SIMAP: Intelligent System for Predictive Maintenance

Anomaly prediction of CT equipment based on IoMT data

BACKGROUND

Large-scale medical equipment, which is extensively implemented in medical services, is of vital importance for diagnosis but vulnerable to various anomalies and failures. Most hospitals that conduct regular maintenance have been suffering from medical equipment-related incidents for years. Currently, the Internet of Medical Things (IoMT) has emerged as a crucial tool in monitoring the real-time status of the medical equipment. In this paper, we develop an IoMT system of Computed Tomography (CT) equipment in the West China Hospital, Sichuan University and collected the system status time-series data. Novel multivariate time-series classification models and frameworks are proposed to predict the anomalies of CT equipment. The important features that are closely related to the equipment anomalies are identified with the model.

METHODS

We extracted the real-time CT equipment status time-series data of 11 equipment between May 19, 2020 and May 19, 2021 from the IoMT, which includes the equipment oil temperature, anode voltage, etc. The arcs are identified as labels of anomalies due to their relationship with decreased imaging quality and CT equipment failures. To improve prediction accuracy, the statistics and transformations of the raw historical time-series data segment in the sliding time window are used to construct new features. Due to the particularity of time-series data, two frameworks are proposed for splitting the training and test sets. Then the Decision Tree, Support Vector Machine, Logistic Regression, Naive Bayesian, and K-Nearest Neighbor classification models are used to classify the system status. We also compare our model to state-of-the-art models.

RESULTS

The results show that the anomaly prediction accuracy and recall of our method are 79% and 77%, respectively. The oil temperature and anode voltage are identified as the decisive features that may lead to anomalies. The proposed model outperforms the others when predicting the anomalies of the CT equipment based on our dataset.

CONCLUSIONS

The proposed method could predict the state of CT equipment and be used as a reference for practical maintenance, where unexpected anomalies of medical equipment could be reduced. It also brings new insights into how to handle non-uniform and imbalanced time series data in practical cases.

Anomaly Detection and Inter-Sensor Transfer Learning on Smart Manufacturing Datasets

Smart manufacturing systems are considered the next generation of manufacturing applications. One important goal of the smart manufacturing system is to rapidly detect and anticipate failures to reduce maintenance cost and minimize machine downtime. This often boils down to detecting anomalies within the sensor data acquired from the system which has different characteristics with respect to the operating point of the environment or machines, such as, the RPM of the motor. In this paper, we analyze four datasets from sensors deployed in manufacturing testbeds. We detect the level of defect for each sensor data leveraging deep learning techniques. We also evaluate the performance of several traditional and ML-based forecasting models for predicting the time series of sensor data. We show that careful selection of training data by aggregating multiple predictive RPM values is beneficial. Then, considering the sparse data from one kind of sensor, we perform transfer learning from a high data rate sensor to perform defect type classification. We release our manufacturing database corpus (4 datasets) and codes for anomaly detection and defect type classification for the community to build on it. Taken together, we show that predictive failure classification can be achieved, paving the way for predictive maintenance.

An Overview of Oil-Mineral-Aggregate Formation, Settling, and Transport Processes in Marine Oil Spill Models

An oil spill is considered one of the most serious polluting disasters for a marine environment. When oil is spilled into a marine environment, it is dispersed into the water column as oil droplets which often interact with suspended particles to form oil-mineral-aggregate (OMA). Knowing how OMA form, settle, and are transported is critical to oil spill modelling which can determine the fate and mass balance of the spilled volumes. This review introduces oil weathering and movement, and the commonly used numerical models that oil spill specialists use to determine how a spill will evolve. We conduct in-depth reviews of the environmental factors that influence how OMA form and their settling velocity, and we review how OMA formation and transport are modelled. We point out the existing gaps in current knowledge and the challenges of studying OMA. Such challenges include having to systematically conduct laboratory experiments to investigate how the environment affects OMA formation and settling velocities, and the need for a comprehensive algorithm that can estimate an OMA settling velocity.

