Fuzzy techniques for assessing workers' risk exposure – acknowledging the need

  • Celina Pinto Leão
  • Susana Pinto da Costa


Risk assessment (RA) is compulsory in Portuguese companies (Law n.º 7/2009, Law n.º 3/2014) consisting of risk analysis and risk evaluation. Risk management (RM) stands for the global process of RA and risk control, whereby measures are taken to monitor or lower the risk levels to acceptable values (Miguel, 2010). RM is a compulsory and valuable tool helping companies to set strategies in order to achieve goals through informed decision-making (ISO 31000). The techniques for RA (ISO 31010) can be selected according to data type (quantitative or qualitative) and data availability. Quantitative RA techniques may be very laborious, expensive and may not be feasible when available data is scarce. Qualitative RA may be very expeditious, but not provide enough detail on the risks that the workers are exposed to. Semi-quantitative RA techniques seems to be a good compromise in the risk indices estimation. For applying these techniques, it is necessary to construct the hierarchy scale of Probability, Severity and Risk Index. The number of classes of each variable may vary between different techniques: Probability range from Impossible to Very Likely in three or more well-defined classes. But are these classes’ boundaries really well-defined as crisp?

Fuzzy approach is useful when dealing quantitatively with imprecision (Varela, et al., 2013). Fuzzy logic (FL) is a mathematical technique that enables computational models of real systems to act on vagueness, enabling the analyst to preserve coherence in his evaluation by making judgements using imprecise information (Rodrigues, et al., 2016; Costa & Arezes, 2012). FL has been applied in a wide range of fields being a valuable tool for professionals who deal with human factors and human contingencies (Costa, 2013), and relevant in both RA and decision-making process of which workers’ safety and health rely on. This work aims at evidencing the pertinence of the use of FL theory to semi-quantitative RA techniques. To accomplish this a systematic literature review (SLR) was conducted.