A critical review on prognostics for stochastic degrading systems under big data

Area of Science:

• Engineering
• Computer Science
• Data Science

Background:

• Remaining Useful Life (RUL) prediction, or prognostics, is crucial for the safety and reliability of stochastic degrading systems.
• The rise of Industry 4.0 and the Internet of Things (IoT) has increased attention on prognostics methods utilizing big data.
• A critical review of the strengths and weaknesses of existing prognostics methods is needed to stimulate new research directions.

Purpose of the Study:

• To critically analyze data-driven prognostics methods for stochastic degrading systems within the context of big data.
• To identify common problems, development trends, and research gaps in current prognostics approaches.
• To explore the emerging area of prognosis under incomplete big data and highlight future opportunities.

Main Methods:

• Analysis of statistical data-driven prognostics methods.
• Review of machine learning (ML)-based prognostics methods.
• Examination of hybrid prognostics approaches combining statistical and ML methods.

Main Results:

• Discussion of the basic research ideas, development trends, and common problems associated with various data-driven prognostics methods.
• Identification of the pros and cons of existing statistical, ML-based, and hybrid prognostics techniques.
• Exploration of prognosis under incomplete big data as a key emerging topic.

Conclusions:

• The paper provides a comprehensive overview of data-driven prognostics for stochastic degrading systems in the big data era.
• It offers insights into challenges and potential opportunities to guide future prognostics research.
• The review aims to stimulate new ideas and advancements in the field of prognostics.