An Agentic AI Framework for Form Design through Implicit Visual Attention

Conventional product form design is a cognitively demanding process, fundamentally limited by its time-consuming nature, reliance on the subjective expertise and skill of the designer, and the opaque translation of inspiration into tangible concepts. This creates a significant productivity barrier, particularly for novice designers. This research introduces a novel, attention-aware framework that restructures this creative workflow by introducing two synergistic systems: EUPHORIA, an immersive Virtual Reality-based moodboarding environment that uses eye-tracking to implicitly capture a designer's aesthetic preferences, and RETINA, an agentic AI pipeline that translates these preferences into concrete form design outputs. The foundational principles were empirically established through two studies. The first study confirmed a positive correlation between a user's implicit attention and their explicit preference. Subsequent study demonstrated that this attention could be directed with external stimuli, showing a strong convergence in visual selections. A comparative study was conducted where four designers solved challenging form design problems through four distinct workflows, ranging from a fully manual conventional method to the proposed end-to-end automated method. The outcomes of each workflow were then assessed by 50 design experts. A worthiness score and a design effectiveness score were calculated based on the rankings and ratings given by the experts, respectively. These scores indicate a substantial improvement in effectiveness and productivity for the end-to-end automated workflow compared to the conventional form design process. Thus, the framework has the potential to democratize the form design endeavour by enabling novice designers to implicitly guide the computers in producing good designs.