Risk, Human Error, and System Resilience: Fundamental Ideas

Abstract

I review and critique basic ideas of both traditional error/risk analysis and the newer and contrasting paradigm of resilience engineering. Analysis of human error has matured and been applied over the past 50 years by human factors engineers, whereas the resilience engineering paradigm is relatively new. Fundamental ideas and examples of human factors applications of each approach are presented and contrasted. Probabilistic risk analysis provides mathematical rigor in generalizing on past error events to identify system vulnerabilities, but prediction is problematical because (a) error definition is arbitrary, and thus it is difficult to infer valid probabilities of human error to input to quantitative models, and (b) future accident conditions are likely to be quite different from those of past accidents. The new resilience engineering paradigm, in contrast, is oriented toward organizational process and is concerned with anticipating, mitigating, and preparing for graceful recovery from future events. Resilience engineering complements traditional error analysis but has yet to provide useful quantification and operational methods. A best safety strategy is to use both approaches.