Purpose The research proposes an integrated hybrid partial least squares structural equation modelling (PLS-SEM) and Bayesian network (BN) approach to analyse the crucial enablers to Maintenance 4.0 adoption in manufacturing industries. Design/methodology/approach A detailed review of the literature is executed to identify the crucial enablers for maintenance adoption in the manufacturing sector. An initial hypothesised model is developed, factors loadings are evaluated, and model fit is confirmed using a PLS-SEM. In addition to that, a BN is modelled to examine the influence of enablers on the adoption of Maintenance 4.0 in the manufacturing sector. Further, to verify the robustness of the outcome of the study, sensitivity analysis is conducted. Findings The analysis reveals that the proposed PLS-SEM model is accurate and valid, categorising the enablers into four groups: organisation-related enablers, digital infrastructure and system support–related enablers, data-related enablers and people-related enablers. A machine learning-based BN technique was implemented to investigate the impact of the identified enablers on the adoption of Maintenance 4.0 in the manufacturing sector. Additional sensitivity analysis confirmed the accuracy of the BN model. Research limitations/implications The proposed methodology, adopted slightly complicated and laborious, provides an accurate and reliable framework for the BN model, which provides a more accurate and clear understanding of the enablers. Practical implications The findings of this research will be highly beneficial to manufacturing industries, enabling them to identify the most influential enablers for successful Maintenance 4.0 adoption and, consequently, improve the probability of success. Originality/value A hybrid framework combining PLS-SEM and machine learning is employed to examine complex systems such as Maintenance 4.0 enablers, allowing both statistical robustness and data-driven insights. The framework offers useful direction for policymakers and industry practitioners by clarifying how critical enablers can be utilized to improve the success rate of successful Maintenance 4.0 adoption.
Purpose This paper aims to address the limitations of Duncan's 1971 economic model for the X‾ control chart by resolving its internal contradictions and enhancing its applicability through integration with preventive maintenance strategies. Design/methodology/approach The research revisits Duncan's 1971 economic model, which assumes both independence and dependence among assignable causes – an inherent contradiction that undermines its realism. A new economic design framework is developed that accommodates multiple assignable causes and incorporates preventive maintenance strategies to improve process control. Findings The proposed model demonstrates improved reliability and extended production time under controlled conditions. It also significantly reduces unplanned downtime and overall operational costs compared to models based on Duncan's assumptions. Originality/value This paper challenges the foundational assumptions of a widely cited economic model and introduces a novel, more realistic approach that combines statistical process control with preventive maintenance. The model offers practical value for industries seeking to optimize quality and reliability in production and service systems.
Purpose This study aims to improve maintenance reliability by strategically allocating the workforce, considering human factors and knowledge management in critical asset maintenance within the public transport sector. Design/methodology/approach The study proposes a quantitative-applied approach integrating multi-criteria decision-making (MCDM), an extended risk priority number (ERPN) for task criticality and a modified priority matrix. The algorithm optimizes resource allocation by mathematically combining operator technical aptitude and willingness. Findings Implementation of the algorithm resulted in an average 47% reduction in repair times for critical machinery. This operational efficiency generated an estimated annual saving of $91386.77, proving that 98% of the economic benefit stems directly from minimizing asset downtime rather than reducing direct labor costs. Research limitations/implications This study was applied in a single transport company, and the results are specific to its operational constraints, which limits direct generalizability to other industrial sectors. Tacit knowledge quantification and task prioritization relied on expert consensus. The model currently assumes full staff availability and does not account for simultaneous unexpected failures. Practical implications The methodology provides maintenance managers with a structured, data-driven tool to transition from subjective, ad-hoc personnel assignments to an objective protocol. It allows for the systematic integration of knowledge management into computerized maintenance management systems (CMMS), optimizing hour-machine productivity across heavy fleets. Social implications The formal recognition of tacit knowledge promotes equity in task allocation and addresses the human reliability gap. By mitigating unequal workloads and recognizing individual technical aptitude, the proposed framework fosters a transparent, motivating and highly engaged work environment, which is critical for sectors operating under severe operational pressure. Originality/value This study addresses a critical gap in industrial fleet maintenance by mathematically operationalizing not only verified tacit knowledge but also operator willingness, integrating human attitude as a quantifiable variable to reduce system execution entropy.
