Measurement-based techniques for evaluation of human interfaces

Summary form only given. In order to enhance objectivity and realism of human reliability assessment (HRA) methods, the Electric Power Research Institute (EPRI) initiated a program to systematically collect data on reliability of control room crew actions during simulator training exercises. The operator reliability data have been used to modify the original human cognitive reliability model (HCR), which feature a skill-, rule-, and knowledge-based classification of human actions to a procedure-based cue-response classification. This improved approach allows for the explicit integration of the impact of operating procedures into probabilistic risk analyses (PRAs). Insights gained from the data are being used by EPRI to provide additional guidance to utility analysts on the detailed logic modeling of human actions and its integration with system logic models in the SHARP1 (systematic human action reliability procedure) framework.<>