Journal of Reliability Science and Engineering

Numerous design optimization methodologies and reliability analysis techniques have been developed to address aleatory and epistemic uncertainties in engineering system design. Aleatory uncertainty is modeled by statistical distributions, while epistemic uncertainty becomes an alternative in cases where data is sparse and cannot be fully captured statistically. Possibility and evidence theories are computationally efficient and robust for quantifying epistemic uncertainty in reliability analysis and design optimization. This paper provides a comprehensive analysis of existing methodologies, challenges, and opportunities in managing uncertainty in engineering systems. Additionally, the concepts and practical applications of possibility and evidence theories are reviewed. Potential future research directions are outlined ultimately. This paper provides the sector with a clear understanding of possibility theory and evidence theory and their developments.

Solar pavement has emerged as an innovative form of building-integrated photovoltaics. Reliability analysis and reliability-based optimal design for solar pavement are infrequent. This research represents the pioneering investigation into the optimal reconfiguration design for solar pavements, considering the interchangeability of solar floor tiles amidst competing risks. A Genetic Algorithm with Munkres Assignment Tuning is proposed to initially determine the Pareto optimality of the reconfiguration design and subsequently enhance the systems reliability through minor adjustments. A comprehensive case study for solar floor tiles with total-cross-tied topology is presented to showcase the main idea.

High reliability and long service life have become the development goals and urgent needs of equipment research and development, especially for important equipment and major projects, which put forward new challenges to the traditional reliability technology. As an effective means to support the high reliability and long life of equipment, accelerated testing (AT) technology has become a hot research topic. On the basis of AT application demand analysis, the technical system of AT is proposed. The current state of research on modelling and analysis methods for AT is reviewed and analysed from five aspects: modelling and analysis of accelerated life tests (ALTs) with a single failure mode, modelling and analysis of accelerated degradation tests (ADTs) with a single failure mode, modelling and analysis of ALTs with multi-failure modes, and modelling and analysis of ADTs with multi-failure modes. And finally, future research directions for techniques related to modelling and analysis of ATs are envisioned.

The complex and costly Earth–Moon logistics necessitates an innovative reliability paradigm to support the deployment and operations of lunar habitation base. To lay the groundwork for the new paradigm, this paper introduces a synergistic approach for optimizing reliability-redundancy, spares inventory, and spacecraft trajectory in the cislunar space. We propose a two-phase planning model to guide the establishment of a long-term human habitat on the Moon. Phase 1 allocates subsystem redundancy and base stock level to meet the reliability and availability goal at low mass cost. Phase 2 optimizes spacecraft fuel type, propellant mass and trajectory to minimize the total launch mass, including the habitat and spare parts of Phase 1. Notably, random spares demand, prolonged resupply, and limited transfer orbits are considered, making it first of its kind in co-optimizing redundancy, spares, rocket staging, and flight route. The redundancy-sparing-trajectory allocation model is demonstrated on the cislunar network with nine nodes and 23 velocity changes. Numerical study shows there exists a strong interdependence of launch mass, inventory lead time, transfer orbit, velocity change, and specific impulse. This study makes an early attempt to integrate space logistics theory into reliability and maintenance planning, potentially opening a new research direction for deploying reliable and cost-effective crewed bases in deep space.

Uncertainty propagation (UP) is a crucial aspect for assessing the influence of input uncertainty on structural responses, holding substantial significance in engineering applications. However, achieving accurate and efficient UP remains challenging, particularly for structures characterized by high nonlinearity and multiple outputs. This study addresses this challenge by proposing a novel adaptive UP method based on the artificial neural network (ANN). In the proposed method, the mean of outputs is analytically derived using the ANN, enabling direct computation of the mean through the weight and bias vectors of the network. An innovative approach is established for solving the standard deviation of outputs, employing several univariate integrals instead of multivariate integrals. The established analytical and univariate integral techniques effectively mitigate post-processing errors commonly encountered when using numerical simulation techniques to estimate the statistical moments of outputs within the ANN context. Furthermore, an adaptive framework is presented, incorporating input space division and an adjustable multi-point addition strategy to enhance computational accuracy in structural UP. Various applications, including highly nonlinear scenarios, multiple outputs, and cases involving finite element models, are presented to demonstrate the effectiveness of the proposed method. The results indicate that the proposed method not only provides accurate estimations of statistical moments but also offers effective estimations of the probability density function of structural outputs.