Purpose The maintenance of public transport systems, for example trains, buses and airplanes, plays a crucial role in ensuring their availability, safety and cost-effectiveness in their respective means of transport. Effective maintenance planning and control (MPC) not only reduce operational disruptions in these public transport systems but also optimize resource allocation and minimize overall costs. However, with the increasing complexity of today’s public transport networks, there is a growing need for advanced, data-driven maintenance approaches. In this regard, artificial intelligence (AI) has emerged as powerful technology in MPC. AI leverages historical and real-time data to predict potential failures, optimize maintenance schedules and improve decision-making processes. Design/methodology/approach This study systematically reviews and examines existing research on AI approaches for MPC across public transport sectors. The preferred reporting items for systematic literature reviews and meta-analysis (PRISMA) guidelines are employed to ensure a transparent and structured screening process. Descriptive analysis are performed to present emerging research trends, while detailed content analysis synthesizes the findings of the most relevant articles in the research domain. Findings The analysis results show that most of the research focuses on the specific aspects such as predicting the remaining useful life and faults diagnoses, in isolation. However, in practice, MPC involves a complex coordination of many activities, for example inspection, resource allocation and workforce and tasks scheduling. Therefore, the authors believe that there is still need for more holistic solutions that can integrate maintenance activities and shopfloor constraints, moving beyond single-component predictions to system-level maintenance optimization that could be validated in real maintenance facilities. Originality/value The original contribution of this work is to provide the state of the art of AI in MPC for public transport.
Purpose This paper proposes a conceptual prescriptive maintenance (PsM) framework for a heterogeneous fleet of degrading vehicles, aiming to synchronize maintenance schedules and reduce inventory and operational costs. Design/methodology/approach A mathematical framework is developed using real-time degradation data from inspections. The policy prescribes two actions: vehicle swaps to optimize fleet management and usage rate adjustments to influence individual degradation trajectories. A simplified case study with numerical examples illustrates the approach. Findings Results show that the policy effectively manages fleet heterogeneity and achieves cost savings compared to conventional maintenance strategies. Originality/value The contribution lies in explicitly linking system usage with degradation. Unlike predictive or condition-based approaches, the policy prescribes both maintenance actions and operational adjustments, broadening the decision space for fleet optimization.
Purpose The purpose of this study is to develop and validate a robust framework for data-driven predictive maintenance (PdM) that estimates the remaining useful life (RUL) of equipment and signals the optimal time to initiate maintenance activities. By integrating statistical modeling and machine learning techniques, the proposed framework aims to minimize unplanned downtimes, reduce maintenance costs and enhance operational efficiency. It addresses critical gaps in existing methods by providing real-time condition monitoring and predictive alerts, enabling maintenance personnel to make informed decisions and optimize maintenance scheduling. Design/methodology/approach This study introduces a data-driven predictive maintenance framework designed to estimate the RUL of equipment and provide timely signals for initiating maintenance tasks. The framework utilizes a combination of statistical modeling and machine learning algorithms. A Weibull distribution with time-varying parameters is employed to model the RUL, while random forest and exponentially weighted moving average (EWMA) control charts are integrated for failure prediction and maintenance signaling. The methodology is validated using a synthetic dataset that simulates real-world scenarios, enabling robust evaluation of the proposed approach in terms of accuracy and reliability. Findings The findings demonstrate that the proposed predictive maintenance framework effectively estimates the RUL of equipment with high accuracy and provides timely signals for initiating maintenance tasks. Validation on a synthetic dataset reveals that the framework consistently predicts failure probabilities and generates alerts well in advance, allowing sufficient time for maintenance planning. The integration of Weibull distribution modeling and Random Forest classifiers enhances the reliability of predictions, while the use of EWMA control charts ensures robust monitoring of failure probabilities. These results highlight the framework's potential to reduce downtimes and maintenance costs. Originality/value This study introduces a novel data-driven predictive maintenance framework for accurately estimating the RUL of equipment and providing timely signals for maintenance actions. Unlike existing methods, the proposed approach integrates a probabilistic model with machine learning techniques, leveraging time-varying Weibull distributions and advanced statistical tools for robust predictions. The framework not only improves failure prediction accuracy but also enhances maintenance planning by signaling appropriate times for action. This contributes to reducing downtime and costs while increasing operational efficiency, offering a significant advancement for Industry 5.0 and smart manufacturing applications.
