Maintenance Costs Research Articles

Sustainability of collective irrigation under water competition between agriculture and civil uses: the case study of Alanya Water Users Association (Türkiye)

This study explores the patterns of sustainable use and management of competing water sectors in the Alanya Water Users Association (WUA) in Southern Türkiye, over eight years (2013-2020) before the SARS-CoV-19 pandemic, focusing on the impacts of the recorded huge growth of tourism. Performance indicators of collective irrigation services are used to identify performance patterns and trends over time. The analysis has revealed a notable increase in water consumption for agricultural activities (+30%), driven by the cultivation of tropical fruits, without a proportional rise in crop production. Concurrently, effective financial management is observed in the WUA, with a consistent reduction in unit Management, Operation and Maintenance costs (-40% to 70%). Additionally, the total water demand has surged (+100%) due to population growth and tourist flux. However, this increase has tensioned water delivery to crops, indicating higher pressures over water availability for all uses. The study has identified three distinct patterns in the technical, financial, and socio-economic performance of the WUA, particularly highlighting the last four years of increased water usage and the disruption caused by the SARS-CoV-19 pandemic in 2020. Despite the current adequacy of water resources, optimized strategies for water management are advocated to address anticipated demographic growth, the introduction of tropical crops and the environmental impacts of climate change.

Performance Enhancement in Air Gap Membrane Distillation Using Heat Recovery Configurations

Experimental investigations were conducted on an air gap membrane distillation (AGMD) unit to evaluate the impact of heat recovery on system performance. The study utilized a steady-state heat source in conjunction with a spiral air gap membrane module for water desalination. The process was implemented using two configurations aimed to minimize heat loss enhancing productivity, gain output ratio (GOR), and thermal efficiency. The experiments involved feedwater with salt concentrations ranging from 10,000 to 30,000 ppm. The measurements were recorded for inlet and outlet temperatures of the heat source, water productivity, and thermal efficiency at various coolant and water feed flow rates. These configurations effectively addressed critical technical challenges in thermal AGMD systems by optimizing heat recovery. The incorporation of heat recovery resulted in a significant increase in water productivity by 37.07% compared to systems without heat recovery while maintaining the same initial heat input. The heat recovery in series (HRSS) and parallel (HRPS) configurations resulted in power consumption reductions of approximately 29.7% and 23.1%, respectively compared to the conventional configuration (CC) at a flow rate of 12 L/min. The levelized cost of water (LCOW) for the AGMD system was comprehensively evaluated and found to be $ 7.71 per cubic meter. This figure reflected the total cost of water production over the system's operational lifespan including capital expenditures, operating and maintenance costs, and energy consumption. The calculated LCOW provides a clear indicator of the system's economic viability, demonstrating the cost required to produce each cubic meter of water through the AGMD.

Effect of aged treatment on improving the performance of HVAF WC–Cr3C2–Ni coating on crystallization roller

The thermal shock resistance and high-temperature performance of WC–Cr3C2–Ni coatings on the surface of the crystallization roll after aged treatment were investigated in this study. The aging treatment at 320 ℃ for 2.5 h facilitated the diffusion of Cu element from the C17200 beryllium copper substrate toward the coating side of WC–Cr3C2–Ni coating, thereby enhancing the bonding between the coatings and substrates. After the aged treatment, the internal structure of the C17200 substrate changed from a dendritic structure to an equiaxed crystal structure, the hardness increased from 225.71 Hv0.5 to 406.52 Hv0.5 (80.4 % increased), and the wear resistance increased.Although the aging treatment reduced the hardness of the coating slightly (from 1257.58 Hv0.5 to 1155.16 Hv0.5, a decrease of 8.1%) and wear resistance (the coefficient of friction, from 0.335 to 0.312, a decrease of 6.9%). However, in the thermal shock test, the thermal shock resistance of the coating after aging treatment was increased from 19 times to 20 times (5.3% increased). However, the thermal shock failure cracks between the coatings and the substrates were transferred from the substrate part to the joint part of the two, which effectively avoided the waste of the crystallization roller substrate in the maintenance process. This research effectively reduces the purchase and maintenance costs of steel mills in actual production and provides theoretical support for cost reduction and efficiency increase in other HVAF (high-velocity air fuel) coating applications.

