Search

Abstract: Inventory pooling occurs when different actors share their inventories with the aim of reducing costs while improving performances. The paper-making industry requires large amounts of ready to install spares aimed at assuring high availability, but presents the drawback of excessive inventory costs. Pooling inventories, with the support of appropriate IT solutions, can be considered an effective strategy to improve availability while reducing the total costs. In the present work a real-life implementation within an important Italian paper-making district has been successfully carried out. Throughout the study the most recent theoretical models have been adapted after a methodical analysis that has been devoted to allow the identification and of the most significant issues and the building of a wide and solid agreement among the participants. Indeed, this required a thorough investigation, since the involved actors are generally competitors and only agree to pool inventories if this leads to increasing profits for themselves. [Copyright &y& Elsevier]

Abstract: To meet always increasing safety requirements in car industry, design and safety assessment methods are developed in order to fit the complexity of new embedded mecatronic systems. Hybrid (discrete/continuous) and dynamic features, specific to these systems, require choosing a suitable formalism. These features should also be considered in safety studies made all through the system design. The aim of this paper is to propose a quantitative analysis method based on the construction of an aggregated Markov graph, which allows a limitation of the combinatorial expansion. This graph is directly deducted from the Petri net modelling of the system. It is composed by a set of functional modes and a set of transitions to which statistical information regarding the system dynamics has been added. [Copyright &y& Elsevier]

Abstract: In this paper a maintenance policy is optimised for a two-unit system with a parallel structure and stochastic dependences. Monitoring problems are taken into account in the optimisation scheme: the failure time of each unit can be not detected with a given probability. Conditions on the system parameters (unit failure rates) and on the non-detection probabilities must be verified to make the optimisation scheme valid. These conditions are clearly identified. Numerical experiments allow to show the relevance of taking into account monitoring problems in the maintenance model. [Copyright &y& Elsevier]

Abstract: This paper deals with a perfectly continuously monitored system which gradually and stochastically deteriorates. The system is renewed by a delayed maintenance operation, which is triggered when the measured deterioration level exceeds an alarm threshold. A mathematical model is developed to study the asymptotic behavior of the reliability function. A procedure is proposed which allows us to identify the asymptotic failure rate of the maintained system. Numerical experiments illustrate the efficiency of the proposed procedure and emphasize the relevance of the asymptotic failure rate as an interesting indicator for the evaluation of the control-limit preventive replacement policy. [Copyright &y& Elsevier]

Abstract: Nowadays, the complex manufacturing processes have to be dynamically modelled and controlled to optimise the diagnosis and the maintenance policies. This article presents a methodology that will help developing Dynamic Object Oriented Bayesian Networks (DOOBNs) to formalise such complex dynamic models. The goal is to have a general reliability evaluation of a manufacturing process, from its implementation to its operating phase. The added value of this formalisation methodology consists in using the a priori knowledge of both the system''s functioning and malfunctioning. Networks are built on principles of adaptability and integrate uncertainties on the relationships between causes and effects. Thus, the purpose is to evaluate, in terms of reliability, the impact of several decisions on the maintenance of the system. This methodology has been tested, in an industrial context, to model the reliability of a water (immersion) heater system. [Copyright &y& Elsevier]

• Digital Twin approach for monitoring and prognosis of ship-specific fatigue damage. • Approach using publicly available wave condition data in lieu of onboard sensors. • Fatigue prognosis using Monte Carlo simulations to account for wave variability. This paper proposes a Digital Twin approach for the monitoring and prognosis of vessel-specific fatigue damage. During design, fatigue damage estimates are based on conservative assumptions regarding operational conditions and structural response. However, variability in the vessel-specific operations from those assumed during design needs to be considered when supporting engineering-based decisions for maintenance deferrals and service life extensions. The use of Digital Twins is proposed to provide this necessary vessel-specific decision support. Digital Twins typically rely on sensor-based data to update their models, however structural health sensors for fatigue monitoring can be prohibitively expensive to install and maintain in ship structures, so the proposed method addresses this by instead combining publicly available vessel-specific operational data (global vessel position data coupled with metocean hindcast data) with computational models to monitor the environmental exposure and track the vessel fatigue accumulation over time. This approach is demonstrated through a case study of a containership that has been in operation for seven years. The results of proposed approach are compared against the fatigue estimate obtained using the design reference wave conditions. The Digital Twin is then used to forecast the remaining fatigue life, in order to support inspection and maintenance scheduling and operational decision-making. [ABSTRACT FROM AUTHOR]