A mission abort policy (MAP) establishes clear, non-ambiguous criteria that define specific system deterioration conditions for discontinuing a primary mission and initiating a rescue procedure (RP) to survive a valuable system performing the mission. A too-late mission abort can incur low system survivability while a too-early abort may unnecessarily compromise mission success probability (MSP). An optimal MAP should strike a balance between these two performance metrics. Therefore, the optimal design of MAPs plays a crucial role in managing the risk of system losses while ensuring a desired level of MSP for critical systems. This article presents a systematic review of MAP research in reliability engineering, classifying and reflecting on the recent intensive research devoted to the mathematical modeling, analysis and optimization of diverse types of MAPs for both single-attempt and multi-attempt mission systems. Potential directions for advancing the state of the art and the state of practice on MAPs are also outlined.

The Industry 5.0 (I5.0) paradigm is expected to further boost the relevance and widespread application of Artificial Intelligence (AI). The actual contribution that it can bring is challenged by current technology, scientific contexts and trends. The objective of this work is to provide an overview of some of the issues that need to be tackled for AI to support the development of reliability and maintenance engineering for I5.0. Three use cases of AI development opportunities are discussed, which are fully compliant with the EU vision of AI enhancements for I5.0: lumping together data of different origins and scales, expert knowledge combined with AI, and causality-based AI. These use cases show that there is room for advanced and successful AI solutions for I5.0, and allow identifying three main elements required to grasp opportunities while identifying, preventing and mitigating risks related to AI development: multidisciplinarity, experience and continuous training.

The calibration of computational models using experimental or operational data to achieve accurate predictions is widely recognized as a crucial challenge in reliability engineering. Bayesian model updating (BMU) has been developed as an appealing methodological framework to achieve this goal, but existing methods range from very approximate but cheap (e.g. Laplace approximation and conjugate priors), less approximate and a bit cheaper (e.g. approximate Bayesian computation), to quite expensive and highly informative techniques such as full Bayesian computation. The goal of this work is to achieve full Bayesian accuracy at a low cost. The approximate Bayesian quadrature has emerged as a highly appealing scheme to achieve this goal. In this work, we develop a family of new acquisition functions with closed-form expressions to accelerate the approximate Bayesian quadrature for addressing the BMU problem with the desired level of accuracy. The proposed method leverages information revealed by both the mean predictions and the posterior covariance of the probabilistic regression model trained for approximating the likelihood function. It thus provides a better trade-off between exploration and exploitation. Results from both numerical and engineering examples show that the proposed method is applicable to multimodel problems, achieving high accuracy and efficiency.

Quantum sensing utilizes quantum effects, such as entanglement and coherence, to measure physical signals. The performance of a sensing process is characterized by error which requires comparison to a true value. However, in practice, such a true value might be inaccessible. In this study, we utilize quantum reliability as a metric to evaluate quantum sensor's performance based solely on the apparatus itself, without any prior knowledge of the true value. We derive a general relationship among reliability, sensitivity, and systematic error, and demonstrate this relationship using a typical quantum sensing process. That is to measure magnetic fields (as a signal) by a spin-$1/2$ particle and using the Stern–Gerlach apparatus to read out the signal information. Our findings illustrate the application of quantum reliability in quantum sensing, opening new perspectives for reliability analysis in quantum systems.

A scientific discipline is not formed by philosophical and generic commentaries; it looks like a building consisting of solid bricks that are precise principles and mathematical laws. Reliability and risk problems embody a new science whose foundations were initiated by the Russian school, and this inquiry means to continue that theoretical research project. Here we address broad topics that are waiting to be carried out. More precisely, this paper describes the decreasing reliability of the system by means of the Boltzmann-like entropy, which yields various trends depending on the structure and behavior of the system. The features typical of infancy, maturity and senility of systems shape the bathtub curve. Annotations on the actual states of the engineering and biological systems are added.

