Machine Failure Detection Based on Projected Quantum Models

TL;DR

The paper tackles proactive industrial maintenance by framing machine-failure detection as online, unsupervised change-point detection on multidimensional time series. It introduces projected quantum models that extract 1-RDM-based features from quantum states, feeding them into a density-ratio-based divergence estimator (uLSIF) to produce an anomaly score $a_s$. Empirical results on synthetic data, bee waggle datasets, and real IoT sensor data—executed on both simulators and IBM hardware—demonstrate that quantum feature transforms can improve change-point detection quality and robustness to noise, with competitive or superior AUC performance and practical feasibility. This work suggests a promising path for integrating near-term quantum processing into predictive maintenance and industrial diagnostics, while highlighting future avenues to understand the underlying mechanisms behind projected quantum features.

Abstract

Detecting machine failures promptly is of utmost importance in industry for maintaining efficiency and minimizing downtime. This paper introduces a failure detection algorithm based on quantum computing and a statistical change-point detection approach. Our method leverages the potential of projected quantum feature maps to enhance the precision of anomaly detection in machine monitoring systems. We empirically validate our approach on benchmark multi-dimensional time series datasets as well as on a real-world dataset comprising IoT sensor readings from operational machines, ensuring the practical relevance of our study. The algorithm was executed on IBM's 133-qubit Heron quantum processor, demonstrating the feasibility of integrating quantum computing into industrial maintenance procedures. The presented results underscore the effectiveness of our quantum-based failure detection system, showcasing its capability to accurately identify anomalies in noisy time series data. This work not only highlights the potential of quantum computing in industrial diagnostics but also paves the way for more sophisticated quantum algorithms in the realm of predictive maintenance.

Figures (13)

• Figure 1: Overview of the proposed framework for machine failure detection based on projected quantum feature extraction and statistical change-point detection approach. Original feature vectors obtained from sensors are transformed to projected quantum features of dimension $d'$ that depends on the quantum circuit used for transformation. New feature vectors are used for statistical divergence estimation to measure anomaly score $a_s$. Large $a_s$ implies that a fault has likely occurred on a machine.
• Figure 2: Schematic diagram of the quantum circuit in our experiment for $n_q=4$. $U_1, \ldots,U_4$ are Haar-random unitary gates with a fixed random seed and $R_{XX}, R_{YY}, R_{ZZ}$ are 2-qubit rotation gates with respect to $XX, YY, ZZ$, respectively. $t$ is set to a fixed constant and $\theta$ represents feature vectors.
• Figure 3: Data values.
• Figure 4: Change scores.
• Figure 5: 1-RDM feature values.