionals to conduct productive literature review using
AI-based tools. The supervising faculty will provide guidance, based on the latest research, on the
multiple steps of the review process, i.e., problem formulation, literature search, screening for
inclusion, quality assessment, data extraction, or data analysis and interpretation. The participants
will learn how to provide the proper instructions in each step, and also assess the accuracy of the
returns.