Cost-optimized probabilistic maintenance for condition monitoring of wind turbines with rare failures

We propose a method, a model, and a form of presenting model results for condition monitoring of a small set of wind turbines with rare failures. The main new ingredient of the method is to sample failure thresholds according to the profit they give to an operating company. The model is a multiple linear regression with seasonal components and external regressors, representing all sensor components except for the selected one. To overcome the scarcity of the training data, we use the median sensor values from all available turbines in their healthy state. The cumulated deviation from the normal behavior model obtained for this median turbine is calibrated for each turbine at the beginning of the test period and after known failures. The proposed form of presenting results is to set a scale for possible costs, control for random maintenance, and show a whole distribution of costs depending on the free model parameters. We make a case study on an open dataset with SCADA data from multiple sensors and show that considering the influence of turbine components is more critical than seasonality. The distribution, the average, and the standard deviation of maintenance costs can be very different for similar minimal costs. Random maintenance can be more profitable than reactive maintenance and other approaches. Our predictive maintenance model outperforms random maintenance and competitors for the whole set of considered turbines, giving substantial savings.

Health monitoring and fault analysis of induction motors: a review

Abstract Induction motors are in high demand for industrial applications due to their reliable and efficient performance compared to other electrical motors. Therefore, the implementation of a predictive maintenance scheme is necessary to minimise financial losses in terms of revenue and maintenance costs. This scheme involves health monitoring system, aimed to enhance motor reliability, extend their useful lifespan and reduce maintenance expenses. Furthermore, traditional health monitoring systems often involve human intervention, which can slow down the diagnosis process. Therefore, a self-sufficient automated system has evolved to improve motor performance and reliability. This paper presents a comprehensive study of induction motor failures and the various methods used for their diagnosis. Additionally, it also discussed the importance of predictive maintenance and the pivotal role played by automated systems in ensuring the sustained efficiency of induction motors in industrial settings.

Optimisation of River State Company (RTC) Fleet Management System

In the transportation industry, fleet maintenance is essential for ensuring operational efficiency, reducing costs, and extending vehicle lifespan. As fleets age, maintenance challenges escalate, impacting profitability. This study focuses on developing a maintenance reliability program for the Rivers State Transport Company (RTC) to optimize fleet management and enhance financial performance. The aim of this study is to assess RTC’s maintenance activities and design a reliability-based program that minimizes costs and maximizes profits. A mixed-method survey research design was adopted, with secondary data collected from RTC’s maintenance workshop records and financial reports spanning from 2014 to 2023. The dataset includes maintenance costs, replacement costs, and income for three vehicle models (Nissan Urvan, Toyota Hiace, and Ford Transit) across multiple years. The Dynamic Recursive Programming Model was employed to analyze vehicle age, cost, and income trends, and Microsoft Excel was used for data processing. The results show a significant rise in maintenance costs as vehicles age. For example, the Nissan Urvan’s maintenance costs increased from ₦1.2 million in 2014 to ₦2.6 million in 2023, representing a 116.7% increase over 9 years. Replacement costs for the Toyota Hiace surged from ₦4 million in 2014 to ₦8.5 million in 2023. Income generated by older vehicles also declined by approximately 25%, with the Ford Transit generating ₦5.1 million in 2023 compared to ₦6.8 million in 2014 due to frequent breakdowns and reduced reliability. The analysis further revealed that replacing vehicles after 6-7 years of operation would maximize financial returns, based on the balance of maintenance and income. In conclusion, maintaining an aging fleet presents a substantial financial burden, with rising maintenance costs, diminishing income, and increasing replacement expenses. This study recommends that RTC adopt a structured vehicle replacement policy guided by the Dynamic Recursive Programming Model. Furthermore, reliability-based maintenance strategies should be implemented to reduce unnecessary expenses and improve the overall performance and profitability of the fleet.

Stochastic economic sizing and placement of renewable integrated energy system with combined hydrogen and power technology in the active distribution network.