Using Transfer Learning to Build Physics-Informed Machine Learning Models for Improved Wind Farm Monitoring

This paper introduces a novel, transfer-learning-based approach to include physics into data-driven normal behavior monitoring models which are used for detecting turbine anomalies. For this purpose, a normal behavior model is pretrained on a large simulation database and is recalibrated on the available SCADA data via transfer learning. For two methods, a feed-forward artificial neural network (ANN) and an autoencoder, it is investigated under which conditions it can be helpful to include simulations into SCADA-based monitoring systems. The results show that when only one month of SCADA data is available, both the prediction accuracy as well as the prediction robustness of an ANN are significantly improved by adding physics constraints from a pretrained model. As the autoencoder reconstructs the power from itself, it is already able to accurately model the normal behavior power. Therefore, including simulations into the model does not improve its prediction performance and robustness significantly. The validation of the physics-informed ANN on one month of raw SCADA data shows that it is able to successfully detect a recorded blade angle anomaly with an improved precision due to fewer false positives compared to its purely SCADA data-based counterpart.

Condition-Based Monitoring and Maintenance: State of the Art Review

Manufacturing firms face great pressure to reduce downtime as well as maintenance costs. Condition-based maintenance (CBM) can be used to effectively manage operations and maintenance by monitoring detailed machine health information. CBM policies and the development of the mathematical models have been growing recently. This paper provides a review of the theoretical and practical development in the field of condition-based maintenance and its current advancements. Standard CBM platform could make it effective and efficient in implementation and performance improvement.

SCADA data as a powerful tool for early fault detection in wind turbine gearboxes

The different operating conditions of wind turbines pose great challenges for efficient and reliable fault detection. Therefore, a good analysis of wind turbine data is essential in assessing the state of the wind turbines, since the traditional threshold cannot provide a timely warning as it indicates that the malfunction has already occurred. This paper presents a new method for analyzing the actual data of the turbines, using aggregated model consisting of the neighborhood comparison method, K-means clustering and decision tree model to diagnose faults. The wind speed of the adjacent turbines is compared with each other, then other parameters of the same wind speed are also compared with each other. The purpose of comparison is that, the wind turbines which are similar in wind speed are similar in performance as well. This approach helps us to discover the abnormal data for turbine performance with in the normal operating range. The abnormal performance of any turbine destroys the similarity relationship between its data and the neighboring unit’s data. The main advantage of this approach is the possibility to detect the beginning of abnormal performance in real time, a case study using real SCADA data is used to validate this approach, which demonstrates its effectiveness and advantages.

Intelligent Condition Monitoring of Wind Power Systems: State of the Art Review

Modern wind turbines operate in continuously transient conditions, with varying speed, torque, and power based on the stochastic nature of the wind resource. This variability affects not only the operational performance of the wind power system, but can also affect its integrity under service conditions. Condition monitoring continues to play an important role in achieving reliable and economic operation of wind turbines. This paper reviews the current advances in wind turbine condition monitoring, ranging from conventional condition monitoring and signal processing tools to machine-learning-based condition monitoring and usage of big data mining for predictive maintenance. A systematic review is presented of signal-based and data-driven modeling methodologies using intelligent and machine learning approaches, with the view to providing a critical evaluation of the recent developments in this area, and their applications in diagnosis, prognosis, health assessment, and predictive maintenance of wind turbines and farms.

An Improved Feature Selection Method Based on Random Forest Algorithm for Wind Turbine Condition Monitoring

Feature selection and dimensionality reduction are important for the performance of wind turbine condition monitoring models using supervisory control and data acquisition (SCADA) data. In this paper, an improved random forest algorithm, namely Feature Simplification Random Forest (FS_RF), is proposed, which is capable of identifying features closely correlated with wind turbine working conditions. The Euclidian distances are employed to distinguish the weight of the same feature among different samples, and its importance is measured by means of the random forest algorithm. The selected features are finally verified by a two-layer gated recurrent unit (GRU) neural network facilitating condition monitoring. The experimental results demonstrate the capacity and effectiveness of the proposed method for wind turbine condition monitoring.

Condition Monitoring of Rail Transport Systems: A Bibliometric Performance Analysis and Systematic Literature Review

Condition monitoring of rail transport systems has become a phenomenon of global interest over the past half a century. The approaches to condition monitoring of various rail transport systems—especially in the context of rail vehicle subsystem and track subsystem monitoring—have been evolving, and have become equally significant and challenging. The evolution of the approaches applied to rail systems’ condition monitoring has followed manual maintenance, through methods connected to the application of sensors, up to the currently discussed methods and techniques focused on the mutual use of automation, data processing, and exchange. The aim of this paper is to provide an essential overview of the academic research on the condition monitoring of rail transport systems. This paper reviews existing literature in order to present an up-to-date, content-based analysis based on a coupled methodology consisting of bibliometric performance analysis and systematic literature review. This combination of literature review approaches allows the authors to focus on the identification of the most influential contributors to the advances in research in the analyzed area of interest, and the most influential and prominent researchers, journals, and papers. These findings have led the authors to specify research trends related to the analyzed area, and additionally identify future research agendas in the investigation from engineering perspectives.

