Self-Supervised Learning for Data Scarcity in a Fatigue Damage Prognostic Problem

TL;DR

This paper explores self-supervised learning to improve RUL predictions in fatigue damage prognosis, demonstrating that pre-trained models outperform non-pre-trained ones with limited labeled data in aerospace-like scenarios.

Contribution

It introduces a self-supervised pre-training approach for RUL estimation in fatigue damage prognosis, effective with scarce labeled data and validated on synthetic aerospace-related datasets.

Findings

Self-supervised pre-training improves RUL prediction accuracy.

Pre-trained models require less computational resources.

Significant performance gains with limited labeled data.

Abstract

With the increasing availability of data for Prognostics and Health Management (PHM), Deep Learning (DL) techniques are now the subject of considerable attention for this application, often achieving more accurate Remaining Useful Life (RUL) predictions. However, one of the major challenges for DL techniques resides in the difficulty of obtaining large amounts of labelled data on industrial systems. To overcome this lack of labelled data, an emerging learning technique is considered in our work: Self-Supervised Learning, a sub-category of unsupervised learning approaches. This paper aims to investigate whether pre-training DL models in a self-supervised way on unlabelled sensors data can be useful for RUL estimation with only Few-Shots Learning, i.e. with scarce labelled data. In this research, a fatigue damage prognostics problem is addressed, through the estimation of the RUL of aluminum alloy panels (typical of aerospace structures) subject to fatigue cracks from strain gauge data. Synthetic datasets composed of strain data are used allowing to extensively investigate the influence of the dataset size on the predictive performance. Results show that the self-supervised pre-trained models are able to significantly outperform the non-pre-trained models in downstream RUL prediction task, and with less computational expense, showing promising results in prognostic tasks when only limited labelled data is available.