Analysis and Consequences on Some Aggregation Functions of PRISM (Partial Risk Map) Risk Assessment Method

Ferenc Bognár,Csaba Hegedűs

doi:10.3390/math10050676

Ferenc Bognár, Csaba Hegedűs

Open Access

https://doi.org/10.3390/math10050676

Copy DOI

Abstract

The PRISM (partial risk map) methodology is a novel risk assessment method developed as the combination of the failure mode and effect analysis and risk matrix risk assessment methods. Based on the concept of partial risks, three different aggregation functions are presented for assessing incident risks. Since the different aggregation functions give different properties to the obtained PRISM numbers and threshold surfaces (convex, concave, linear), the description of these properties is carried out. Similarity analyses based on the sum of ranking differences (SRD) method and rank correlation are performed and robustness tests are applied related to the changes of the assessment scale lengths. The PRISM method provides a solution for the systematically criticized problem of the FMEA, i.e., it is not able to deal with hidden risks behind the aggregated RPN number, while the method results in an expressive tool for risk management. Applying new aggregation functions, proactive assessment can be executed, and predictions can be given related to the incidents based on the nature of their hidden risk. The method can be suggested for safety science environments where human safety, environmental protection, sustainable production, etc., are highly required.

Highlights

Nowadays, the development of risk assessment methodologies is clearly visible in the industry and service sector as well
Descriptive statisticss of the set of partial risk map (PRISM) numbers produced by the presentedd aaggggrreeggaattioionnfufunnctcitoionnssanadndthtehseest eotf otrfatdriatdiointiaolnRaPl NRPnNumnbuemrsbaeres dareescdriebsecdriabned aconmd pcaormedp.aSroemd.eSkoemy erelkaetiyonrselbaetitownesenbethtwe aeegngrethgeatiaogngfruengcattiioonns foufnthcteioPnRsISoMf tmhethPoRdISaMre malseothpordesaerneteadlsion tphriessseenctteidoni.nRtohbius sstencetsisonte.sRtsoobfutshtenersasnkteinstgss obfastehde ornanPkRinISgMs bnausmedbeorns PdRetIeSrMminnuemd byerdsidffeetreernmt iangegdrebgyadtiiofnfefruentctaigognrseagraetidoenscfruinbcetdioansswareell.described as well
Changing to the PRISM aggregation functions from RPN(m), the number of distinct vTaablulees1.(DNe)scdrripotpivseosftaf tfirstoicms r1e2la0tetdoto19th, e4d2i,ffaenrednt5P2RfIoSrMthneumAb(emrs),aMnd(mthe),traandditiSo(nmal)RvPeNrsniounms, breesrp. ectively, that decrease the variability of the values

Summary

Introduction

The development of risk assessment methodologies is clearly visible in the industry and service sector as well. One typical development direction is to combine different mathematical methodologies with a platform risk assessment methodology such as FMEA—failure mode and effect analysis [1,2], RM—risk matrix [3], HAZOP—hazard and operability analysis [4], FTA—fault tree analysis [5], etc. The typical aim of these studies is to develop the platform methodology, increasing its strengths and/or decreasing its weaknesses by adding new, typically mathematical features. Another major development direction is to combine a risk assessment methodology with another one [6,7,8]. Throughout the decades of development, the reliability, effectiveness, usefulness, applicability, etc., of the platform risk assessment methodologies were significantly increased by dominantly mathematic-based methodological developments [9,10,11,12]

Objectives

Methods

Results

Conclusion