• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2026, Vol. 62 ›› Issue (7): 324-342.doi: 10.3901/JME.260381

• 数字化设计与制造 • 上一篇    

扫码分享

人智协同设计中的信任问题:挑战、进展和展望

罗仕鉴, 郭和睿, 李庆龄, 易珮琦   

  1. 浙江大学计算机科学与技术学院 杭州 310027
  • 收稿日期:2025-04-21 修回日期:2025-09-06 发布日期:2026-05-25
  • 作者简介:罗仕鉴,男,1974年出生,博士,教授,博士研究生导师。主要研究方向为工业设计、智能设计、服务体验设计。E-mail:sjluo@zju.edu.cn
    郭和睿(通信作者),男,2000年出生,博士研究生。主要研究方向为用户体验与产品创新设计,智能设计,设计评价。E-mail:heruiguo@zju.edu.cn
    李庆龄,女,2000年出生,硕士研究生。主要研究方向为用户体验与产品创新设计,交互设计,智能设计。E-mail:22360475@zju.edu.cn
    易珮琦,女,2001年出生,博士研究生。主要研究方向为用户体验与产品创新设计,交互设计,智能设计。E-mail:yipeiqi@zju.edu.cn
  • 基金资助:
    国家自然科学基金资助项目(52475287)。

Trust Issues in Design for Human-AI Collaboration: Challenges, Progress and Prospects

LUO Shijian, GUO Herui, LI Qingling, YI Peiqi   

  1. College of Computer Science and Technology, Zhejiang University, Hangzhou 310027
  • Received:2025-04-21 Revised:2025-09-06 Published:2026-05-25

摘要: 随着生成式人工智能技术的不断发展,面向垂直领域的人工智能生成设计成为新的范式,其中人类-人工智能组队工作中的信任关系对设计结果的创新性、效率与质量有重要影响。然而,当前相关研究缺少完整、系统的讨论。因此为了明确人智协同设计中信任关系相关研究的趋势、热点和方法,从信任的研究发展出发,通过对文献的聚类和分析,定义了信任的影响因素;立足设计领域,构建人智协同设计信任层次模型,系统探讨技术、交互与体验等三个方面的研究内容与体系结构,总结了信任测量、校准、修复、决策与评价的关键技术。从理论、方法、技术、系统和实践五个方面,提出人智协同信任的研究趋势和展望,为未来相关研究提供参考。

关键词: 人智协同, 生成设计, 生成式人工智能, 信任关系, 信任模型

Abstract: With the continuous development of generative AI technology, vertical-oriented AI generative design has become a new paradigm,whose trust relationship in human-AI team work has a significant impact on the innovation, efficiency and quality of design results. However, the current related research lacks a complete and systematic discussion. Therefore, in order to clarify the trends, hotspots and methods of the research related to the trust relationship in design for human-AI collaboration, the influencing factors of trust are defined by clustering and analyzing the literature from the development of research on trust. Based on the design field, a trust level model for human-AI collaboration in design is constructed, the research content and architecture of the three aspects of technology, interaction and experience are systematically explored. The key technologies of trust measurement, calibration, repair, decision-making, and evaluation are summarized. Research trends and perspectives of human-AI collaboration trust in design are presented in terms of theory, methods, technology, systems, and practice, which can provide references for future related research.

Key words: human-AI collaboration, generative design, generative AI, trust relationships, trust models

中图分类号: