Bayesian risk-based decision method for model validation under uncertainty

This paper develops a decision-making methodology for computational model validation, considering the risk of using the current model, data support for the current model, and cost of acquiring new information to improve the model. A Bayesian decision theory-based method is developed for this purpose, using a likelihood ratio as the validation metric for model assessment. An expected risk or cost function is defined as a function of the decision costs, and the likelihood and prior of each hypothesis. The risk is minimized through correctly assigning experimental data to two decision regions based on the comparison of the likelihood ratio with a decision threshold. A Bayesian validation metric is derived based on the risk minimization criterion. Two types of validation tests are considered: pass/fail tests and system response value measurement tests. The methodology is illustrated for the validation of reliability prediction models in a tension bar and an engine blade subjected to high cycle fatigue. The proposed method can effectively integrate optimal experimental design into model validation to simultaneously reduce the cost and improve the accuracy of reliability model assessment.     

Detection of equipment aging and determination of the efficiency of a corrective measure

When too many failures occur on a given piece of equipment, the dependability engineer needs to decide whether these failures are attributable to poor initial design or if they are due to a phenomenon of aging. If aging is confirmed, the problem is then to determine the moment at which the process began and what corrective measure (generally, a modification in the design or in the preventive maintenance program) is the best suited to delay the occurrence of the failure. This measure will thus make it possible to extend the lifetime of the equipment.The method is based on the simple hypothesis of a model of step aging and on Bayesian techniques.The principal benefit of this method is the determination of the time at which aging begins (and the related uncertainties), the evolution in the failure rate of the component in its initial state and once modified, and the probability of success of the corrective measure. The IBTV software was developed to implement this methodology.     

A Dynamic Risk Assessment Methodology for Maintenance Decision Support

The failure mode and effect analysis (FMEA) is a widely applied technique for prioritizing equipment failures in the maintenance decision‐making domain. Recent improvements on the FMEA have largely focussed on addressing the shortcomings of the conventional FMEA of which the risk priority number is incorporated as a measure for prioritizing failure modes. In this regard, considerable research effort has been directed towards addressing uncertainties associated with the risk priority number metrics, that is occurrence, severity and detection. Despite these improvements, assigning these metrics remains largely subjective and mostly relies on expert elicitations, more so in instances where empirical data are sparse. Moreover, the FMEA results remain static and are seldom updated with the availability of new failure information. In this paper, a dynamic risk assessment methodology is proposed and based on the hierarchical Bayes theory. In the methodology, posterior distribution functions are derived for risk metrics associated with equipment failure of which the posterior function combines both prior functions elicited from experts and observed evidences based on empirical data. Thereafter, the posterior functions are incorporated as input to a Monte Carlo simulation model from which the expected cost of failure is generated and failure modes prioritized on this basis. A decision scheme for selecting appropriate maintenance strategy is proposed, and its applicability is demonstrated in the case study of thermal power plant equipment failures. Copyright © 2016 John Wiley & Sons, Ltd.     

What to do next: Using problem status to determine the course of action

Formal decision support tools are little used in engineering design. This paper explores the reasons for this and presents a method which is tailored to problems characterized by teams of stakeholders with inconsistent views who generate multiple alternatives and criteria, and who work to reach consensus. This method is especially designed to support activity when much of the information is qualitative, immature, and there is a diversity of views. The methodology assists the team in determining which alternative attributes to invest time in refining in their effort to reach consensus. The underlying mathematical structure (a Bayesian model of multi-attribute team decision making) is presented. This model supports team member belief about an alternative's ability to meet a criteria on two dimensions, knowledge and confidence. The methodology forces recording the rationale used to reach the final decision. A running example is used to explain the details.This research has been supported by the National Science Foundation under grant DDM-931996. The opinions in this paper are the authors' and do not reflect the position of the NSF or Oregon State University.     

Analysis of OREDA data for maintenance optimisation

This paper provides estimates for the average rate of occurrence of failures, ROCOF (“failure rate”), for critical failures when also degraded failures are present. The estimation approach is exemplified with a data set from the offshore equipment reliability database “OREDA”. The suggested modelling provides a means of predicting how maintenance tasks will affect the rate of critical failures.     

Using AHP in determining the prior distributions on gas pipeline failures in a robust Bayesian approach

The paper proposes a robust Bayesian approach to support the replacement policy of low-pressure cast-iron pipelines used in metropolitan gas distribution networks by the assessment of their probability of failure. In this respect, after the identification of the factors leading to failure, the main problem is the historical data on failures, which is generally limited and incomplete, and often collected for other purposes. Consequently, effective evaluation of the probability of failure must be based on the integration of historical data and knowledge of company experts. The Analytic Hierarchy Process has been used as elicitation method of expert opinion to determine the a priori distribution of gas pipeline failures. A real world case study is presented in which the company expertise has been elicited by an ad hoc questionnaire and combined with the historical data by means of Bayesian inference. The robustness of the proposed methodology has also been tested.     

The Graphical Analysis for Maintenance Management Method: A Quantitative Graphical Analysis to Support Maintenance Management Decision Making

This article proposes a logical support tool for maintenance management decision making. This tool is called the Graphical Analysis for Maintenance Management (GAMM), a method to visualize and analyze equipment dependability data in a graphical form. The method helps for a quick and clear analysis and interpretation of equipment maintenance (corrective and preventive) and operational stoppages. Then, opportunities can be identified to improve both operations and maintenance management (short–medium term) and potential investments (medium–long term). The method allows an easy visualization of parameters, such as the number of corrective actions between preventive maintenance, the accumulation of failures in short periods of time, and the duration of maintenance activities and sequence of stops of short duration. In addition, this tool allows identifying, a priori, anomalous behavior of equipment, whether derived from its own function, maintenance activities, misuse, or even equipment designs errors. In this method, we used a nonparametric estimator of the reliability function as a basis for the analysis. This estimator takes into account equipment historical data (total or partial) and can provide valuable insights to the analyst even with few available data. Copyright © 2012 John Wiley & Sons, Ltd.     

Modelling maintenance effects on manufacturing equipment performance: results from simulation analysis

Equipment maintenance and system reliability are important factors affecting the organisation’s ability to provide quality and timely services to customer. While maintenance remains an important function to manufacturing, it is only recently that attempts have been made to quantify its impact on equipment performance. In this research, an approach of linking maintenance with equipment performance is developed using simulation modelling. The modelling approach involves defining probabilistic models and assumptions affecting system performance, such as: the probabilistic model for the initial failure rate/intensity of the equipment; the probabilistic model for the system deterioration and corresponding effect; the probabilistic model for the random times of corrective maintenance (CM) and preventive maintenance (PM) that takes into the account the types of maintenance plans/policies and the potential dependency between CM and PM times; and the probabilistic model for the random effects of CM and PM on the reliability of the equipment. Using a continuous manufacturing equipment, the model is used to analyse the impact of deterioration, failures and maintenance (policies, timing and efficiency) on equipment performance. It is shown that modelling the effect maintenance provides a basis of evaluating maintenance efforts with the potential application in performance evaluation and decision support while investigating opportunities for manufacturing equipment performance improvement.     

An option generation and selection methodology for process equipment selection

The option generation and selection (OGS) methodology forms part of a general approach for the design of agile chemical plants based on business, product and process knowledge, with support from information models. This paper describes an equipment OGS tool that encompasses the principles of combinatorial process and plant design. The main components of the methodology are: an equipment option generation model described as a set of objects, and the net relationships between them, and an equipment option selection model which consists of procedures for equipment selection. The two models are supported by databases containing information specific to each equipment type, the concept on which the equipment is based, and relationships with other equipment types. Robust, systematic and complete forms of these models can be used as the basis of an expert system for process equipment design, with equipment selected using these tools satisfying the requirements of both specific processes and families of processes (that contain common features, similar functional groups or similar raw material requirements for process operations). Application of the methodology also allows the evaluation of options for reconfiguring existing plant.     

Risk-based maintenance planning of subsea pipelines through fatigue crack growth monitoring

Research and development in the field of risk-based maintenance of offshore structures has recently attracted large attention due to the significant level of accident risk and the cost associated with maintenance in such remote facilities. The uncertainties associated with the deterioration of these facilities require a sound decision making methodology for maintenance planning. This paper presents a dynamic risk-based methodology for maintenance scheduling of subsea pipelines subjected to fatigue cracks. The developed method can assist the asset managers to select the optimum approach for mitigating the consequences of failure while minimizing the maintenance costs. A Bayesian network is developed to model the probabilistic deterioration process and then it is extended to an influence diagram for estimating the expected utility of each decision alternative. Observation of damage state is included in the model to enhance decision making capacity. To demonstrate the applicability of the methodology, three cases with different fatigue crack incidents on a pipeline are considered. Based on the monitoring results, the model is able to determine whether the maintenance should be performed or not. The economic risk associated with maintenance is also minimized by suggesting the optimum maintenance technique among multiple possible methods such as welding or major repair.     

