Fuzzy logic-based approach for identifying the risk importance of human error
Introduction
The reliability and safety assessment of operational systems should not only focus on hardware failure but also include human error. A study by Trager (1985) showed that 50–70% of the risk at nuclear power facilities was because of human errors. In a large-scale and complex industrial system, human is prone to produce various errors by the effects of error-forcing conditions. If a potential human error has a high occurrence probability or potential severe effects, this error is termed critical human error. To prevent and reduce human errors, it is important to identify these potentially critical human error modes by human error risk assessment.
A variety of human error identification (HEI) techniques have been developed for identifying critical human errors. Kirwan (1998) outlined and reviewed 38 approaches of human error identification, categorizing them into many types of error identification approach. These also include first generation and second-generation human reliability analysis (HRA) methods. The “first generation” method of HRA, like technique for human error rate prediction (THERP) (Swain and Guttmann, 1983), accident sequence evaluation program (ASEP) (Swain, 1987), which is a simplified version of the THERP, and human cognition reliability (HCR) (Hannaman et al., 1985), success likelihood index methodology (SLIM) (Embrey, 1984), and the human error assessment and reduction technique (HEART) (Williams, 1992), are based on a fact that human has inherent deficiencies just like mechanical or electrical components. In first generation human reliability analysis, operator actions are broken into sub-tasks up to a defined degree of resolution. Most of the basic human error probabilities (HEPs) are given by expert judgments and then they are modified by the factors representing the effects of the environment in the scope of uncertainty. Those factors are called Performance Shaping Factors (PSFs) or Performance Influencing Factors (PIFs). The second-generation method like cognitive reliability and error analysis method (CREAM) (Hollnagel, 1998), a technique for human error analysis (ATHEANA) (Cooper et al., 1996), SPAR-H (Gertman et al., 2005) and MDTA (Kim et al., 2005, Kim et al., 2008) are based on the cognitive model of human decisions and actions. They attempt to identify Errors of Commission (EOC) and incorporate contextual factors into their qualitative and quantitative analyses. All these methods are well suited for supporting basic or generic Quantitative Risk Assessment (QRA). They provide the probabilities of human errors and thus meet the primary requirement of reliability analysis. However, all these methods focus strongly towards quantification, in terms of success/failure of action performance, with lesser attention paid to the effects of individual human error on system. These result in limitations in the discovery of real critical human error modes, and do not satisfy the objective of system safety or risk assessment.
Some researchers have studied the above issues. For instance, Whittingham and Reed (1989) developed the Human Error Mode Effect and Criticality Analysis (HEMECA) to identify the prioritization of human error modes on the baisis of the principle of hardware-oriented failure mode and effect analysis (FMEA). Yu et al. (1999) also developed the Human Error Criticality Analysis (HECA) method. It is used to identify the potentially critical human errors and tasks in the human operation system by constructing human error criticality matrix. Its horizontal axis and vertical axis are respectively the criticality index number (i.e. the HEP multiplied by the EEP) of human error modes and safety or cost severity classification. It considered not only the HEP, but also the Error-Effect Probability (EEP) and Error Consequence Severity (ECS). These three indices are integrated into the human error risk assessment model to assess the risk prioritization of human errors or tasks. However, the above methods do not take the relative weights of the HEP, EEP and ECS into account. They cannot define the risk importance (i.e. risk magnitude or risk criticality) of human errors for the lack of the classification of Risk Criticality Level (RCL). In addition, Gertman et al. (2001) and Lee et al. (2004) used Conditional Core Damage Probabilities (CCDPs) to measure human error contribution to risk in operating events by statistical analysis of event reports. However, This kind of method do not considers the effects of individual human error on system, and requires a lot of event reports.
Human error risk assessment is a process to determine the risk magnitude of each human error mode to assist decision-making. The reliability of results of risk assessment highly relies on the correctness of the risk model, the availability and accuracy of the risk data. However, risk assessors often face the circumstances where the risk data are incomplete or accompanied by high uncertainty. For example, one of the major criticisms of current HRA techniques is the need for expert judgment to evaluate HEP (Kim, 2001, Mosleh and Chang, 2004). Additionally, in many circumstances, the effects of human error modes on system cannot be explicitly evaluated because of the complex structures and functions of the system, and the complex interactions between human and machines. Therefore, it is necessary to develop a new human error risk assessment method which can model the uncertainty to identify critical human errors. Under such conditions, fuzzy logic approaches are very practical. The fuzzy logic method can better simulate the complicated process and treat qualitative or imprecise or vague knowledge and information (Klir and Yuan, 1995). When the available information from the process is qualitative, inexact, vague or uncertain, the notion of the membership function utilized by fuzzy theory is then most adequate for depicting this knowledge. Therefore, the fuzzy logic methodology provides a tool for directly working with the linguistic terms used in making the risk factor assessment, and has currently had many applications in safety and risk analysis field such as system reliability and risk assessment (Bowles and Pelaez, 1995, Sii et al., 2001; Yadav et al., 2003, Guimaraes and Lapa, 2007, Markowski et al., 2009) and human reliability analysis (Onisawa, 1988, Cai et al., 1991, Auflick, 1999, Kim and Bishu, 2006, Kim et al., 2006, Konstandinidou et al., 2006; Marseguerra and Zio Enrico Librizzi, 2007, Zioa et al., 2009), etc. The problem is that they neither consider the risks caused by human error nor the effects of human errors on system. Thus this paper proposes a fuzzy logic-based comprehensive framework to assess the risk of human error and determine the risk importance of human error.
