Transfer Learning for Smart Manufacturing: A Stepwise Survey

Shufei Li, Pai Zheng

Research output: Journal article publicationConference articleAcademic researchpeer-review

5 Citations (Scopus)


Nowadays, industrial companies embrace the cutting-edge artificial intelligence (AI) techniques to achieve smart manufacturing over the entire organization. However, effective data collection and annotation still remain as a big challenge in many manufacturing scenarios. Transfer learning, serving as a breakthrough of learning sharing knowledge and extracting latent features from scarce data, has attracted much attention. Transfer learning in literature mainly focuses on the definitions and mechanisms of interpretation while lacking a systematic implementation scheme for manufacturing. To fulfill this gap and facilitate industrial resource use efficiency, this paper attempts to systematize strategies of transfer learning in today's smart manufacturing in a step-by-step manner. Twenty representative transfer learning works are investigated from the perspectives of manufacturing activities along the engineering product lifecycle. Meanwhile, the potential availability of industrial dataset is also briefly introduced. It is hoped this research can provide a clear guide for both academics and industrial practitioners to design appropriate learning approaches according to their own industrial scenarios.

Original languageEnglish
Pages (from-to)37-42
Number of pages6
Issue number5
Publication statusPublished - 2020
Event3rd IFAC Workshop on Cyber-Physical and Human Systems, CPHS 2020 - Beijing, China
Duration: 3 Dec 20205 Dec 2020


  • domain adaptation
  • manufacturing intelligence
  • smart manufacturing
  • Transfer learning

ASJC Scopus subject areas

  • Control and Systems Engineering


Dive into the research topics of 'Transfer Learning for Smart Manufacturing: A Stepwise Survey'. Together they form a unique fingerprint.

Cite this