Fuzzy belief propagation in constrained Bayesian networks with application to maintenance decisions

Bayesian networks have been widely applied to domains such as medical diagnosis, fault analysis, and preventative maintenance. In some applications, because of insufficient data and the complexity of the system, fuzzy parameters and additional constraints derived from expert knowledge can be used to enhance the Bayesian reasoning process. However, very few methods are capable of handling the belief propagation in constrained fuzzy Bayesian networks (CFBNs). This paper therefore develops an improved approach which addresses the inference problem through a max-min programming model. The proposed approach yields more reasonable inference results and with less computational effort. By integrating the probabilistic inference drawn from diverse sources of information with decision analysis considering a decision-maker's risk preference, a CFBN-based decision framework is presented for seeking optimal maintenance decisions in a risk-based environment. The effectiveness of the proposed framework is validated based on an application to a gas compressor maintenance decision problem.     

Determining the optimal maintenance action for a deteriorating repairable system

This paper develops a decision model for risk management of the deterioration of a repairable system. When a failure occurs in a deteriorating system, an optimal maintenance decision that includes the possibility of system replacement, as compared to mere deterioration reduction, should be made. There are many uncertainties associated with deterioration, however, so the decision may require a probabilistic analysis. Here, a well-known nonhomogeneous Poisson process with a power law intensity function is used to model the uncertain behavior of the deteriorating system. A Bayesian statistical approach is adopted to allow for the uncertainty of the parameters of the power law intensity function, which imposes a conjugate prior distribution of the parameters. A power law maintenance cost function and the failure cost are analyzed to determine the magnitude of failure risk reduction by minimizing the expected cost incurred from the maintenance action and future failures. A numerical example is given.     

Using Taguchi design,regression analysis and simulation to study maintenance float systems

This paper proposes a new strategic approach for the maintenance float decision models. This new approach incorporates Taguchi experimental design, Taguchi's ANOVA procedure and regression analysis. Taguchi experimental design is used to generate the input variables into the simulation program. The results are analysed using Taguchi's ANOVA procedure. Input variables found significant are subsequently applied in a regression model. Predictor models for the system utilization are developed and their validity tested. Cost oriented decision models are further developed to show the applicability of our models to decision situations. The major advantage of the strategic approach proposed here is that it saves time and reduces the cost of running simulation. This approach will also offer a decision support and improve the effectiveness of maintenance float decisions.     

Improving postponement operation in warehouse: an intelligent pick-and-pack decision-support system

Today, many manufacturers prefer to shift their pick-and-pack postponement operations to third-party logistics companies to lower operational cost and improve mass customisation flexibility. The type of pick-and-pack operation is complicated, as it usually involves a large number of stock keeping units and several warehouse operations. Surprisingly, in most 3PL, the complicated postponement operations are still mainly managed by an experienced warehouse manager, which is an unreliable methodology and often creates many mistakes. A hybrid intelligent system integrating Case-Based Reasoning and Fuzzy Logic is proposed to support the manager in making pick-and-pack postponement decisions. The system comprises a solution-refining module that proposes and holds the details of past pick-and-pack operations. This approach will be of benefit to managers as effective decision support of pick and pack planning can be provided. A case study is used to illustrate the applicability and effectiveness of the approach. Implications of the proposed approach are discussed, and suggestions for further work are outlined.     

Probabilistic seismic demand model and fragility estimates for critical failure modes of un-anchored steel storage tanks in petroleum complexes

Estimation of seismic vulnerability of existing equipment in petroleum complexes, traditionally has been done by using general fragility curves which have been developed for a group of equipment without considering their physical specifications. Although these kinds of fragility are suitable for preparing emergency plans and governmental decision making, for the sake of seismic upgrading of existing facilities, specific equipment fragilities are needed. In this paper un-anchored steel storage tanks, which is among the most vulnerable equipment in petroleum facilities, are selected for a case study. Probabilistic demand models for Elephant Foot Buckling (EFB) and welding failure at the connection between the bottom plate and shell are developed based on FEM analytical data, and a Bayesian updating rule is used to assess the unknown demand model parameters. The approach properly accounts for both aleatory and epistemic uncertainties. Developed probabilistic demand models are used to estimate the fragility of critical failure modes of selected tanks and real world data. The approach can be applied for other failure modes and other equipment, as well.     

