HUMAN ERROR ANALYSIS IN RAILWAY SAFETY CASES; PANACEA OR POISON?

摘要

This paper outlines the techniques applied and lessons learnt when applying formal human error and reliability techniques to support the production of a railway safety case in the UK. This paper argues that the optimum techniques for addressing human error may need to be adapted from those advocated by human factors professionals in order to achieve a pragmatic, valid and justifiable safety argument within the often limited timescales and costs. In particular, during this work it was found that formal classification of errors did not assist in identifying human error events or specifying mitigations against these events. In addition the application of formal human reliability techniques did not adequately address the potential for human violation (Williams, 1997). It is concluded that formal methods can only be used as supporting techniques to a qualitative approach and that quantified human reliability techniques should not be used definitively due to their inherent weaknesses.