Purpose Rolling bearing failures remain a primary cause of induction motor breakdowns, creating significant reliability and maintenance challenges. This study aims to enhance fault diagnosis robustness under variable-speed conditions by addressing the limitations of fixed-speed datasets and single-model deep learning approaches. Design/methodology/approach An ensemble deep learning model (EDLM) is developed by integrating convolutional neural networks (CNN), deep belief networks (DBN) and stacked autoencoders (SAE). The framework applies a weighted-fusion strategy to exploit the complementary strengths of the base learners. Diagnostic performance is evaluated using three input modalities: raw vibration signals, spectrograms and infrared thermal images. Findings The EDLM consistently outperformed individual models across all metrics. Among the modalities, infrared thermal images achieved the highest diagnostic accuracy (98%), demonstrating superior capability in capturing subtle fault features under variable-speed conditions. Originality/value To the best of current knowledge, this study presents the first ensemble of CNN, DBN and SAE for bearing fault diagnosis under variable-speed conditions. By validating multi-sensor inputs and highlighting the diagnostic advantage of infrared thermography, the work provides a reliable and scalable solution for real-world machinery health monitoring.
Purpose The maintenance function is crucial for maintaining competitiveness, safety and environmental responsibility. As demands for quality and production efficiency increase, optimized maintenance becomes more essential. Industry 4.0 and 5.0 introduce new generations of maintenance, highlighting technical and human-centered approaches. However, manufacturing companies still face many challenges in implementation and use. Prior research lacks studies that support the manufacturing industry and have not been practically connected to it. This research explores the implementation and use of smart maintenance technologies in large Swedish manufacturing companies, offering practical recommendations for industry practitioners and contributing to the field of smart maintenance research.Design/methodology/approach The research is based on 12 semi-structured interviews with respondents from 11 large manufacturing companies representing varying levels of experience and maturity in smart maintenance technologies. The empirical data were analyzed qualitatively to identify themes associated with such technologies.Findings This research identifies and describes three themes associated with smart maintenance technologies in large manufacturing companies: organizational, human and technical. These themes do not outline an implementation process, but rather synthesize the practical experiences of participating companies, increasing the understanding of how smart maintenance technologies are implemented and used in industrial practice.Originality/value This research highlights developments in maintenance for both practitioners and researchers. It compiles insights from participating companies using smart maintenance technologies to improve understanding of their practical application in industry.
Purpose This study proposes to identify the relationship between asset management (AM) maturity and business performance, thereby building a theoretical model that relates AM competencies to business performance. Design/methodology/approach This is tested by analyzing the application of partial least squares structural equation modeling (PLS-SEM) in a sample of 70 asset-intensive Brazilian companies that have conducted a maturity assessment with multiple evaluators. Findings The results reveal that AM maturity influences business performance, whereas AM capabilities related to strategic AM have an impact on different AM competencies. In addition, having the capability to exploit Risk and Review competence demands maturity in Leadership and People and Asset Information. Practical implications The finding motivates AM decision-makers to develop AM Maturity by fostering the improvement of AM capabilities, since AM Maturity contributes to business performance. In addition, this prompts such decision-makers to explore the technologies and practices that have been applied in mature organizations whether in Brazil or in other countries. Originality/value These findings significantly advance the theoretical underpinnings of AM in validating the structural model that demonstrates the relationship between AM dimensions, AM maturity and business performance, which can be used in asset-related decisions that are expected to have an impact on business performance.
