Value of Verbalizations in 3D CAD Workflows

This research article investigates the role of think-aloud computing in enhancing AI assistance for 3D computer-aided design (CAD) workflows. The study involves a three-part experimental design with ten designers and ten experts, focusing on capturing designers' verbalizations during CAD modeling tasks. The first part records think-aloud verbalizations to uncover designers' intent, while the second part compares expert feedback provided with and without access to these verbalizations. The final part includes interviews with designers to evaluate the quality of the feedback received. Findings indicate that verbalizations reveal essential insights into design rationale, future actions, and challenges that are not captured by traditional geometric data. This study highlights the potential of think-aloud data to inform context-aware AI tools, improving the feedback process and aligning expert guidance with designers' goals. The implications for future AI assistance in CAD are discussed, emphasizing the need for intelligent prompts to facilitate effective communication of design intent.