ABSTRACT
In this chapter, the authors use the automobile crash example to explain the need for and structure of a causal Bayesian Network (BN). They explain why popular methods such as risk registers and heat maps are insufficient to properly handle risk assessment. The authors describe the causal approach to risk assessment and show that how it overcomes the limitations of the popular methods. They describe a simple causal approach to risk assessment that overcomes basic flaws and limitations of other commonly used approaches. The authors address some of the core limitations of the traditional statistical approaches to risk assessment by using causal or explanatory models. The accuracy of the risk assessment is crucially dependent on the fidelity of the underlying model used to represent the risk; the simple formulation is insufficient. Many risk management standards and guidelines assume that once a mitigation action is put in place that it will never degrade, be undermined.