Modelling the reliability of search and rescue operations with Bayesian Belief Networks

This paper uses a Bayesian Belief Networks (BBN) methodology to model the reliability of Search And Rescue (SAR) operations within UK Coastguard (Maritime Rescue) coordination centres. This is an extension of earlier work, which investigated the rationale of the government's decision to close a number of coordination centres. The previous study made use of secondary data sources and employed a binary logistic regression methodology to support the analysis. This study focused on the collection of primary data through a structured elicitation process, which resulted in the construction of a BBN. The main findings of the study are that statistical analysis of secondary data can be used to complement BBNs. The former provided a more objective assessment of associations between variables, but was restricted in the level of detail that could be explicitly expressed within the model due to a lack of available data. The latter method provided a much more detailed model, but the validity of the numeric assessments was more questionable. Each method can be used to inform and defend the development of the other. The paper describes in detail the elicitation process employed to construct the BBN and reflects on the potential for bias.     

Challenges to Bayesian decision support using morphological matrices for design: empirical evidence

A novel Bayesian design support tool is empirically investigated for its potential to support the early design stages. The design support tool provides dynamic guidance with the use of morphological design matrices during the conceptual or preliminary design stages. This paper tests the appropriateness of adopting a stochastic approach for supporting the early design phase. The rationale for the stochastic approach is based on the uncertain nature of the design during this part of the design process. The support tool is based on Bayesian belief networks (BBNs) and uses a simple but effective information content–based metric to learn or induce the model structure. The dynamically interactive tool is assessed with two empirical trials. First, the laboratory-based trial with novice designers illustrates a novel emergent design search methodology. Second, the industrial-based trial with expert designers illustrates the hurdles that are faced when deploying a design support tool in a highly pressurised industrial environment. The conclusion from these trials is that there is a need for designers to better understand the stochastic methodology for them to both be able to interpret and trust the BBN model of the design domain. Further, there is a need for a lightweight domain-specific front end interface is needed to enable a better fit between the generic support tool and the domain-specific design process and associated tools.     

Optimal Bayesian maintenance policy for a gearbox subject to two dependent failure modes

Most maintenance optimization models of gear systems have considered single failure mode. There have been very few papers dealing with multiple failure modes, considering mostly independent failure modes. In this paper, we present an optimal Bayesian control scheme for early fault detection of the gear system with dependent competing risks. The system failures include degradation failure and catastrophic failure. A three‐state continuous‐time–homogeneous hidden Markov model (HMM), namely the model with unobservable healthy and unhealthy states, and an observable failure state, describes the deterioration process of the gear system. The condition monitoring information as well as the age of the system are considered in the proposed optimal Bayesian maintenance policy. The objective is to maximize the long‐run expected average system availability per unit time. The maintenance optimization model is formulated and solved in a semi‐Markov decision process (SMDP) framework. The posterior probability that the system is in the warning state is used for the residual life estimation and Bayesian control chart development. The prediction results show that the mean residual lives obtained in this paper are much closer to the actual values than previously published results. A comparison with the Bayesian control chart based on the previously published HMM and the age‐based replacement policy is given to illustrate the superiority of the proposed approach. The results demonstrate that the Bayesian control scheme with two dependent failure modes can detect the gear fault earlier and improve the availability of the system.     

PLM Software Selection Model for Project Management Using Hierarchical Decision Modeling With Criteria From PMBOK® Knowledge Areas

Abstract:

This article presents a methodology for product lifecycle management (PLM) software selection utilizing the nine knowledge areas of the Project Management Body of Knowledge (PMBOK®) (A Guide to the Project Management Body of Knowledge (PMBOK® Guide, 2008). The methodology utilizes a five-process gate approach where the PLM offerings are researched, sorted, paired down, evaluated, and implemented. The primary decision model utilizes hierarchical decision modeling (HDM). Prioritizations are assigned to the PMBOK® knowledge areas utilizing a pair-wise comparison survey and then assessed across the different PLM system offerings. The proposed decision methodology is meant to serve only as a guide to assist with selection of PLM software from a project management perspective. To validate the model, several applications are presented that apply the selection methodology to different industries: semiconductor, information technology (IT), and automotive supply.     

A Bayesian stochastic programming approach to an employee scheduling problem

Bayesian forecasting models provide distributional estimates for random parameters, and relative to classical schemes, have the advantage that they can rapidly capture changes in nonstationary systems using limited historical data. Unlike deterministic optimization, stochastic programs explicitly incorporate distributions for random parameters in the model formulation, and thus have the advantage that the resulting solutions more fully hedge against future contingencies. In this paper, we exploit the strengths of Bayesian prediction and stochastic programming in a rolling-horizon approach that can be applied to solve real-world problems. We illustrate the methodology on an employee production scheduling problem with uncertain up-times of manufacturing equipment and uncertain production rates. Computational results indicate the value of our approach.