From Fragmented Inputs to Actionable Design Intent

Design teams rarely start with a complete and coherent requirements package. Instead, they work with fragmented inputs such as briefs, standards, legacy specifications, expert assumptions, performance targets, and evolving stakeholder constraints. This session will explore how AI-enabled workflows combining information normalization, uncertainty quantification, human-in-the-loop review, retrieval-augmented generation, threshold-based routing, and decomposition/aggregation can convert incomplete information into actionable, traceable design intent. Although developed in product design and simulation, the insights directly transfer across industries.