Scholarlink

High-quality documentation is essential for reproducible and scalable Open Source Hardware. However, many projects apply existing documentation templates inconsistently, resulting in missing or outdated information. This study uses a two-phase approach to evaluate documentation quality and explore automation support. First, we benchmarked a minimal Open Source Hardware documentation template against seven representative hardware repositories to identify recurring gaps in adoption. Second, we tested Large Language Models (ChatGPT and Gemini AI) as Natural Language Processing assistants to automate documentation tasks such as detecting changes, updating files, and highlighting missing information. Results show that while the template already covers essential documentation categories, projects often fail to apply it consistently. Large Language Model-based assistants can help maintainers by reducing manual effort, improving cross-file consistency, and prompting for incomplete assets. Together, these findings outline a practical path toward semi-automated, maintainable, and reproducible documentation workflows for the Open Source Hardware community.

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

Gap Analysis of Documentation Template for Open Source Hardware and Adoption in Documentation Workflows with Large Language Models

  • 会议论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年

查看全文
  • Neeraj Chodankar,
  • Sagar Didaga Ramakrishna,
  • Sejal More,
  • Atharva Chitrakar,
  • Julien Colomb,
  • Robert Mies

摘要

High-quality documentation is essential for reproducible and scalable Open Source Hardware. However, many projects apply existing documentation templates inconsistently, resulting in missing or outdated information. This study uses a two-phase approach to evaluate documentation quality and explore automation support. First, we benchmarked a minimal Open Source Hardware documentation template against seven representative hardware repositories to identify recurring gaps in adoption. Second, we tested Large Language Models (ChatGPT and Gemini AI) as Natural Language Processing assistants to automate documentation tasks such as detecting changes, updating files, and highlighting missing information. Results show that while the template already covers essential documentation categories, projects often fail to apply it consistently. Large Language Model-based assistants can help maintainers by reducing manual effort, improving cross-file consistency, and prompting for incomplete assets. Together, these findings outline a practical path toward semi-automated, maintainable, and reproducible documentation workflows for the Open Source Hardware community.