A large number of methods have been proposed and some have been widely used to quantify human error rates particularly for probabilistic risk analyses of technological systems. In recent years investigators andpractitioners have pointed out a number of major limitations of these models, including proper handling of errors of commission. At the same time new approaches that attempt to model human-systems interactions based on a more fundamental understanding of the causes and mechanisms of human error, are mostly in their early stages of development and often require a different structure of risk and reliability ananlysis models. There are also instances of application of "evolutionary models which offer improved qualitative insights but do not yet provide more reliable quantitative measures.

The purpose of this workshop is to discuss the required features and characteristics of a practical yet theoretically sound evolutionary approach that bridges the gap beteween current methods and the new advanced models. The specific objectives include:

• Development of a clear picture of the state of the art in human reliability analysis.
• Develo0ment of an understanding of those facets of human cognition and behavior that need to be incorporated in evolutionary models.
• Ientification of the general characteristics and data requirements of evolutionary models.

Main	Organization	Sponsors	Publication