The current study concentrates on the planning (sitting and sizing) of a renewable integrated energy system that incorporates power-to-hydrogen (P2H) and hydrogen-to-power (H2P) technologies within an active distribution network. This is expressed in the form of an optimization model, in which the objective function is to reduce the annual costs of construction and maintenance of integrated energy systems. The model takes into account the planning and operation model of wind, solar, and bio-waste resources, as well as hydrogen storage (a combination of P2H, H2P, and hydrogen tank), and the optimal power flow constraints of the distribution network. Electrical and hydrogen energy are administered in an integrated energy system. The modeling of the uncertainties regarding the quantity of load and renewable resources is achieved through stochastic optimization using the Unscented Transformation method. The novelties of the scheme include the sizing and placement of a combined hydrogen and power-based renewable integrated energy system, the consideration of the impacts of bio-waste units, P2H, and H2P systems on the planning of the integrated energy system and the operation of the active distribution network, and the modeling of uncertainties using the Unscented Transformation method to reduce the calculation time. The study's results demonstrate the scheme's ability to improve the technical conditions of the distribution network by considering the optimal planning of integrated energy systems. In comparison to the network power flow, the operation status of the network has been improved by approximately 23-45% through the optimal siting, sizing, and energy management of hydrogen storage equipment, as well as renewable resources in the form of integrated energy systems. In other words, optimal energy management and planning of the integrated energy systems in the distribution network has been able to reduce energy losses and voltage drop by 44.5% and 42.4% compared to the load flow studies. In this situation, peak load carrying capability has increased by about 23.7%. In addition, compared to the case of the network with renewable resources, the overvoltage has decreased by about 43.5%. Also, Unscented Transformation method has a lower calculation time than scenario-based stochastic optimization.

Optimal maintenance and pricing strategy for a periodic review production system with fixed maintenance costs and limited maintenance capacity

In recent times, the escalating complexity of advanced production systems has led to increased exposure to various uncertainties, which impact systems' reliability. To maintain the reliability of systems, reduce maintenance costs and increase revenues, we establish an effective joint optimal maintenance and pricing strategy for a periodic review production system, which is of practical importance. Compared with the existing literature, we make a contribution to considering limited maintenance capacity and fixed maintenance costs simultaneously, and developing joint optimal maintenance and pricing strategies. We initially construct a dynamic programming model for this problem, and prove the objective function is strong CK-concave. We then show that the optimal maintenance strategy is partially characterized by two thresholds, and the optimal pricing strategy depends on the optimal number of operational machines after repairs. Numerical results show that the optimal maintenance and pricing strategy is quite robust, and is not much affected by various parameters.

Utilization of Machine Learning and Explainable Artificial Intelligence (XAI) for Fault Prediction and Diagnosis in Wafer Transfer Robot

Faults in the wafer transfer robots (WTRs) used in semiconductor manufacturing processes can significantly affect productivity. This study defines high-risk components such as bearing motors, ball screws, timing belts, robot hands, and end effectors, and generates fault data for each component based on Fluke’s law. A stacking classifier was applied for fault prediction and severity classification, and logistic regression was used to identify fault components. Additionally, to analyze the frequency bands affecting each failed component and assess the severity of faults involving two mixed components, a hybrid explainable artificial intelligence (XAI) model combining Shapley additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME) was employed to inform the user about the component causing the fault. This approach demonstrated a high prediction accuracy of 95%, and its integration into real-time monitoring systems is expected to reduce maintenance costs, decrease equipment downtime, and ultimately improve productivity.

Unsupervised fault detection using frequency-wise angular filtering in contaminated vibration signals

Manufacturing processes involve multiple machines within a production line. Unexpected faults in machines reduce productivity and increase maintenance costs. Engineers face difficulties in managing numerous machines individually and controlling them immediately. For automatic condition monitoring, several studies have focused on multivariate statistical process control and fault detection based on artificial intelligence. These methods require labeled data or assume that the training data contains only normal patterns. However, obtaining labeled data in the industry is challenging because engineers must manually label the data. Contaminated signals containing fault patterns in unlabeled training data significantly degrade the performance of fault detection in the model. This study proposes Unsupervised fault detection with Frequency-wise Angular Filtering (UFAF) to improve the performance of fault detection in contaminated vibration signals. The UFAF extracts angular features to estimate the normal samples for use only during model training. This filtering strategy is repeated at every epoch and is eventually optimised to use only high-quality normal samples during model training. An experiment using SpectraQuest gearbox datasets confirms the excellent performance for contaminated signals, as angular features are effective in identifying normal and fault signals. The UFAF is practical and applicable in industries wherein it is difficult to collect labeled data.

