Semi-supervised Learning with Deep Generative Models for Asset Failure Prediction

TL;DR

This paper introduces a semi-supervised deep generative modeling approach for predicting asset failure, effectively utilizing limited health status data to improve reliability predictions.

Contribution

It combines deep neural networks with non-linear embedding to enhance prognostic accuracy with sparse health status labels, advancing asset failure prediction methods.

Findings

Significant improvement in RUL estimation with only 1% labeled data

Deep generative models effectively regularize complex prognostic models

High prediction accuracy with limited health status information

Abstract

This work presents a novel semi-supervised learning approach for data-driven modeling of asset failures when health status is only partially known in historical data. We combine a generative model parameterized by deep neural networks with non-linear embedding technique. It allows us to build prognostic models with the limited amount of health status information for the precise prediction of future asset reliability. The proposed method is evaluated on a publicly available dataset for remaining useful life (RUL) estimation, which shows significant improvement even when a fraction of the data with known health status is as sparse as 1% of the total. Our study suggests that the non-linear embedding based on a deep generative model can efficiently regularize a complex model with deep architectures while achieving high prediction accuracy that is far less sensitive to the availability of health status information.