An Assessment Framework for Human-Machine Collaboration Design: Analyzing the Influence of Different Collaboration Modes on E-Bike Design

This study presents a method for assessing human-machine collaboration design and investigates the effects of three collaboration modes on design outcomes and the design scenarios adapted by these modes. Nine designers were invited to engage in the electric bicycle design project, and the designs were assessed using qualitative and quantitative methodologies. The research findings demonstrate that the collaborative co-creation mode excels in innovation and feasibility, rendering it suitable for intricate innovation projects; the generative-driven mode outperforms in sustaining design goal consistency, making it appropriate for preliminary exploration during the conceptual design phase; the human-centered design mode is constrained by designers' competencies, exhibits overall weak performance, and is suitable for proficient designers to swiftly visualize, verify, and optimize solutions. The study emphasizes the imperative for designers to develop transdisciplinary abilities to effectively assess and incorporate AI-generated design results, while the proposed framework offers significant theoretical direction and practical insights for employing AI tools in design practice.