Decision-Making Model for Life Cycle Management of Aircraft Components

This paper presents a novel decision-making framework for the life cycle management of aircraft components, integrating advanced data analytics, artificial intelligence, and predictive maintenance strategies. The proposed model addresses the challenges of balancing safety, reliability, and cost-effectiveness in aircraft maintenance. By using real-time health monitoring systems, failure probability models, and economic analysis, the framework enables more informed and dynamic maintenance strategies. The model incorporates a comprehensive approach that combines reliability assessment, economic analysis, and continuous re-evaluation to optimize maintenance, replacement, and life extension decisions. The optimization method on the base of genetic algorithm (GA) is employed to minimize total life cycle costs while maintaining component reliability within acceptable thresholds. The framework’s effectiveness is demonstrated through case studies on three distinct aircraft components: mechanical, avionics, and engine. These studies showcase the model’s versatility in handling different failure patterns and maintenance requirements. This study introduces a data-driven decision-making framework for optimizing the life cycle management of aircraft components, focusing on reliability, cost-effectiveness, and safety. To achieve optimal maintenance scheduling and resource allocation, a GA is employed, allowing for an effective exploration of complex solution spaces and enabling dynamic decision-making based on real-time data inputs. The GA-based optimization approach minimizes total life cycle costs while maintaining component reliability, with the framework’s effectiveness demonstrated through case studies on key aircraft components. Key findings from the case study demonstrate significant cost reductions through optimization, with mechanical components showing a 10% more reduction in total life cycle costs, avionics components achieving a 14% more cost reduction, and engine components demonstrating a 7% more decrease in total costs. The research also presents an optimized dynamic maintenance schedule that adapts to real-time component health data, extending component lifespans and reducing unexpected failures. The framework effectively addresses key industry challenges such as no fault found events while minimizing unexpected failures and enhancing the overall reliability and safety of aircraft maintenance practices. Sensitivity analysis further demonstrates the model’s robustness, showing stable performance under varying failure rates, maintenance costs, and degradation rates. The study contributes a scalable approach to predictive maintenance, balancing safety, cost, and resource allocation in dynamic operational environments.

Study of municipal solid waste treatment using plasma gasification by application of Aspen Plus

Abstract Plasma gasification is a viable and efficient technique for handling solid waste, including hazardous waste, while also holding the potential for energy recovery. The objective of this study is to analyze the treatment of municipal solid waste (MSW) using plasma gasification by application of Aspen Plus software. An earlier proposed model was used to analyze the effect of employing different types of gasifying agents as plasma gas, on the composition of syngas produced. The lower heating value, cold gasification efficiency and carbon conversion efficiency were calculated and compared in each case. A sensitivity analysis study was also carried out to observe the effect of variation in plasma gas flow rate and feed flow rate on the composition of syngas generated. The capital, operational and maintenance costs of the process were determined using existing correlations. For a feed rate of 20,000 kg per hour of MSW, the highest yield of syngas (CO (33.48 %), and H2 (34.30 %) with the highest LHV (7.9 MJ/N-m3)) were produced when air was employed as plasma gas. The cold gas efficiency and the carbon conversion efficiency were at their peak when air was used as the gasifying agent. The sensitivity analysis revealed that as the flow rate of plasma gas increases syngas production decreases while the increase in MSW flow rate results in an increase in syngas production. Additionally, the cost analysis revealed that for a plasma gasification plant that can handle 500 tons of MSW per day, the estimated capital and annual operational and maintenance costs are $125,496,721 and $9,833,892 respectively.

Assessment of Oil Paints for Corrosion Protection of Reinforcing Steel Bars in Concrete

Corrosion becomes a threat to reinforced concrete structural integrity during the service life of the structural member. This situation has become a global concern because of the economic and safety implications. Hence, many corrosion inhibition techniques including coating with materials (organic and inorganic) were used to prevent its devastating effects. The present study used three oil paint brands (mat, red-oxide and gloss Leyland oil paints in Ghana) to variedly coat reinforcing steel bars to ascertain their corrosion protection capacity in concrete. There were zero, one, two and three paint coats of average film thickness, 72 µm, 129 µm and 220 µm respectively explored. An Electrochemical process involved simulated concrete pore solution made of 10% sodium chloride (NaCl) in distilled water with three electrode systems to represent a corrosion medium was used to determine corrosion protection capacity of the paints. The results indicate over 90% corrosion resistance of the paints used. Comparing the corrosion rates of the coated and uncoated reinforcing steel bars, while it would take a one coat steel bar 342.5 years and 237 years respectively for spraying and manual brushing modes to reduce in diameter by 0.616mm, it would take only one year for an uncoated steel bar to reduce by this same 0.616mm. The protection capacity proved better with increasing coats on all corrosion variables for the selected paints used. The implication is that the selected oil paint brands are capable of protecting reinforcing steel bars in concrete against corrosion and therefore recommended for application to reduce maintenance cost. Statistically, no significant difference existed between coating mode, number of coatings, paint type and steel type on the corrosion variables for the coated steel bars. However, a further study about the durability of the coating in the concrete during its service time is recommended.