Purpose The purpose of this study is to develop a more realistic cumulative damage replacement framework for systems subjected to shocks and prior wear. Existing models assume new equipment with no residual degradation, which can misguide maintenance decisions for fairly- used machinery common in developing economies. This study incorporates initial damage and a shock-dampening factor to capture reduced vulnerability to shocks. It evaluates age-based, shock-based, damage-based and joint replacement policies using an industrial washing machine, with the goal of identifying cost-effective strategies and improving maintenance planning for partially worn, shock-resistant systems. Design/methodology/approach A modified cumulative damage model is proposed, incorporating two key factors: initial residual damage and shock-dampening factor that attenuates the impact of external shocks. Three practical single replacement policies: age-based, shock-count-based and damage-threshold-based are formulated and analytically evaluated. Furthermore, a joint criteria maintenance model is developed that provides the jointly optimal result for any fairly used and new machines. These models are applied to an industrial washing machine case study to compare the cost performance of each policy and a unified joint policy. Findings The modified cumulative damage model incorporating prior wear and shock dampening accurately predicts optimal replacement decisions for fairly- used systems. For the industrial washing machine, optimal single-policy decisions occur at 400 h, 8 shocks and 700 damage units, with the age-based strategy yielding the lowest cost. The joint policy identifies an optimal solution at 450 h, 8 shocks and 800 damage units, achieving a lower overall cost than the combined single-policy costs. Compared with the baseline model assuming no prior damage or shock resistance, the proposed framework provides more realistic replacement intervals and avoids misleading maintenance delays. Research limitations/implications The study assumes constant shock-arrival behaviour and a fixed shock-dampening factor, which may not fully capture machines operating in highly variable environments. Parameter estimation relies on external data, limiting precision where local data are unavailable. The model focuses on non-repairable systems and may require adaptation for partially repairable assets. Future work should incorporate dynamic shock resistance, variable operating conditions and improved field data collection to enhance accuracy. These extensions would expand applicability across diverse industrial settings and strengthen the generalizability of cumulative damage-based maintenance policies. Practical implications The proposed policies provide maintenance managers with realistic replacement intervals for fairly- used machinery, preventing premature failure and reducing unplanned downtime. By integrating prior wear and shock attenuation, it supports more accurate lifecycle cost estimation and improves planning for second-hand equipment common in resource-constrained economies. The joint policy helps avoid redundant replacements, offering a more economical strategy than single-criterion approaches. Overall, the model strengthens decision-making for preventive replacement, enhances equipment availability and reduces long-term operational costs. Social implications Improved maintenance planning for fairly- used machinery supports operational continuity in developing economies where second-hand equipment is widely used. By reducing unexpected failures, downtime and repair expenses, the model contributes to safer work environments and more reliable service delivery in sectors such as manufacturing, healthcare and utilities. Cost-effective replacement strategies can also lower the financial burden on small and medium enterprises, promoting economic resilience. The approach ultimately enhances productivity and supports sustainable equipment use in resource-limited settings. Originality/value This study introduces a generalized cumulative damage framework that, for the first time, explicitly incorporates both residual wear and shock-dampening behaviour into replacement modelling for fairly- used systems. It extends traditional models by enabling simultaneous evaluation of age, shock count and cumulative damage through a joint policy that minimizes redundant replacements. The model unifies cases for new and used equipment, reducing to existing formulations when no prior damage or shock resistance is present. Its ability to capture realistic degradation mechanisms offers significant value for maintenance engineering, especially in contexts dominated by second-hand machinery.
Purpose Several studies have focused on maximizing the efficiency of solar energy conversion in photovoltaic (PV) systems. The accumulation of dirt on the surfaces of PV modules is a primary environmental factor that reduces performance. Periodic cleaning, implemented as preventive maintenance, mitigates these losses; however, it also incurs costs. Minimizing total system cost therefore requires determining the optimal cleaning interval based on techno-economic considerations. This study develops a computational optimization framework to determine the preventive maintenance interval for PV modules, balancing energy yields against operational costs.Design/methodology/approach The proposed computational tool was developed in a Python-based integrated development environment. It was designed to be generic and flexible, allowing users to input key parameters, such as the installed system capacity, cost of preventive maintenance, energy compensation tariff, average peak sun hours, and power-loss function due to soiling, allowing for both linear and exponential models. Based on these inputs, the tool calculates the optimal maintenance interval that minimizes the total system cost for various operating scenarios.Findings This study shows that using an optimized maintenance interval rather than relying on generic fixed schedules can lead to significant annual savings. The simulation results demonstrate the sensitivity of the optimal interval to variations in maintenance costs and soiling-related energy generation losses. Additionally, differences in installation type, local environmental conditions, and system scale affect the maintenance strategies.Originality/value The proposed tool advances the current state of practice by integrating both linear and asymptotic exponential models to represent the output power loss caused by soiling, thereby accommodating diverse operational conditions of PV modules, unlike most tools currently available. The optimization process employed analytical and numerical techniques to determine the optimal preventive maintenance interval. Using a linear formulation, the analytical approach ensured convergence to a global minimum. For an exponential formulation, the method combined analytical feasibility conditions with numerical procedures to identify optimal solutions efficiently. In addition, the study introduces a graphical diagram that illustrates the relationship between preventive maintenance cost and the optimal maintenance interval for different installed capacities. These diagrams function as practical decision-support instruments, enabling the rapid assessment of cost-effective maintenance strategies tailored to specific system configurations.