The paper is organized as follows. Section 2 briefly introduces the basic components of fuzzy logic system. Section 3 describes a comprehensive methodology of assessing the risk of human error in human operational system, which includes three stages: the preliminary phase, the measure phase of risk indices and the fuzzy inference phase. Section 4 presents a case example to demonstrate the proposed approach. Section 5 presents some concluding remarks.
Section snippets
Short description of fuzzy inference system
Fuzzy logic was originally introduced by Zadeh (1965) as a mathematical way to represent vagueness in everyday life. In contrast to classical logical systems, fuzzy logic considers modes of reasoning that are approximate rather than exact. Fuzzy logic starts with the concept of a fuzzy set. A fuzzy set is a set without a crisp, clearly defined boundary. The fundamental difference between fuzzy logic and conventional modeling techniques is on the definition of sets. Traditional set theory is
Risk assessment model of human error
This paper constructs risk assessment model of human error on the basis of fuzzy approximate inference as shown in Fig. 3. It includes the following stages: (1) The preliminary phase. (2) The measurement phase of risk indices of human error. (3) Fuzzy inference phase.
Case study
After an initiating event in a nuclear power plant, operators should respond to the emergency accident and the errors might take place because of the effects of context on human activities. A case of steam generator tube rupture (SGTR) accident in a PWR nuclear power plant (Zhang, 2006) is used to demonstrate the proposed identification method of fuzzy logic-based risk importance of human error.
Conclusion and discussion
Human error is the main reason that leads to accident occurrence. Therefore, the pressing problem is how to identify the critical human error and the risk importance of human errors for purposely preventing the occurrence of human errors. This paper presents a new human error risk assessment method based on fuzzy logic to determine risk importance of human error. The conclusions are obtained as follows:
- (1)
In many situations, human error risk analysis is a complex task which is of great uncertainty
Acknowledgements
The paper was supported by National Natural Science Foundation Program of China (70873040). We would like to acknowledge and thank those who provide data and suggestions. The anonymous reviewers and the editor of this paper are also gratefully acknowledged for their constructive comments and suggestions.
References (38)
- et al.
Fuzzy logic importance of failures in a system failure mode, effects and criticality analysis
Reliability Engineering and System Safety
(1995) - et al.
Fuzzy inference to risk assessment on nuclear engineering systems [J]
Applied Computing
(2007) Human error identification techniques for risk assessment of high risk systems – part 1: review and evaluation of techniques
Applied Ergonomics
(1998)- et al.
A systematic approach to analysing errors of commission from diagnosis failure in accident progression
Reliability Engineering and System Safety
(2005) - et al.
The MDTA-based method for assessing diagnosis failures and their risk impacts in nuclear power plant
Reliability Engineering and System Safety
(2008) Human reliability analysis in the man–machine interface design review
Annals of Nuclear Energy
(2001)- et al.
A probabilistic approach for determining the control mode in CREAM
Reliability Engineering and System Safety
(2006) - et al.
A fuzzy modeling application of CREAM methodology for human reliability analysis
Reliability Engineering and System Safety
(2006) - et al.
Model-based human reliability analysis: prospects and requirements
Reliability Engineering and System Safety
(2004) - et al.
Fuzzy logic for process safety analysis
Journal of Loss Prevention in the Process Industries
(2009)
An approach to human reliability in man–machine systems using error possibility
Fuzzy Sets and Systems
A fuzzy-logic-based approach to qualitative safety modeling for marine systems
Reliability Engineering and System Safety
A Technique for Human Error Analysis. NUREG/CR-6350
The SPAR-H Human Reliability Analysis Method, NUREG/CR-6883
Cited by (65)
How to account artificial intelligence in human factor analysis of complex systems?
2023, Process Safety and Environmental ProtectionExploring the effect of basalt fibers on maximum deviator stress and failure deformation of silty soils using ANN, SVM and FL supported by experimental data
2022, Advances in Engineering SoftwareCitation Excerpt :FL is based on the concept of fuzzy sets. A fuzzy set does not have a clearly defined boundary [44]. In traditional logic, an object is either included or not in a set.
Estimation of the likelihood of severe accident management decision-making using a fuzzy logic model
2020, Annals of Nuclear EnergyAllocation of performance shaping factors in the risk assessment of an offshore installation
2020, Journal of Loss Prevention in the Process IndustriesUsing stakeholders’ judgement and fuzzy logic theory to analyze the risk influencing factors in oil and gas pipeline projects: Case study in Iraq, Stage II
2020, International Journal of Critical Infrastructure ProtectionCitation Excerpt :The results of the survey (which are probability and severity) were used as inputs for this model. In accordance with the conventional fuzzy systems, the process of analyzing RIFs is achieved over three stages namely, fuzzification, knowledge base, and defuzzification, as shown in Fig. 2 [13,27,29]. Fuzzification is about providing crisp inputs for the FIS by generating sets of membership functions.