Resumen de The derivation of hazard criteria from historical knowledge
A set of accidents in the offshore engineering industry was analysed in order to generate a set of criteria that could be applied both to designs and design processes to reveal how susceptible they were to hazard. This involved two main tasks: the analysis of causation around the site of a moderately large number of accidents, and the analysis of causation around the site of a number of flawed design decisions. The accident dataset was taken from accident investigators' reports, while the design decision dataset was taken from practising designers' observations on a set of design reservations raised by a verification authority. In each case, the outcome was a set of criteria: one set concerning the way in which design elements ultimately contributed to accidents, and one set concerning the way in which the design process failed during the course of design activity. A framework in which these criteria could be applied is laid out, and the limitations of this approach in comparison with alternative approaches are discussed.