Purpose The automobile sector in India is one of the key segments of Indian economy as it contributes to 4% of India's gross domestic product (GDP) and 5% of India's Industrial production. Through this study, the effect of risk and frangibility in the Indian automobile supply chains were observed.Design/methodology/approach A Failure Mode and Effects Analysis (FMEA) technique is used to prioritize the various types of risk into zones, namely high, medium and low risk factors. Also, the best-worst method (BWM) is used to find out the weightings of different risk factors to address frangibility in supply chains.Findings Various risk factors in the automobile supply chains were observed and prioritized in the study. Using risk priority ranking technique like FMEA, the risk factors under main risk drivers of supply chains were ranked considering their overall effects, into various risk zones like high, medium and low. Further, the BWM was used to weigh and rank these risks.Research limitations/implications The results can help organizations to prioritize various risk factors and propose appropriate risk mitigation strategies to upsurge supply chain reliability, efficiency and responsiveness.Practical implications By identifying to what risk zone the particular risk factor belongs; managers can come up with a proper risk mitigation strategy. This can help them understand how much resources need to be invested, thereby saving useful production time and cost.Social implications The theoretical implications of the study are based on the Normal Accident Theory perspectives, where it is assumed that complex and tightly coupled systems can inevitably fail. Managers in the automobile industry can consider reducing complex and tight coupling, while designing their supply chain networks.Originality/value The risk factor identification and prioritization of the same in the Indian automobile industry is scarcely attempted in literature. A combination of FMEA and BWM methods was used for effectively prioritizing the same.
Purpose The main purpose of this study is to find out the relationship between the existing conditions in the railway lines, so that by using it we can take a main and initial step to deal with chain failures. When the experts of maintenance and repairs in railway lines know that a failure in a geometric parameter or in general a change in the state of a geometric parameter will affect the condition of other parameters, they can better deal with the failures and make important decisions. Design/methodology/approach In this study, the studied data were analyzed using data mining methods such as regression, decision tree, Bayesian classification and associative rules. The main approach in this study is to discover the connections in the geometry of railway lines. The steps to reach the results of the research are as follows: (1) collection of railway line failure data, (2) cleaning the collected data, (3) entering the collected data into the mentioned models and (4) discovering the relationship between the conditions of each of the geometric parameters and the breakdowns that occurred on the track. Findings There is a non-negligible relationship between track gauge (GAU) failure and other geometric parameters. The conditions of the gauge track geometric parameters are most affected by the conditions of the alignment level and twist parameters, that is, failure in these parameters occurs like a chain. Using the results of association rules can be effective in reducing the inspection (according to the explanation in the article). Originality/value This paper predicts GAU conditions by analyzing other track parameters. The linear polynomial regression method and Bayesian classification are used to identify the fault causing GAU failure. Additionally, GAU failure is estimated based on the failure of other parameters through regression analysis.