State-of-the-Art Review of Microcapsule Self-Repairing Concrete: Principles, Applications, Test Methods, Prospects

Cement-based materials are widely used in construction worldwide, but they are vulnerable to environmental stressors and thermal fluctuations, leading to the formation of internal cracks that compromise structural integrity and durability. Traditional repair methods such as surface coatings, grouting, and groove filling are often costly and labor-intensive. In response, self-repairing technologies for cement-based materials have emerged as an innovative and promising solution, offering the potential to significantly extend the lifespan of structures and reduce maintenance costs. A particularly novel approach is the development of microcapsule-based self-repairing concrete. In this system, repair agents are encapsulated within microcapsules and combined with curing agents in the concrete matrix. When cracks form, the microcapsules rupture, releasing the repair agents to autonomously heal the damage. This self-repairing mechanism is characterized by its high efficiency, durability, environmental sustainability, and versatility, making it a promising alternative to traditional repair methods. Recent research has focused on the development of microcapsules with various core materials, such as TDI (toluene diisocyanate), IPDI (isophorone diisocyanate), or epoxy resin, as well as composite shell materials including paraffin wax, PE (polyethylene) wax, nano-SiO2, and nano-CaCO3. A novel advancement in this area involves the enhancement of microcapsules through the incorporation of magnetic nanomaterials into the shell, providing new possibilities for self-repairing systems that address cracks in cement-based materials.

Weld Seam Recognition and Tracking With Omnidirectional Wall-Climbing Robot for Spherical Tank Inspection

Abstract Weld seams of in-service pressure storage equipment, such as spherical tanks, require regular inspection to ensure safe operation. Wall-climbing robots can replace manual operations, increasing inspection efficiency and reducing maintenance costs. High precision and fast weld seam identification and tracking are beneficial for improving the automated navigation and spatial positioning of wall-climbing robots. This study proposes a weld seam recognition and tracking method with the omnidirectional wall-climbing robot for spherical tank inspection. Based on deep learning networks, the robot has a front-mounted camera to recognize weld seams and extract weld paths. Weld seam deviation data (drift angle and offset distance) were used in real time to provide feedback on the robot's relative position. For the robot to quickly correct deviations and track weld seams, a seam path-tracking controller based on sliding mode control was designed and simulated. Weld recognition experiments revealed that the robot can accurately recognize and extract weld paths, and the recognition time for each image was approximately 0.25 s. In the weld seam tracking experiments, the robot could successfully track longitudinal and transverse weld seams at different speeds (from 0.05 to 0.2 m/s). During the process of weld seam tracking, the robot angle error was kept within ±3 deg, and the maximum offset distance was less than ±35 mm. Field tests on a 3000-m3 spherical tank were conducted to verify the practicability and effectiveness of the weld seam tracking system. This robotic system can autonomously complete weld seam identification and tracking, which promotes the automation of spherical tank inspection and maintenance.

A Coverage Hole Recovery Method for 3D UWSNs Based on Virtual Force and Energy Balance

Underwater wireless sensor networks (UWSNs) have been applied in lots of fields. However, coverage holes are usually caused by complex underwater environment. Coverage holes seriously affect UWSNs’ performance and quality of service; thus, their recovery is crucial for 3D UWSNs. Although most of the current research recovery algorithms demand hole detection, the number of additional mobile nodes is too large, the communication and computing costs are high, and the coverage and energy balance are poor. Therefore, these methods are not suitable for UWSN hole repairing. In order to enhance the performance of hole recovery, a coverage hole recovery method for 3D UWSNs in complex underwater environments based on virtual force guidance and energy balance is proposed. The proposed method closely combines the node energy and considers complex environmental factors. A series of multi-dimensional virtual force models are established based on energy between nodes, area boundaries, zero-energy holes, low-energy coverage holes, underwater terrain, and obstacle forces. Then, a coverage hole recovery method for 3D UWSNs based on virtual force guidance and energy balance (CHRVE) is proposed. In this method, the direction and step size of mobile repairing node movement is guided by distributed computation of virtual forces, and the nodes are driven towards the target location by means of AUV or other carrier devices. The optimal position to improve coverage rate and node force balance is obtained. Simulation experiments show good adaptability and robustness to complex underwater terrain and different environments. The algorithm does not require precise coverage hole boundary detection. Furthermore, it balances network energy distribution significantly. Therefore, this method reduces the frequency of coverage hole emergence and network maintenance costs.