High reliability and long service life have become the development goals and urgent needs of equipment research and development, especially for important equipment and major projects, which put forward new challenges to the traditional reliability technology. As an effective means to support the high reliability and long life of equipment, accelerated testing (AT) technology has become a hot research topic. On the basis of AT application demand analysis, the technical system of AT is proposed. The current state of research on modelling and analysis methods for AT is reviewed and analysed from five aspects: modelling and analysis of accelerated life tests (ALTs) with a single failure mode, modelling and analysis of accelerated degradation tests (ADTs) with a single failure mode, modelling and analysis of ALTs with multi-failure modes, and modelling and analysis of ADTs with multi-failure modes. And finally, future research directions for techniques related to modelling and analysis of ATs are envisioned.

A mission abort policy (MAP) establishes clear, non-ambiguous criteria that define specific system deterioration conditions for discontinuing a primary mission and initiating a rescue procedure (RP) to survive a valuable system performing the mission. A too-late mission abort can incur low system survivability while a too-early abort may unnecessarily compromise mission success probability (MSP). An optimal MAP should strike a balance between these two performance metrics. Therefore, the optimal design of MAPs plays a crucial role in managing the risk of system losses while ensuring a desired level of MSP for critical systems. This article presents a systematic review of MAP research in reliability engineering, classifying and reflecting on the recent intensive research devoted to the mathematical modeling, analysis and optimization of diverse types of MAPs for both single-attempt and multi-attempt mission systems. Potential directions for advancing the state of the art and the state of practice on MAPs are also outlined.

Quantifying uncertainty and updating reliability are essential for ensuring the safety and performance of engineering systems. This study develops a hierarchical Bayesian modeling (HBM) framework to quantify uncertainty and update reliability using data. By leveraging the probabilistic structure of HBM, the approach provides a robust solution for integrating model uncertainties and parameter variability into reliability assessments. The framework is applied to a linear mathematical model and a dynamical structural model. For the linear model, analytical solutions are derived for the hyper parameters and reliability, offering an efficient and precise means of uncertainty quantification and reliability evaluation. In the dynamical structural model, the posterior distributions of hyper parameters obtained from the HBM are used directly to update the reliability. This approach relies on the updated posteriors to reflect the influence of system uncertainties and dynamic behavior in the reliability predictions. The proposed approach demonstrates significant advantages over traditional Bayesian inference by addressing multi-source uncertainty in both static and dynamic contexts. This work highlights the versatility and computational efficiency of the HBM framework, establishing it as a powerful tool for uncertainty quantification and reliability updating in structural health monitoring and other engineering applications.

Time-variant reliability analysis for a single failure mode has advanced considerably to exhibit great potential. However, in practical engineering applications, structures/systems generally invoke multiple failure modes. It poses as lasting challenge for the current reliability analysis, and simultaneously entails very high, if not intractable, computational cost. Therefore, developing accurate time-variant system reliability model that accounts for multiple failure modes and its commensurate solution strategy is imperative. To this objective, this study develops an innovative method to assess time-variant system reliability by utilizing Kriging model. First, multiple failure modes are decomposed into separate time-variant reliability assessments, each addressing a single failure mode. As inspired by the discretization of stochastic processes, for each time-variant function in the structural system is assumed as a set of time-invariant reliability problems. Subsequently, for each time-invariant reliability issue, the most probable point (MPP) is determined, and the function is linearized and expanded accordingly. On this basis, Kriging model of MPP at discrete time is constructed. To enhance the model's precision, active learning techniques are employed for ongoing model updates. Finally, the reliability index and failure probability are determined by solving the time-invariant system reliability models. The proposed method's applicability and effectiveness are demonstrated through exemplifications of variable systems.

This is the inaugural editorial.