An Overview on Fault Diagnosis, Prognosis and Resilient Control for Wind Turbine Systems

Wind energy is contributing to more and more portions in the world energy market. However, one deterrent to even greater investment in wind energy is the considerable failure rate of turbines. In particular, large wind turbines are expensive, with less tolerance for system performance degradations, unscheduled system shut downs, and even system damages caused by various malfunctions or faults occurring in system components such as rotor blades, hydraulic systems, generator, electronic control units, electric systems, sensors, and so forth. As a result, there is a high demand to improve the operation reliability, availability, and productivity of wind turbine systems. It is thus paramount to detect and identify any kinds of abnormalities as early as possible, predict potential faults and the remaining useful life of the components, and implement resilient control and management for minimizing performance degradation and economic cost, and avoiding dangerous situations. During the last 20 years, interesting and intensive research results were reported on fault diagnosis, prognosis, and resilient control techniques for wind turbine systems. This paper aims to provide a state-of-the-art overview on the existing fault diagnosis, prognosis, and resilient control methods and techniques for wind turbine systems, with particular attention on the results reported during the last decade. Finally, an overlook on the future development of the fault diagnosis, prognosis, and resilient control techniques for wind turbine systems is presented.

Anomaly Detection of Power Plant Equipment Using Long Short-Term Memory Based Autoencoder Neural Network

Anomaly detection is of great significance in condition-based maintenance of power plant equipment. The conventional fixed threshold detection method is not able to perform early detection of equipment abnormalities. In this study, a general anomaly detection framework based on a long short-term memory-based autoencoder (LSTM-AE) network is proposed. A normal behavior model (NBM) is established to learn the normal behavior patterns of the operating variables of the equipment in space and time. Based on the similarity analysis between the NBM output distribution and the corresponding measurement distribution, the Mahalanobis distance (MD) is used to describe the overall residual (OR) of the model. The reasonable range is obtained using kernel density estimation (KDE) with a 99% confidence interval, and the OR is monitored to detect abnormalities in real-time. An induced draft fan is chosen as a case study. Results show that the established NBM has excellent accuracy and generalizability, with average root mean square errors of 0.026 and 0.035 for the training and test data, respectively, and average mean absolute percentage errors of 0.027%. Moreover, the abnormal operation case shows that the proposed framework can be effectively used for the early detection of abnormalities.

Determining the initial and subsequent impact of artificial intelligence adoption on economy: a macroeconomic survey from ASEAN

The adoption of AI is an ongoing phenomenon in today’s economy in all the industries. The purpose of this paper is to examine the economic impact of AI adoption in the region of ASEAN. To achieve this objective, structural questionnaire was developed for the various industry experts in targeted region. A sample of 240 experts was finally obtained over a time span of 6 weeks through online structural questionnaire approach. For measuring AI adoption, twelve items, initial economic impact (seven items), and subsequent economic impact (six items) were finally added in the questionnaire. For analyses purpose, descriptive statistics, structural equation modelling, and regression analyseswereapplied, examining the both initial and subsequent economic impact of AI adoption. Findings through structural model indicates that overall both initial and subsequent impact are significantly determined by AI adoption in related industries. Additionally, in depth analyses for the individual AI items as their initial and subsequent economic impact indicate that Usage of the data for AI adoption, clear strategy for AI adoption, successful mapping for AI adoption and overall positive attitude towards AI adoption have their significant and positive influence on initial economic indicators. Whereas, as per subsequent economic impact, factors like effective usage of data for AI adoption, assessing the right skills of individuals for AI adoption and positive attitude towards AI adoption are significantly impacting on material investment, capital investment, increasing unemployment, higher economic output, higher return on capital and higher wages for the existing labor. These findings have provided an outstanding evidence in the field of AI and its economic impact in the region of ASEAN and can be considered as initial contribution in related fields. Both industry exports and macroeconomic decision makers can significantly utilize the findings to develop their conceptual framework and understanding for the integration between AI adoption and economy. Additionally, this study can work as reasonable justification for implementing the more adoption of AI in various industries as it has positive economic outcome (both initial and subsequent). However, one of the key limitations of this study is limited sample size and only 240 industry exports were targeted from selected industries in ASEAN. Future study could be reimplemented on similar topic with expanding the sample size for better findings and more generalization.

