Stanley Fong

Stanley Fong

Auto-generated profile, based on publicly available information

Stanley Fong

Department of Civil and Environmental Engineering, University of Waterloo, Waterloo, ON, Canada

Research Focus & Metrics

Academic Career Length

9Years (2018-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
23 December 2021 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
SFStanley Fong
Stanley Fong
SNSriram Narasimhan
Sriram Narasimhan
In the literature pertaining to condition-based maintenance using machine learning, the class of unsupervised approaches has yet to be fully explored. In particular, the topics of fault prediction and unsupervised degradation modeling, which have significant importance to practice, have not been widely studied. In this article, we propose a novel unsupervised approach for early degradation detection and fault prediction, termed the Bayesian one-class support vector machine (B-OCSVM), which is based on OC-SVM and the hierarchical Bayesian framework. The proposed approach aims to address key gaps in the literature by proposing a new unsupervised OC-SVM hyperparameter estimation method for early degradation detection and new degradation threshold, which is used to predict when a detected degradation process will manifest as a fault. Unlike traditional prognostic approaches, the proposed degradation threshold can be defined without prior knowledge pertaining to the fault. Fault prediction is achieved using a novel prognostic in conjunction with a hierarchical Bayesian degradation framework. We validate and benchmark the proposed approach against the state of the art using run-to-failure data from three datasets to demonstrate the earlier detection point of the proposed approach compared to existing approaches and to showcase how the proposed degradation threshold and novel prognostic can be used effectively to predict when faults will occur. In addition, the results show that B-OCSVM generalizes well to a number of different operating conditions and fault types, which signifies that the early detection and prognostic elements of the proposed approach have strong practical significance in real-world applications.

30 September 2019 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Open access
SFStanley Fong
Stanley Fong
JHJinane Harmouche
Jinane Harmouche
SNSriram Narasimhan
Sriram Narasimhan
Jerome AntoniJerome Antoni
Jerome Antoni
Vibration analysis is a powerful tool for condition monitoring of rotating machinery. In the non-stationary case, this analysis often involves de-noising and extraction of the time-varying harmonic components buried within the vibration signal. However, the complexity of many contemporary techniques especially in relation to non-stationary signals and their dependence on prior knowledge of the system kinematics in order to be effective is an inhibitor to autonomous fault detection and monitoring of non-stationary systems. In this paper, a non-parametric, blind spectral pre-processing approach to simultaneously de-noise and extract harmonic content from non-stationary vibration signals is presented. The proposed approach utilizes Mean Shift Clustering in conjunction with the Short-Time Fourier Transform to separate time-varying harmonics from background noise within the frequency spectrum, without the need for a priori knowledge of the system. The technique is fully invertible, allowing the time signals corresponding to the separated time-varying harmonic and residual components to be reconstructed. The performance of the proposed technique is compared against existing pre-processing methods, and validated using several industrial datasets: first, using vibration data obtained from a low-speed, non-stationary industrial automated people mover gearbox, next using vibration data from an aircraft engine containing outer race faults, and finally, using non-stationary vibration data from a wind turbine containing frequent speed fluctuations.

26 April 2018
Verified
Conference Paper
Closed access
SFStanley Fong
Stanley Fong
AAAli Ashasi-Sorkhabi
Ali Ashasi-Sorkhabi
GPGuru Prakash
Guru Prakash
SNSriram Narasimhan
Sriram Narasimhan
MRMike Riseborough
Mike Riseborough
Cable-driven automated people movers (APMs) are complex, dynamic systems that are comprised of many critical rotating components such as bearings, gears, shafts, and line sheaves. Consequently, these systems are susceptible to a number of faults such as bearing failures, shaft misalignments, or gear tooth deterioration. Due to the continual operative nature of many APM systems, downtimes due to faults or superfluous maintenance can result in significant economic losses. Recently, preventative maintenance strategies have become an attractive alternative to traditional run-to-failure or scheduled maintenance strategies. Condition-based maintenance is a preventative maintenance strategy that works by detecting incipient faults within the critical components of a system through the identification of anomalies within condition indicators extracted from their vibration signatures. A fault is detected when a condition indicator exceeds a prescribed failure threshold. Traditionally, these failure thresholds are based on historical precedents or failure data obtained from similar specimens. However, due to their longevity and bespoke construction, many APMs lack the historical failure data required to define these failure thresholds in the traditional way. In this paper, an automated CBM framework developed specifically for a cable-driven APM gearbox is presented. The framework utilizes a state-based Gaussian mixture modeling approach to account for variations in operating conditions resulting from frequent start and stop phases, variations in passenger load, and fluctuations in cable tension, among other variables. Failure threshold setting is accomplished using a hierarchical Bayesian approach, which utilizes past and present data to set thresholds that adapt to changing health state of the machine over its lifespan. In addition to real-time monitoring, the proposed framework utilizes the collected data to model the degradation of the system in order to generate an estimate of its remaining useful life. The combination of short-term preventative maintenance (real-time fault detection) and long-term predictive maintenance coupled with automation, makes the presented approach a highly efficient tool for maintenance planning.