Achieving Consistency in the Use of Human Factors Analytic Methods for Reliability Assessment

Abstract

In highly complex systems, the human has a role to play in optimising system availability, maintenance and safety. However within modern technological systems it is also recognised that human error can be a major contributor to system failures. There are a range of techniques available to quantify and predict human error. Many of these methods are used in parallel with human factors task analysis to identify the performance goals and objectives of the human. A problem exists in the use of these methods. Often there are inconsistencies in the analysis due to variations in interpretation by analysts. So analyses can result in different, but equally valid, inputs into the human factors design of the system. This paper aims to explore the reasons why variability exists and to look at potential solutions to overcome discrepancies between analysts. Case studies from the high risk industries and defence industries are used to illustrate. Analysts with varying ranges of experience are examined. The difficulties encountered are assessed and the approaches used to resolve them outlined.