Scholarlink

Scenario-Based Design (SBD) processes often struggle with biases, overgeneralization, and limited inclusivity, when designing for complex domains like Autonomous Vehicles (AVs). Through focus groups with non-technical end-users, this study defines requirements for SBD End-User Development (EUD) tools. By thematically analyzing and affinity mapping transcript and note data from 16 participants that engaged with a typical SBD workflow, five core needs for SBD software are identified: (1) context-rich foundations, (2) dynamic persona templates, (2) real-time scenarios reflecting socio-behavioral edge-cases, (3) strong persona-scenario synthesis and natural end-user interaction, (4) explainable and transparent interactions, and (5) data-driven yet interpretable processes and outputs. Findings suggest how EUD design tools can balance human creativity and input with computational rigor, enabling designers to iteratively refine designs while ensuring transparency, inclusivity, and explainability.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Scenario-Based Design Tools: From End-User Reflections to Requirements

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Benjamin Cham,
  • Genovefa Kefalidou

摘要

Scenario-Based Design (SBD) processes often struggle with biases, overgeneralization, and limited inclusivity, when designing for complex domains like Autonomous Vehicles (AVs). Through focus groups with non-technical end-users, this study defines requirements for SBD End-User Development (EUD) tools. By thematically analyzing and affinity mapping transcript and note data from 16 participants that engaged with a typical SBD workflow, five core needs for SBD software are identified: (1) context-rich foundations, (2) dynamic persona templates, (2) real-time scenarios reflecting socio-behavioral edge-cases, (3) strong persona-scenario synthesis and natural end-user interaction, (4) explainable and transparent interactions, and (5) data-driven yet interpretable processes and outputs. Findings suggest how EUD design tools can balance human creativity and input with computational rigor, enabling designers to iteratively refine designs while ensuring transparency, inclusivity, and explainability.