Digital twins are a significant way to achieve fault detection of various smart manufacturing, which provide a new paradigm for complex industrial process monitoring. Wastewater treatment processes play a crucial role in water recycling, its failures may cause risks of adverse environmental impacts. This paper studies the digital twins fault detection framework based on the convolutional autoencoder for wastewater treatment processes monitoring. The designed digital twins fault detection framework can simulate the sludge bulking failure and the toxic impact failure conditions in the virtual space to construct the simulation data with continuous updating through wastewater data. The simulation data is divided into rate of change information sub-block, original sub-block, and cumulative information sub-block using the multi-block modeling strategy to fully explore the hidden information. Further, the sliding window method is utilized to resample the reconstructed sub-blocks to enhance the effects of the detection performance. Bayesian fusion is adopted, and the final decision is made based on the fused statistical value and the control limit. The comparison experiments tested on the digital twins fault detection framework demonstrate the superiority and feasibility of detection performance. [ABSTRACT FROM AUTHOR]

• Description of the state-of-art research in fleet maintenance and identification of the gaps of existing work. • Optimisation model incorporating preventive and predictive maintenance while balancing workload to meet fleet demand. • Examination of the role of component criticality as well as the precision of the RUL prognosis. • A universal model that can be adapted for multiple sectors with similar operational constraints. The optimization of fleet maintenance management is of utmost importance to ensure the efficient and reliable operation of asset fleets. Traditional maintenance strategies are often reactive or rely on predetermined schedules, which can lead to inefficient resource allocation and increased operational costs. The advent of digital technologies has allowed the integration of predictive maintenance into fleet management. This paradigm shift towards a more data-driven approach enables fleet management to dynamically respond to issues identified through sensors and algorithms that detect anomalies and provide prognostic insights regarding the remaining useful life of components. However, a notable deficiency exists in the integration of predictive maintenance with calendar-based preventive maintenance and fleet operational allocation, thereby impeding value creation for businesses. This paper presents an optimisation model to address this issue by incorporating preventive and predictive maintenance while simultaneously striving to balance the workload to meet operational demand and mitigate potential penalties resulting from failure to meet these demands. The paper also examines the role of component criticality as well as the precision of the RUL (Remaining Useful Life) prognosis. Through the experiments conducted, it has been demonstrated that the allocation of the fleet is subject to change depending on the level of criticality of monitored components. These findings reveal the potential risks and penalties that could arise from an insufficient definition of failure impact severity when integrating predictive maintenance with existing preventive approaches and operational workload balance. Additionally, the experiments underscore the impact of RUL distributions precision on total operating costs, showing the confidence intervals through a sensitivity analysis. [ABSTRACT FROM AUTHOR]

• Two-kind combined reliability systems are introduced in terms of two reliability systems: (n , f , k) and < n , f , k >. • Some important concepts such as irreducible structure and structure equivalence are introduced. • We present an important criterion: four-class combined reliability systems are needed to be considered in essential. • A coordinate system is introduced for representing the working and failure regions for the combined reliability systems. • A procedure for calculations of the combined reliability systems is given. In the paper, two-element (f , k) and < f , k > combined systems are developed, and two-kind combined reliability systems and some related new concepts such as irreducible structure and structure equivalence are introduced, which can be used not only in real practical situations, but also enrich the reliability contents. We proved that four-class combined reliability systems need to be considered actually, which facilitates the handling of these combined reliability systems. Meanwhile, a coordinate system with two axes, the number of total failed components and the number of maximal consecutive failed components, is introduced in the paper for representing the working and failed areas for the combined reliability systems, which brings out a useful tool for understanding the working and failure mechanisms, and system state transition ways, for the combined reliability systems. To summarize the reliability calculation, a procedure for calculations of the combined reliability systems is presented, which is an efficient guidance for these reliability computations. Finally, two numerical examples are given to illustrate the computation procedure in which the detailed steps, special on using Finite Markov Chain Imbedding Approach (FMCIA), are shown. We believe that the present paper can shed light on researches on combined reliability systems. [ABSTRACT FROM AUTHOR]