Probabilistic Machine Learning for Uncertainty-Aware Diagnosis of Industrial Systems

2025-09-23

AI Summary

In industrial system fault diagnosis, deep neural networks lack reliable predictive confidence estimation, rendering consensus-based diagnosis vulnerable to false positives. This paper proposes an ensemble probabilistic machine learning framework tailored for consensus-driven diagnosis, which—uniquely and systematically—integrates uncertainty quantification into the diagnostic pipeline to enable dynamic modeling of prediction uncertainty and automated decision calibration. The method operates without high-fidelity physical models, relying solely on historical operational data for training and inference. Extensive multi-case experiments demonstrate consistent improvements over conventional data-driven approaches across key metrics—including accuracy, F1-score, and false positive rate. Ablation studies confirm that explicit uncertainty modeling substantially enhances diagnostic robustness. By bridging probabilistic reasoning with consensus mechanisms, the framework establishes a novel paradigm for trustworthy fault diagnosis in safety-critical and sensitive applications.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Knowledge Representation and Reasoning: Diagnosis and Abductive Reasoning
• Reasoning under Uncertainty: Relational Probabilistic Models

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Deep neural networks has been increasingly applied in fault diagnostics, where it uses historical data to capture systems behavior, bypassing the need for high-fidelity physical models. However, despite their competence in prediction tasks, these models often struggle with the evaluation of their confidence. This matter is particularly important in consistency-based diagnosis where decision logic is highly sensitive to false alarms. To address this challenge, this work presents a diagnostic framework that uses ensemble probabilistic machine learning to improve diagnostic characteristics of data driven consistency based diagnosis by quantifying and automating the prediction uncertainty. The proposed method is evaluated across several case studies using both ablation and comparative analyses, showing consistent improvements across a range of diagnostic metrics.

Problem

• Quantifying prediction uncertainty in neural network fault diagnosis
• Improving diagnostic reliability for consistency-based industrial systems
• Reducing false alarm sensitivity through probabilistic ensemble methods

Innovation

• Ensemble probabilistic machine learning for uncertainty quantification
• Automating prediction uncertainty in data-driven diagnosis
• Improving diagnostic metrics through uncertainty-aware framework