HUMAN ERROR IN STRUCTURAL RELIABILITY - II: REVIEW OF MATHEMATICAL MODELS

A review of some pertinent literature is made to show that there is now sufficient evidence that classical reliability calculations are not able to correctly predict the incidence of structural failure or malfunction. The main factor which has not been included in such analyses is human error. Interest in the importance of human error, whether of the gross or variability type, has increased considerably in recent years. Several approaches have been suggested for predicting the effect of human error: herein only statistically based models are considered. An attempt is made to compare and critically review the various suggested techniques, and a further model developed in outline. For the abstracts of parts I and III see TRIS nos 394165 and 394167. (Author/TRRL)