Effect of Time History on Normal Behaviour Modelling Using SCADA Data to Predict Wind Turbine Failures

Operations and Maintenance (O&M) can make up a significant proportion of lifetime costs associated with any wind farm, with up to 30% reported for some offshore developments. It is increasingly important for wind farm owners and operators to optimise their assets in order to reduce the levelised cost of energy (LCoE). Reducing downtime through condition-based maintenance is a promising strategy of realising these goals. This is made possible through increased monitoring and gathering of operational data. SCADA data are useful in terms of wind turbine condition monitoring. This paper aims to perform a comprehensive comparison between two types of normal behaviour modelling: full signal reconstruction (FSRC) and autoregressive models with exogenous inputs (ARX). At the same time, the effects of the training time period on model performance are explored by considering models trained with both 12 and 6 months of data. Finally, the effects of time resolution are analysed for each algorithm by considering models trained and tested with both 10 and 60 min averaged data. Two different cases of wind turbine faults are examined. In both cases, the NARX model trained with 12 months of 10 min average Supervisory Control And Data Acquisition (SCADA) data had the best training performance.

Designing predictive maintenance systems using decision tree-based machine learning techniques

Purpose

The after-sale service industry is estimated to contribute over 8 percent to the US GDP. For use in this considerably large service management industry, this article provides verification in the application of decision tree-based machine learning algorithms for optimal maintenance decision-making. The motivation for this research arose from discussions held with a large agricultural equipment manufacturing company interested in increasing the uptime of their expensive machinery and in helping their dealer network.

Design/methodology/approach

We propose a general strategy for the design of predictive maintenance systems using machine learning techniques. Then, we present a case study where multiple machine learning algorithms are applied to a particular example situation for an illustration of the proposed strategy and evaluation of its performance.

Findings

We found progressive improvements using such machine learning techniques in terms of accuracy in predictions of failure, demonstrating that the proposed strategy is successful.

Research limitations/implications

This approach is scalable to a wide variety of applications to aid in failure prediction. These approaches are generalizable to many systems irrespective of the underlying physics. Even though we focus on decision tree-based machine learning techniques in this study, the general design strategy proposed can be used with all other supervised learning techniques like neural networks, boosting algorithms, support vector machines, and statistical methods.

Practical implications

This approach is applicable to many different types of systems that require maintenance and repair decision-making. A case is provided for a cloud data storage provider. The methods described in the case can be used in any number of systems and industrial applications, making this a very scalable case for industry practitioners. This scalability is possible as the machine learning techniques learn the correspondence between machine conditions and outcome state irrespective of the underlying physics governing the systems.

Social implications

Sustainable systems and operations require allocating and utilizing resources efficiently and effectively. This approach can help asset managers decide how to sustainably allocate resources by increasing uptime and utilization for expensive equipment.

Originality/value

This is a novel application and case study for decision tree-based machine learning that will aid researchers in developing tools and techniques in this area as well as those working in the artificial intelligence and service management space.

A Survey of Condition Monitoring and Fault Diagnosis toward Integrated O&M for Wind Turbines

Wind power, as a renewable energy for coping with global climate change challenge, has achieved rapid development in recent years. The breakdown of wind turbines (WTs) not only leads to high repair expenses but also may threaten the stability of the whole power grid. How to reduce the operation and the maintenance (O&M) cost of wind farms is an obstacle to its further promotion and application. To provide reliable condition monitoring and fault diagnosis (CMFD) for WTs, this paper presents a comprehensive survey of the existing CMFD methods in the following three aspects: energy flow, information flow, and integrated O&M system. Energy flow mainly analyzes the characteristics of each component from the angle of energy conversion of WTs. Information flow is the carrier of fault and control information of WT. At the end of this paper, an integrated WT O&M system based on electrical signals is proposed.