Blade tip clearance measurement method based on blade tip timing without the once per revolution sensor

Abstract The rotor blade tip clearance (BTC) of an aeroengine is a critical parameter that significantly impacts its performance and safety. Accurately determining dynamic BTC has been a focal point of research in experimental testing. Among the effective measurement methods for dynamic BTC, the laser triangulation method based on blade tip timing (BTT) stands out. However, this method typically relies on an once-per-revolution (OPR) sensor for rotor speed and blade serial number information during measurement, leading to increased installation and maintenance costs as well as additional measurement uncertainties. In response to this challenge, this paper presents two signal processing method that eliminate the need for an OPR sensor. The first method, referred to as "none OPR method 1," closely follows the measurement principles of the traditional method but employs a skewed dual light probe (SDLP) signal to extract rotational speed information, thus obviating the need for an OPR sensor. The second method, "none OPR method 2," introduces a novel data processing approach that enables BTC determination without any source of rotor speed information. Both methods utilize dynamic and static BTC matching techniques to identify blade serial numbers and achieve BTC measurement without an OPR sensor on a simulated rotor experimental bench. Comparative analysis with traditional method BTC measurements reveals that, under stable rotor speeds, all methods achieve similar levels of accuracy and repeatability errors within 0.025mm. However, when dealing with variable speed conditions, the traditional method's accuracy is significantly affected. None OPR method 1 mitigates the impact of rotor speed changes to some extent, while none OPR method 2 remains unaffected by such changes. Importantly, both none OPR methods demonstrate the capability to measure blade-by-blade tip clearances akin to the traditional method.

Modeling and application of maintenance activities in a multi-site company

In an environment where competition is increasingly fierce, companies are constantly looking for new solutions to keep their market share, meet their increasingly demanding customers, and meet the needs of their staff. Companies must develop new innovative products not only of good quality and minimum cost, but also they must meet new constraints, particularly those related to respect for the environment and health / safety. This work is part of the implementation of a maintenance structure in a multi-site production company, with fixed complex homogeneous machines. These machines require, for level 4 and 5 maintenance activities, very expensive mobile specific tools and maintenance characterized by activities with a great variability of tasks in nature, specificity and duration, thus affecting the five maintenance levels. This involves the intervention of technicians of various qualifications and skills, resulting in various hourly labor costs. It is then a question of dimensioning the resources by balancing the loads, by assigning to each level of maintenance a corresponding level of competence. This article concerns the modeling of the activities of this type of maintenance, by proposing a Local Maintenance Workshop (LMW) equipped with a set of mobile tools. The goal is to minimize the cost of maintenance and improve availability.

Improving Freight Train Wheel Monitoring with Smart Sensors

Onboard monitoring of freight car axleboxes enhances safety, reduces maintenance costs, and improves track conditions by preventing secondary damage. Installing wireless sensors on freight cars without a nearby power source should be cost-effective, given the large quantities involved. To address this, a new wireless smart sensor node has been deployed. The sensor automatically recognizes stable operating conditions, detects wheel rotational speed from vibrations, performs real-time condition monitoring, and transmits the results to the cloud. This study outlines the smart sensor concept and the pilot field test conducted with real freight cars. The results demonstrate the ability to estimate wheel rotational speed from vibrations and the potential for detecting wheel out-of-roundness (OOR) using a newly developed condition indicator for low-power real-time operations.

M-CLUSTER: multistage clustering for unsupervised train wheel condition monitoring

Wheel out-of-roundness (OOR) is a common wheel defect that raises maintenance costs and increases the risk of failure or damage to track components. This paper proposes a novel multistage clustering framework (M-CLUSTER) for unsupervised condition monitoring of train wheels. The framework initially extracts time-domain features from raw acceleration responses collected by rail-mounted sensors. Sensitive features are then selected through an unsupervised feature selection algorithm called local learning-based clustering (LLC). Next, a detector model is trained using density-based spatial clustering of applications with noise (DBSCAN), a data clustering method effective for clusters with similar density. Since this algorithm does not originally involve separate training and testing phases, a new two-step mechanism is introduced: (1) training on a healthy dataset and (2) testing on an unlabelled dataset. Finally, the severity of train wheel defects is classified by K-means, with cluster validity indices (CVI) automatically determining the number of severity clusters (classes). The framework’s efficiency is demonstrated through the detection of defective wheels using the Alfa Pendular passenger model. Results indicate that M-CLUSTER accurately identifies train wheel flats and polygonal wear without labelled data, achieving 98% accuracy by selecting 10 features from the set.