DesignGPT: Multi-Agent Collaboration in Design

Shiying Ding,Xinyi Chen,Yan Fang,Wenrui Liu,Yiwu Qiu,Chunlei Chai

2023-11-20

Abstract:Generative AI faces many challenges when entering the product design workflow, such as interface usability and interaction patterns. Therefore, based on design thinking and design process, we developed the DesignGPT multi-agent collaboration framework, which uses artificial intelligence agents to simulate the roles of different positions in the design company and allows human designers to collaborate with them in natural language. Experimental results show that compared with separate AI tools, DesignGPT improves the performance of designers, highlighting the potential of applying multi-agent systems that integrate design domain knowledge to product scheme design.

Artificial Intelligence,Human-Computer Interaction

What problem does this paper attempt to address?

The paper aims to address the challenges faced by generative AI in product design workflows, particularly issues related to interface usability and interaction patterns. The paper proposes the DesignGPT multi-agent collaboration framework, which allows human designers to collaborate with AI agents representing different roles in a design company using natural language. Experimental results show that DesignGPT improves designers' performance compared to using AI tools alone, highlighting the potential of integrating design domain knowledge into multi-agent systems to support product concept design. Specifically, the paper attempts to address the following points:

1. **Bridging the gap between design thinking and machine thinking**: By introducing standardized operating procedures (SOPs), the quality and consistency of generative AI outputs are improved.
2. **Supporting the entire design process**: Existing AI design tools often do not consider the complete design workflow, whereas DesignGPT supports the entire process from requirement analysis to final solution output through the collaboration of multiple AI roles.
3. **Improving design efficiency and quality**: Through multi-agent collaboration, design interruptions are reduced, and the novelty, completeness, and feasibility of design solutions are enhanced.

By constructing a multi-agent collaboration system that includes a knowledge base of design thinking and the design process, the paper demonstrates how DesignGPT can support designers in innovative design during the conceptual phase.