SE Validation Limited

V&V and Safety of Systems Involving Human Factors (HF)

Introduction

Humans can have an enormous impact on the performance of any system, so the potential beneficial and detrimental impacts from the human need to be understood and recorded. It can be argued that the human operator should be more important in a safety case than the designed equipment. As task complexity increases, so does the probability of a human error. This may be error on the part of the equipment developers, or the operators, or both. The discipline of human factors seeks to ensure that the contribution of errors that the human makes to potential accidents are as low as reasonably practicable (ALARP). Often the focus of human factors analysis is on the error likelihood, but we believe it is equally important to remember, promote and encourage the reliability of human operators in whole system assessment.

Identification

Human error identification techniques are numerous, and there are many papers on each technique. They may be categorised by type of approach and five broad classifications have been developed; taxonomies, psychologically based, cognitive modelling, cognitive simulations and reliability oriented. Several analytical-method classifications are also derived; check-lists, flowcharts, group-based, psychological, representation, cognitive, task analysis, affordance-based, commission identification and crew interactions. We generally recommend that customers use a combination of techniques and analytical methods.

Quantification

Similarly, there are many quantification techniques e.g. HEART and THERP. Consistently exact values of human error rates to three or four significant figures (as may be available for material reliability), are not possible. Human behaviour is not that regular. Often however, that is not the principle requirement. This may be more on the lines of getting data that is useful for the purpose i.e. for comparative purposes, or simply to determine values to better than a one significant figure 'guestimate'. At SE Validation, we are very aware of the role that humans can have in system capability and system fallibility. We have direct experience of many of the techniques noted above and can therefore help customers select those most appropriate for identifying key tasks and the errors that may occur and quantifying the likelihood of occurrence. This helps customers reduce the impact and probability of inevitable human error.

Papers

'Quantification of military helicopter aircrew human error probabilities for safety models', MOD Equipment Safety Assurance Symposium, October 2003.

'Validating a Process for Understanding Human Error Probabilities in Complex Human Computer Interfaces', Coping with Complexity Workshop, Glasgow University, March 2005.