What question did this study set out to answer?

The study aims to develop a framework that integrates affective and functional requirements in product design using AI techniques.

March 18, 2026Open Access

Data–driven product form design integrating affective and functional requirements via TGRA–QFD and Stable Diffusion

Key Points

The study aims to develop a framework that integrates affective and functional requirements in product design using AI techniques.
Mined user reviews to extract affective and functional requirements.
Applied Rank Correlation Analysis to calculate weights of affective requirements.
Developed Time-weighted Grey Relational Analysis to determine functional requirements weights.
Utilized Quality Function Deployment to map requirements to design features.
Conducted VR eye-tracking to evaluate and select optimal product designs.
The proposed model significantly improved design efficiency.
Strong consistency was found between VR evaluations and subjective questionnaire responses.
The aesthetic measurement formula effectively identified the best form for the product.

Abstract

The integration of affective and functional requirements has become a critical issue in achieving user-centered and competitive product design. Traditional approaches tend to focus on either affective or functional aspects in isolation, resulting in an incomplete understanding of user requirements. To address this gap, this study proposes an AI-driven product form design framework that simultaneously extracts and integrates affective requirements (ARs) and functional requirements (FRs) from online reviews. The research involves four stages. First, user reviews of the target product are mined to extract ARs and FRs. Then, the weights of the ARs are calculated based on the Rank Correlation Analysis (RCA), and an innovative Time-weighted Grey Relational Analysis (TGRA) method is proposed to determine the weights of the FRs by considering the influence of temporal factors. Second, Quality Function Deployment (QFD) maps the ARs and FRs to concrete design features (DF). Third, these features are converted into structured prompts for the Stable Diffusion Model (SDM) to produce product concept sketches. Finally, a VR eye-tracking experiment provides both sensory and cognitive evaluations to identify the optimal concept. A case study on an aerial-photography drone (APD) selects the best form using an aesthetic measurement formula and validates it through VR eye-tracking. The results indicate a strong consistency between the experimental findings and the subjective questionnaire outcomes, confirming the effectiveness of the proposed design model. This model enhances design efficiency while ensuring the visual aesthetics of the product.

Bookmark

View Full Paper