A combined mono- and multi-turbine approach for fault indicator synthesis and wind turbine monitoring using SCADA data

The monitoring of wind turbines using SCADA data has received lately a growing interest from the fault diagnosis community because of the very low cost of these data, which are available in number without the need for any additional sensor. Yet, these data are highly variable due to the turbine constantly changing its operating conditions and to the rapid fluctuations of the environmental conditions (wind speed and direction, air density, turbulence, …). This makes the occurrence of a fault difficult to detect. To address this problem, we propose a multi-level (turbine and farm level) strategy combining a mono- and a multi-turbine approach to create fault indicators insensitive to both operating and environmental conditions. At the turbine level, mono-turbine residuals (i.e. a difference between an actual monitored value and the predicted one) obtained with a normal behavior model expressing the causal relations between variables from the same single turbine and learnt during a normal condition period are calculated for each turbine, so as to get rid of the influence of the operating conditions. At the farm level, the residuals are then compared to a wind farm reference in a multi-turbine approach to obtain fault indicators insensitive to environmental conditions. Indicators for the objective performance evaluation are also proposed to compare wind turbine fault detection methods, which aim at evaluating the cost/benefit of the methods from a production manager's point of view. The performance of the proposed combined mono- and multi-turbine method is evaluated and compared to more classical methods proposed in the literature on a large real data set made of SCADA data recorded on a French wind farm during four years : it is shown than it can improve the fault detection performance when compared to a residual analysis limited at the turbine level only.

Innovative airport 4.0 condition-based maintenance system for baggage handling DCV systems

Purpose

The purpose of this paper is to present the findings of a recent study conducted with the objective of addressing the problem of failure of baggage carts in the high-speed baggage tunnel at Heathrow Terminal 5 by the development of an innovative condition-based maintenance (CBM) system designed to meet the requirements of 21st century airport systems and Industry 4.0.

Design/methodology/approach

An empirical experimental approach to this action research was taken to install a vibration condition monitoring pilot test in the north tunnel at Terminal 5. Vibration data were collected over a 6-month period and analysed to find the threshold of good quality tyres and those with worn bearings that needed replacement. The results were compared with existing measures to demonstrate that vibration monitoring could be used as a predictive model for CBM.

Findings

The findings demonstrated a clear trend of increasing vibration velocity with age and use of the baggage cart wheels caused by wheel mass unbalanced inertia that was transmitted to the tracks as vibration. As a result, preventative maintenance is essential to ensure the smooth running of airport baggage. This research demonstrates that a healthy wheel produces vibration of under 60 mm/s whereas a damaged wheel measures up to 100 mm/s peak to peak velocity and this can be used in real-time condition monitoring to prevent baggage cart failure. It can also run as an autonomous system linked to AI and Industry 4.0 airport logic.

Originality/value

Whilst vibration monitoring has been used to measure movement in static structures such as bridges and used in rotating machinery such as railway wheels (Tondon and Choudhury, 1999); this is unique as it is the first time it has been applied on a stationary structure (tracks) carrying high-speed rotating machinery (baggage cart wheels). This technique has been patented and proven in the pilot study and is in the process of being rolled out to all Heathrow terminal connection tunnels. It has implications for all other airports worldwide and, with new economic sensors, to other applications that rely on moving conveyor belts.

Feature Selection Algorithms for Wind Turbine Failure Prediction

It is well known that each year the wind sector has profit losses due to wind turbine failures and operation and maintenance costs. Therefore, operations related to these actions are crucial for wind farm operators and linked companies. One of the key points for failure prediction on wind turbine using SCADA data is to select the optimal or near optimal set of inputs that can feed the failure prediction (prognosis) algorithm. Due to a high number of possible predictors (from tens to hundreds), the optimal set of inputs obtained by exhaustive-search algorithms is not viable in the majority of cases. In order to tackle this issue, show the viability of prognosis and select the best set of variables from more than 200 analogous variables recorded at intervals of 5 or 10 min by the wind farm’s SCADA, in this paper a thorough study of automatic input selection algorithms for wind turbine failure prediction is presented and an exhaustive-search-based quasi-optimal (QO) algorithm, which has been used as a reference, is proposed. In order to evaluate the performance, a k-NN classification algorithm is used. Results showed that the best automatic feature selection method in our case-study is the conditional mutual information (CMI), while the worst one is the mutual information feature selection (MIFS). Furthermore, the effect of the number of neighbours (k) is tested. Experiments demonstrate that k = 1 is the best option if the number of features is higher than 3. The experiments carried out in this work have been extracted from measures taken along an entire year and corresponding to gearbox and transmission systems of Fuhrländer wind turbines.