Scholarlink

<p>This structured narrative review examines when digital twin-enabled predictive maintenance for MEMS is economically and technically defensible. We show that the strongest business case arises only when degradation is observable, sufficient lead time exists for intervention, and the avoided cost of failure exceeds monitoring and integration costs. The review distinguishes digital twins from digital models and simulation frameworks, summarizes MEMS-specific sensing and packaging constraints, and introduces a transparent review protocol, a benchmarking checklist, and a decision matrix for unit-level, fleet-level, screening, and redesign-first strategies. Overall, the literature supports a conditional rather than universal case for MEMS PdM, with value concentrated in high-criticality applications that combine actionable data, calibrated uncertainty, and organizational readiness.</p>

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

Digital twins and artificial intelligence for predictive maintenance in microelectromechanical systems (MEMS): a techno-economic perspective

  • 期刊论文

  • 发表日期：

    2026年7月2日

查看全文
  • Tingfa Zhou,
  • Chao Hu,
  • Dong Liu,
  • Yuanyuan Duan,
  • Zhong Zhang

摘要

This structured narrative review examines when digital twin-enabled predictive maintenance for MEMS is economically and technically defensible. We show that the strongest business case arises only when degradation is observable, sufficient lead time exists for intervention, and the avoided cost of failure exceeds monitoring and integration costs. The review distinguishes digital twins from digital models and simulation frameworks, summarizes MEMS-specific sensing and packaging constraints, and introduces a transparent review protocol, a benchmarking checklist, and a decision matrix for unit-level, fleet-level, screening, and redesign-first strategies. Overall, the literature supports a conditional rather than universal case for MEMS PdM, with value concentrated in high-criticality applications that combine actionable data, calibrated uncertainty, and organizational readiness.