Machine Learning based Laser Failure Mode Detection

TL;DR

This paper introduces an LSTM-based machine learning approach for detecting laser failure modes, significantly improving classification accuracy over traditional threshold methods and classical ML models.

Contribution

The study presents a novel LSTM-based fault detection method for laser degradation analysis, outperforming existing threshold and classical ML techniques.

Findings

LSTM model achieves 95.52% accuracy in laser failure detection.

LSTM outperforms threshold-based and classical ML models.

Synthetic data effectively trains the fault detection system.

Abstract

Laser degradation analysis is a crucial process for the enhancement of laser reliability. Here, we propose a data-driven fault detection approach based on Long Short-Term Memory (LSTM) recurrent neural networks to detect the different laser degradation modes based on synthetic historical failure data. In comparison to typical threshold-based systems, attaining 24.41% classification accuracy, the LSTM-based model achieves 95.52% accuracy, and also outperforms classical machine learning (ML) models namely Random Forest (RF), K-Nearest Neighbours (KNN) and Logistic Regression (LR).