مدل فازی فرایند تحلیل شبکه ای (ANP) برای شناسایی ریسک رفتار معیوب (FBR) در سیستم کار
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|6083||2008||13 صفحه PDF||سفارش دهید||محاسبه نشده|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Safety Science, Volume 46, Issue 5, June 2008, Pages 771–783
Work system safety is a function of many factors, besides it is dynamic and complex. There may be relations and dependencies among the safety factors. Therefore, work system safety should be analyzed in a holistic manner. In this study, the faulty behavior risk (FBR) which is significant in work system safety is tried to be determined through analytical network process (ANP) which is an extension of analytical hierarchy process and allows analysis of complex systems. Besides, there are many difficulties and limitations in measuring the faulty behavior factors. For this reason, the weights of factors and sub-factors necessary to calculate the FBR are determined by using fuzzy ANP and by this way it was possible to make better decisions in this process.
In the work system, fault is usually defined as the deviation from the reasonable, designed and expected behavior standard (Ozgüven, 2003). Faulty behaviors are the behaviors which either decrease or have the potential to decrease the safety and performance of the work system. Unwanted or inappropriate behaviors are also considered faulty behaviors (Sabancı, 1999). Work systems are usually classified into three groups, namely manual, mechanical and automated systems, according to the ergonomics aspects (McCormick, 1982). No matter how complex and comprehensive the work systems are, the human and machinery factors determine the performance and running of the work systems (Kurt, 1993). Synchronous and harmonious running of the human and machine factors in the work system depends on the safety performance of the work system. The safety performance is possible if only there is an effective safety management. In this context, measuring the probable risks of factors that cause faulty behaviors in the work systems is important for the safety performance. In the literature of the safety management, there are many methods proposed or developed in order to measure the safety performance. The main methods are the accident statistics, the accident control chart, and the attitude scales (Brauer, 1994). However, most of these methods are passive or subjective methods. For example, the accident statistics only shows the performance of the safety management in the past, whereas the accident control chart demonstrates the current state of the safety management. The attitude scale includes a measurement which is both subjective and qualitative. For these reasons, none of these methods can evaluate the current state of the safety management quantitatively and estimate the occurrence probability of unwanted events (Chen and Yang, 2004). However, in his study, Qien, 2001a and Qien, 2001b has taken a broad step towards the development of a risk indicator that capable of providing a signal or warning. His methodology includes an eight-step procedure including identification of risk influencing factors, assessment of potential change in risk influencing factors, assessment of effect of change in risk, and selection of risk indicators (Chen and Yang, 2004). In another study, which was carried out to determine the current safety performance indicator of the work system, it is proposed to link the safety climate with the actual characteristics of the workplace obtained from worker questionnaires (Dedobbeleer and Beland, 1991). Williamson et al. (1997) linked the safety climate with the perception of workers about their work environments through, again, questionnaires. Both approaches have yet proved their effectiveness in forecasting or predicting incident occurrence. In fact, it seems to date no proven indicator, methodology or any mean are known to be capable of predicting the occurrence of incidents in a workplace (Chen and Yang, 2004). Chen and Yang (2004) who observed the deficiency of the studies in the literature, have developed a predictive risk indicator (PRI) based on unsafe acts or conditions in a petrochemical plant. The unsafe observation results are quantified by a simple rating based on estimates of probability of danger (PD), frequency of work exposure (FE), number of persons at risk (RN) and maximum of probable loss (MPL). The ratings are combined according to the geometric average to give the risk index. As it can be seen from the studies mentioned above, problem of work system safety has been analyzed from different point of views in the literature. However, as stated by Chen and Yang (2004), the safety in a work system is always dynamic and depends not only on the perception of workers but also on many other complicating factors such as the management enforcement of safety regulations, worker’s attitude toward safety, workplace ergonomics, etc. Work system safety management is a function of many factors (Grote and Künzler, 2000, Champoux and Brun, 2003, Fang et al., 2004 and Mearns et al., 2003) and at the same time a factor of the work system could affect another factor. Moreover, there may be inter-relations among the factors of work system. For this reason, work system safety should be analyzed from a holistic point of view. The main difference of this study from the others in the literature is its modeling of work system safety problem in a holistic manner. Just as in this study, analytic network process (ANP), which is an extension of analytic hierarchy process (AHP) and allows analysis of complex systems, is used to determine the faulty behavior risk of the work system. Additionally, difficulties or limitations can be encountered many times while measuring the risk levels of work system factors. For example, it is not possible to measure qualitative factors such as safety culture, sensory adaptation, tendency of risky behavior, competition, management–worker relationships, exactly. Therefore, measuring qualitative factors by using fuzzy numbers instead of using crisp numbers helps both making decisions easier and obtain more realistic results. In this context, fuzzy ANP is used to determine the weights of factors/sub-factors which are required for computation of FBR in this study.
نتیجه گیری انگلیسی
Faulty behaviors may cause results which affect operation of work system, continuity of production and health of workers negatively. Therefore it is necessary to determine the factors which cause faulty behaviors and the risk level of the work system due to these factors in the context of safety and performance of work system. In this study, an early warning model which will forecast the faulty behavior risk in the work system is proposed. Since faulty behavior causing factors have a complex structure, ANP approach is used in the proposed FBR model. Using the proposed FBR model possible risk of the work system can be determined before it occurs. Therefore, decisions could be made according to the risk level calculated by the FBR model. If the calculated risk level is not appropriate for the work system safety, either corrective actions are taken in the work system or the work system is stopped and redesigned. FBR of two work systems are determined by using the model proposed in this study. It is observed that organizational factors have the highest weight among the basic factors which cause faulty behaviors in the work system and it is followed by job related factors and personal factors. Task context, work flow and monotony have the highest three global weights among the sub-factors of faulty behaviors in the given order. All three sub-factors are related to the job. This shows that, it is important to design the work flow and context of work system in a way that decreases FBR. Although the application of the model proposed in this study is specific to a facility, it can be modified so that it could be used in other firms. Besides, FBR model proposed in this study, includes the dependencies and relations among the main factors of the first stage. In further research, the relations among sub-factors which are in the third stage of FBR model can be analyzed via fuzzy ANP.