2026 ASEE Annual Conference & Exposition

We present our recent progress on the Manufacturing Adviser, a GPT-based question answering tool, designed to provide adaptive tutoring in manufacturing education. In our earlier work, we created a structured question-answer (Q&A) framework using WH-type prompts organized across four manufacturing related topics: process, sub-process, parameter, and design-for-manufacturing. The current phase introduces mechanisms for evaluating and adapting to learner expertise through reverse Q&A. The Adviser asks questions to gauge user understanding before or during an interaction. By evaluating the specificity, correctness, and reasoning within responses, the system estimates a learner’s proficiency and adapts subsequent questions posed at or asked by the users accordingly. This approach supports more effective scaffolding of learning content and enables the Adviser to act as a semi-autonomous tutor rather than a simple question-answering tool. This enhancement aims to make the Manufacturing Adviser more context-aware, and user-friendly to promote the education of and broaden participation in the future manufacturing workforce.