Purpose The objective of study is to develop a computationally efficient decision-support model to generate actionable maintenance plans under strict budgetary and performance constraints. This study addresses the critical challenge of resource allocation optimization for pavement maintenance in regional road networks, focusing on the inherent conflict between minimizing agency cost and maximizing network quality under stringent budget and safety constraints.Design/methodology/approach The study develop an optimization framework using multi-objective particle swarm optimization (MOPSO) specifically adapted for discrete maintenance variables. To ensure rigorous adherence to operational constraints, the external penalty function (EPF) method was integrated into the algorithm to strictly manage non-negotiable budget caps and mandated minimum quality levels. The model was applied to a case study of N = 25 intercity road sections in Northwest Iran, utilizing localized unit maintenance costs (denominated in Toman/m2) and an expert-defined quality grading system. Analysis across eight distinct management scenarios - covering weak, medium and high initial quality networks under varying budget levels - was conducted to generate actionable Pareto optimal fronts.Findings The results demonstrated a clear non-linear relationship: networks with a weak initial quality (Qavg = 0.47) required the maximum budget (up to 32.44 Billion Toman) to achieve a final cumulative quality of 22.37. In contrast, higher-quality networks (Qavg = 0.67) achieved a superior final quality of 24.15 with a substantially lower budget (as low as 24.73 Billion Toman). This quantitatively confirms that a proactive maintenance strategy focused on preserving high-quality assets yields a superior cost-efficiency and return on investment. The study's principal contribution is providing a robust, fast-acting, and constraint-compliant decision-support tool that bridges the research gap between sophisticated MOO (multi-objective optimization) theory and practical, localized pavement management under conditions of financial austerity in developing economies.Originality/value The core methodological innovation lies in the explicit application of the EPF method to strictly enforce the non-negotiable budget cap and a mandated minimum quality level (Q min). This rigorous constraint handling transforms the complex multi-objective problem into an unconstrained form, using a squared penalty for quality violations to prioritize network integrity.
Purpose Existing Department of War methods for assigning facility maintenance budgets rely on uniform funding level models from an average facility of the corresponding building type. Such an approach does not factor in facility age, condition, unique inventory or environmental factors. Current optimization approaches tend to scale poorly in scenarios that may require millions of individual component-level decisions. There is a need for more prescriptive, reliability-focused facility sustainment funding methods capable of being applied at a large scale. Design/methodology/approach An enhanced statistical reliability-centered maintenance (RCM) framework is proposed that integrates meaningful covariates like climate zones and inspection data. These models are used to assign optimal maintenance intervals to maximize the return-on-investment (ROI) from renewal activities that extend equipment service life. This approach is further amended to incorporate budget constraints and adjust metrics to compensate for deferred maintenance. Findings Case studies against 30 facilities from general administrative buildings and barracks show that the average sustainment levels using the developed method are reasonable compared to existing budget levels and industry standards, but provide much more prescriptive funding tailored to the individual facility inventory. Time tests also show the approach to be a highly scalable method, capable of generating millions of component-level decisions without the need for high-performance computing resources. Practical implications The developed framework provides a scalable approach to assigning facility maintenance funding rooted in reliability-centered maintenance. Compared to competing approaches, it implements better risk-based decisions via covariate-based reliability models and can be scaled without high-performance computing resources or excessive decision heuristics. Social implications Because the developed model helps more efficiently allocate funding, it provides a better use of budgets, particularly in large-scale public domains. In addition, the method implements climate-based reliability models that can adjust risk-based decisions due to climate change. Originality/value The proposed approach develops a novel, scalable approach to assigning optimum maintenance activities suitable for large organizations. Reliability models with climate covariates are built from historical component inspections to inform the model. Compared to competing approaches that implement integer programming methods, this approach transforms the problem into a continuous variable optimization problem so that more efficient algorithms provide scalable solutions. Additionally, the method introduces an adjustment approach to better manage the risk of deferred maintenance.
Purpose The model feature extraction is enhanced by suppressing noise interference and optimizing feature sensitivity to improve its robustness in practical applications. Design/methodology/approach A fault diagnosis method for rolling bearings based on convolutional neural networks in a strong noise environment. Findings Experiments show that this model demonstrates high robustness and generalization ability under noisy conditions. Originality/value It provides a novel framework for industrial fault diagnosis to solve the problem of fault signals being submerged by noise.
