Scenario to specification: promises and pitfalls of AI in developing user-centered engineering specifications with interdisciplinary teams

Abstract

This study evaluates a framework for translating user scenarios into engineering specifications with AI integration. We conducted workshops with AI-assisted and non-AI teams to assess AI's impact on usability, efficiency, and collaboration. We collected data through surveys, interviews, and observations. Results indicate the framework is moderately to highly usable. While AI improved efficiency, it did not enhance output comprehensiveness or collaboration. Information overload and limited contextual understanding hindered AI integration. The study highlights AI's potential as a technical consultant and interdisciplinary bridge, emphasizing the need for domain-specific training and enhanced interactivity capabilities to optimize human-AI collaboration. These findings underscore AI's role in engineering design, contributing to scalable methods for developing user-centered specifications.