PurposeBalancing cost-effectiveness and availability against safety and environmental stewardship remains a persistent challenge in asset management. Conventional Risk-Based Inspection (RBI) often depends on imprecise qualitative assessments, while modern data-driven methods frequently lack transparency. This study addresses this gap by introducing the multidimensional risk analysis (MDRA) framework, which offers a comprehensive, transparent, and actionable risk profile for safety-critical assets.Design/methodology/approachA hybrid intelligence system utilizing a weighted, stacked ensemble classifier was developed to predict the probability of failure (PoF). Bootstrap resampling was employed to generate 95% confidence intervals for uncertainty quantification. To address the limitations of the "black box" nature of standard AI, SHapley Additive exPlanations (SHAP) were integrated to ensure full model interpretability. Finally, a hierarchical, rule-based engine translated these probabilistic outputs into actionable inspection plans for industrial assets.FindingsThe application of the MDRA framework successfully demonstrated the capacity to generate reliable and explainable inspection priorities. By integrating deterministic rules with probabilistic machine learning insights, the system produced robust, data-driven inspection plans that accounted for uncertainty. The framework proved capable of identifying the equilibrium between six critical operational values: safety, integrity, reliability, availability, environmental responsibility, and cost-effectiveness, translating complex risk data into clear, actionable maintenance decisions.Originality/valueThis paper presents a new framework that overcomes key limitations of traditional risk-based methods and opaque advanced analytics. Its main contribution is the integration of explainable AI (XAI) and uncertainty quantification into safety-critical asset management. By combining model interpretability with probabilistic risk estimates, the MDRA framework improves the transparency, robustness, and reliability of data-driven risk assessments. This dual focus on explanation and uncertainty allows for more confident and context-aware prioritization of inspection and maintenance activities, supporting operational efficiency while upholding safety and environmental standards.
PurposeThis paper presents a proposal and verification of a new immersive methodology that employs a collaborative VR tool to assess both qualitative and quantitative aspects of maintainability in the early stages of new product development.Design/methodology/approachThe industry 4.0 era and its technological advancements are transforming product development by integrating tools such as Virtual Reality, enabling more efficient design processes, real-time collaboration, and the simulation of product concepts in digital environments prior to physical production.FindingsAn experiment involving 21 participants was conducted, in which maintainability analyses were performed using the proposed methodology to compare two design alternatives (A and B) across two different case studies (1 and 2). The most suitable solution among the evaluated alternatives was identified using the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), a multi-criteria decision-making method. Furthermore, quantitative ergonomic assessments were conducted through the immersive Virtual Reality environment in combination with the Ovako Working Posture Analysis System (OWAS), a posture classification method.Originality/valueThe results demonstrate strong potential for the proposed methodology, as evidenced by participants' expressed interest in incorporating the VR tool into their regular workflows.
PurposePreventive maintenance (PM) policies for repairable systems often assume constant effectiveness and uniform failure consequences. This paper integrates age-dependent PM effectiveness with severity-aware estimation to produce availability-optimizing PM schedules that are both condition- and consequence-sensitive.Design/methodology/approachWe develop a virtual-age non-homogeneous Poisson process model in which each PM type \(k\) induces an age-dependent restoration factor phi k(t) is an element of [0,1]. Corrective-maintenance (CM) failures enter the likelihood with downtime-derived weights f_c, yielding a severity-weighted pseudo-likelihood. Two calibrated baselines are estimated at OEM intervals: (1) unweighted (OEM) and (2) severity-aware (weighted). We then optimize PM schedules under both unweighted and weighted phi k(t) using a genetic algorithm, with feasibility bounds subject to operational constraints t1 is an element of [200,300]\), t2 is an element of [450,550]\), t3 is an element of [800,1000]\) hours.FindingsRelative to OEM estimation, severity-aware fitting increases inferred failure intensity and systematically lowers phi k(t) for light and moderate PMs, producing more conservative reliability and improved model fit. Incorporating age-dependent phi k(t) reshapes the optimization landscape: weighted schedules favor slightly earlier moderate PM (Type II) while maintaining comparable availability. Back-testing shows closer alignment between predicted and observed failure trajectories, particularly for high-downtime events, demonstrating the value of consequence-aware, age-based PM modeling beyond availability alone.Research limitations/implicationsSeverity is proxied by downtime; multidimensional consequence metrics, i.e. safety, cost, quality and schedule, are future work. Sparse counts at extreme ages can widen uncertainty in phi k(t). Unit heterogeneity suggests hierarchical extensions.Practical implicationsThe method operationalizes risk-sensitive PM: planners can jointly tune timing and consequence-aware effectiveness to reduce severe disruptions without over-servicing. The approach is data-driven and CMMS-ready.Originality/valueThis is the first synthesis to jointly (1) model PM effectiveness as a function of age and (2) estimate it via a severity-weighted virtual-age likelihood, then (3) optimize schedules under real feasibility constraints using real downtimes in chronological order.
